Lithium ion battery detection system for electric energy storage

Abstract

The invention discloses a lithium ion battery detection system for electric energy storage. The lithium ion battery detection system comprises a battery working condition simulation module, a data acquisition and sensing module, a performance evaluation and calculation module, a safety early warning and diagnosis module and a data management and man-machine interaction module. The working condition simulation module generates a simulation charging and discharging instruction based on park load curve superposition random disturbance. The data acquisition module acquires electrical, thermal and environmental parameters with high precision. The performance evaluation module calculates capacity, internal resistance, SOH, SOP and the like, and fine analysis can be carried out through an EIS test and an equivalent circuit model. And the safety early warning module realizes early warning in manners of identifying internal resistance gradient on line, constructing an electric-thermal coupling model to predict temperature rise, defining a dynamic inconsistency index and the like. The data management module stores data, generates a report and can communicate with a park energy management system to report key parameters. According to the invention, real working condition simulation and multi-dimensional performance evaluation of the energy storage battery are realized, and the detection effectiveness and the power station management level are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power detection, in particular to a lithium ion battery detection system for power energy storage. BACKGROUND

[0002] With the increase of renewable energy penetration and the implementation of peak-valley electricity price policy, energy storage power stations equipped with lithium ion batteries have been widely used in industrial and commercial parks to realize demand management, backup power, energy time shift and smooth fluctuation. As a key asset, the performance degradation and safety state of the core component of lithium ion battery cluster directly affect the operation efficiency, economic benefit and safety of the entire power station.

[0003] Currently, the detection of energy storage batteries is mostly carried out in laboratory environment, using standard charge-discharge procedures (such as constant current constant voltage charging, constant current discharging) to evaluate their capacity and internal resistance. However, this detection method has significant shortcomings: first, it fails to simulate the complex working conditions that the battery faces in real park operation, such as frequent and random power scheduling, cycling at partial state of charge, etc., resulting in a disconnection between the evaluation results and the actual operating performance; second, the detection parameters are relatively isolated, lacking coordinated analysis and comprehensive evaluation of battery health status, power status, and safety state; third, the early warning means are single, mostly relying on absolute threshold alarms of voltage and temperature, lacking early warning capability for the gradual change of battery internal resistance, inconsistency between single cells, and abnormal electro-thermal coupling that may lead to thermal runaway; fourth, the detection data lack interaction with upper-level platforms such as park energy management systems, failing to form a closed loop from detection to operation and maintenance.

[0004] Therefore, there is an urgent need to develop a special lithium ion battery detection system that can simulate real operating conditions, achieve multi-dimensional performance evaluation, have advanced safety warning functions, and be linked with park management systems. SUMMARY

[0005] The purpose of the present application is to provide a solution to the above-mentioned problems.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: A lithium ion battery detection system for power energy storage, comprising: a battery working condition simulation module, configured to generate simulation charge-discharge control instructions including charging, discharging, standing and random power fluctuation according to a preset or externally imported typical load curve of an industrial and commercial park, to drive the battery test equipment to apply highly similar working conditions to the measured battery as real operation.

[0007] Data acquisition and sensing module, for real-time and high-precision acquisition of electrical parameters (such as voltage and current), thermal parameters (such as surface temperature), and environmental parameters (such as ambient temperature) of the battery monomer or the entire battery cluster during the test process.

[0008] Performance evaluation and calculation module, for calculating and evaluating key performance indicators of the battery, including actual capacity, DC / AC internal resistance, energy efficiency, state of health (SOH), and state of power (SOP), based on the collected real-time data through built-in algorithms and models.

[0009] Safety warning and diagnosis module, for early warning and intelligent diagnosis of potential thermal runaway risks, battery cluster inconsistency intensification, and various potential faults through multi-source feature fusion analysis (combining electrical, thermal, and time series data) and model prediction technology.

[0010] Data management and human-computer interaction module, for centralized storage and management of all test data, evaluation results, and warning records, and for providing users with detection scheme configuration, real-time data display, historical result query, and report generation functions through a visual graphical interface.

[0011] Further, the current instruction generation process of the battery working condition simulation module can be specifically designed as follows: first, a standard basic charge-discharge current curve I_base(t) is fitted based on historical operation data. In order to simulate the random fluctuation characteristics of actual load, a random power disturbance term ΔP(t) obeying normal distribution N(0, σ²) is superimposed on the basic curve, where the standard deviation σ can be set according to the load fluctuation characteristics of the target park. Finally, the simulated charge-discharge current instruction I_sim(t) is calculated through the formula I_sim(t) = I_base(t) + ΔP(t) / V_avg, where V_avg is the average voltage of the battery cluster. This method can effectively reproduce the non-stationary and random power scheduling in actual operation.

[0012] Further, in the performance evaluation and calculation module, the state of health SOH_c of the battery can be calculated by the capacity method, and the calculation formula is: SOH_c = (C_actual / C_rated) x 100%, wherein C_actual is the actual complete charge and discharge capacity measured in the current test, and C_rated is the nominal capacity of the battery out of the factory. At the same time, the instantaneous power state SOP(t) of the battery can be estimated by the formula SOP(t) = min(SOC(t) * V(t) / R_min, (1 - SOC(t)) * V(t) / R_min, P_max), which comprehensively considers the current state of charge SOC(t), voltage V(t), allowable minimum direct current resistance R_min and the limitation of battery rated maximum power P_max, to evaluate the maximum power that the battery can safely release or absorb in the current state.

[0013] Further, the performance evaluation and calculation module can also obtain more detailed internal state information by performing electrochemical impedance spectroscopy test. For the measured alternating current impedance spectrum data, a second-order RC equivalent circuit model is used for fitting. The voltage time domain response U(t) of the model can be expressed as: U(t) = OCV(SOC) + I(t) * R_ohm + I(t) * R_ct * (1 -exp(-t / (R_ct * C_dl))) + I(t) * R_diff * (1 - exp(-t / (R_diff * C_diff))). By identifying the model parameters, the ohmic resistance R_ohm, charge transfer resistance R_ct, double-layer capacitance C_dl, diffusion resistance R_diff and diffusion capacitance C_diff can be obtained. The sum of R_ohm and R_ct can be used as the direct current resistance R_dc to more accurately represent the internal state of the battery, which is used to assist the evaluation of SOH and SOP, and the resistance measured by the simple direct current pulse method can better reflect the essence of battery aging.

[0014] Further, the safety warning and diagnosis module can integrate an online parameter identification algorithm based on recursive least squares method. The algorithm can identify the ohmic resistance R_ohm and polarization resistance R_p of the battery in real time. By monitoring the change of the resistance over time, a warning rule is set: when the resistance change rate AR_rate exceeds the preset threshold θ_R in the last N sampling periods, the resistance abnormality warning is triggered. The change rate AR_rate is calculated by the formula AR_rate = |R(k) - R(k-N)| / (R(k-N) * N), wherein R(k) is the resistance value identified at the kth sampling time. This mechanism can capture the gradual change trend of the resistance and realize early degradation warning.

[0015] Furthermore, the safety early warning and diagnosis module can construct a thermal runaway early warning model based on the principle of electro-thermal coupling. This model calculates the total heat generation rate Q_gen inside the battery, with the formula: Q_gen = I² * (R_ohm + R_ct) + |I| * (T * ΔS / nF) + A * exp(-E_a / (R_g * T)) * f(SOC). This formula covers the three main heat sources: Joule heating, reaction heat, and side reaction heat. Simultaneously, by solving the heat balance equation m * C_p *dT / dt = Q_gen - h * A_s * (T - T_amb) in real time, the battery's temperature rise trend in the near future can be predicted. Here, m, C_p, h, A_s, and T_amb represent the battery mass, specific heat capacity, heat dissipation coefficient, surface area, and ambient temperature, respectively. When the battery temperature T_pre predicted by the model exceeds the safety threshold T_safe within a set time window, the system will trigger a high-level thermal runaway risk alarm.

[0016] Furthermore, the safety warning and diagnosis module specifically designs a warning method for battery cluster inconsistency issues. It calculates the standard deviation σ_v and range ΔV_max of the voltages of all individual cells within the cluster and introduces the time derivative to analyze their changing trends. The voltage inconsistency index UI is defined as: UI = α * (σ_v / V_avg) + β * (ΔV_max / V_avg) + γ * |d(σ_v) / dt|, where α, β, and γ are preset weighting coefficients, and V_avg is the average voltage of the individual cells. The warning threshold UI_th(t) is not a fixed value but a dynamic value related to the current average SOH and average SOC of the battery cluster. When the real-time calculated UI exceeds this dynamic threshold, the system determines that the battery cluster has a serious risk of inconsistency deterioration.

[0017] Furthermore, the data acquisition and sensing module must meet the data accuracy requirements under different testing scenarios. When performing electrochemical impedance spectroscopy (EIS) testing, the sampling frequency should be no less than 1 kHz to ensure the capture of high-frequency impedance information; when performing routine charge-discharge monitoring, the sampling frequency should be no less than 10 Hz. The acquired electrical parameters must include at least the total voltage and current of the battery cluster, and the voltage of each individual cell; the thermal parameters must include at least the temperature of at least three key points on the battery surface (e.g., the positive electrode, negative electrode, and center) and the ambient temperature.

[0018] Furthermore, the data management and human-computer interaction module not only stores and manages data but also generates comprehensive professional testing reports. These reports include historical SOH decay curves, internal resistance trend graphs, inconsistency analysis charts, and thermal runaway risk assessment conclusions. Simultaneously, this module supports standard communication protocols (such as Modbus TCP / IP, IEC 61850, etc.), enabling data communication with the park's energy management system. It proactively reports key operational parameters such as the battery system's available capacity and current maximum callable power, obtained through system evaluation, achieving closed-loop linkage between testing and operation.

[0019] The beneficial effects of this invention are: This invention employs a current command generation method based on historical load curves superimposed with random disturbances. This method subjects the tested battery to charging, discharging, resting, and random power fluctuations highly similar to those experienced in a real-world industrial park, significantly enhancing the practical guiding value of performance evaluation results. The system not only assesses basic capacity and internal resistance but also calculates multi-dimensional indicators such as SOH, SOP, and energy efficiency through models and algorithms. Furthermore, it can obtain AC internal resistance spectra through EIS testing for more refined internal state analysis, achieving a comprehensive "check-up" of battery performance. Simultaneously, the system integrates a multi-level early warning mechanism based on gradual internal resistance trends, electro-thermal coupling model prediction, and a dynamic inconsistency index. This mechanism can detect potential safety hazards significantly earlier than traditional voltage / temperature threshold alarms, providing a valuable time window for preventing serious accidents such as thermal runaway. The early warning threshold can be dynamically adjusted according to the battery state (e.g., the inconsistency index threshold is related to SOH and SOC), making the early warning more accurate. The system also utilizes online identification algorithms such as recursive least squares, enabling it to possess self-learning and adaptive capabilities. Through the data interface with the park's energy management system, accurate detection and evaluation results (such as available capacity and maximum callable power) can be directly fed back to the operation decision-making level to optimize charging and discharging strategies and maintenance plans. This truly forms a data-driven closed loop from detection to operation and maintenance, improving the intelligent management level and economy of the entire energy storage power station. Attached Figure Description

[0020] Figure 1 This is a block diagram of the overall structure of the system of the present invention; Figure 2 This is a flowchart of the battery operating condition simulation module generating simulated current in this invention; Figure 3 This is a schematic diagram of the second-order RC equivalent circuit model in this invention; Figure 4 This is a flowchart of the safety early warning and diagnosis module in this invention; Figure 5 This is a schematic diagram illustrating the calculation and early warning of the voltage inconsistency index in this invention. Detailed Implementation

[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 some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0022] like Figure 1 As shown, the present invention provides a lithium-ion battery testing system for power storage, which mainly includes five functional modules: a battery operating condition simulation module, a data acquisition and sensing module, a performance evaluation and calculation module, a safety early warning and diagnosis module, and a data management and human-computer interaction module. Each module communicates at high speed through the system's internal data bus.

[0023] I. Battery Operating Condition Simulation Module This module is the starting point for detection, and its goal is to generate charging and discharging commands that reflect the actual operating status of energy storage batteries in industrial and commercial parks. The load characteristics of industrial and commercial parks include: high power consumption (discharging) during daytime production periods, charging during off-peak hours at night, idle or low-power operation during midday breaks or production adjustments, and random power changes caused by equipment start-ups and shutdowns and fluctuations in renewable energy.

[0024] like Figure 2 As shown, the workflow of this module is as follows: Base curve loading: Select a typical 24-hour net load curve for the industrial park (i.e., the difference between total electricity consumption and photovoltaic / wind power generation in the park) from the database or import it from the user. Normalize the curve to obtain the base power curve P_base(t) expressed as a percentage of nominal power.

[0025] Random Disturbance Superposition: To simulate uncertainties in actual operation, a random power disturbance term ΔP(t) is introduced. This disturbance term is generated independently within each discrete time step Δt, following a normal distribution with a mean of 0 and a standard deviation of σ, i.e., ΔP(t) ~ N(0, σ²). The standard deviation σ can be set based on the historical fluctuation statistical analysis of the park's load, for example, set to 5%-15% of the current value of P_base(t). This simulates the subtle fluctuations that still exist after smoothing control.

[0026] Current command generation: The base power is superimposed with the random disturbance to obtain the simulated power command P_sim(t) = P_base(t) + ΔP(t). Considering that the power P = V * I, and that the battery voltage changes during charging and discharging, to simplify control, the average voltage V_avg of the current battery cluster (which can be fed back from the data acquisition module or approximated using the nominal voltage) is used for conversion to generate the simulated current command I_sim(t): I_sim(t) = P_sim(t) / V_avg, When P_sim(t) > 0, it indicates a discharge command; when P_sim(t) < 0, it indicates a charging command; when P_sim(t) = 0, it indicates a resting state.

[0027] Boundary protection: The generated I_sim(t) also needs to be compared with the battery's rated maximum charge and discharge currents I_max_charge and I_max_discharge for limiting to ensure test safety.

[0028] Through the above process, this module can generate highly realistic test conditions, making subsequent performance evaluations more valuable.

[0029] II. Data Acquisition and Sensing Module This module is responsible for accurately and synchronously capturing the battery's response under simulated operating conditions. Its performance directly affects the accuracy of subsequent evaluation and diagnosis.

[0030] Electrical parameter acquisition: Total voltage and total current: Measurements are performed using high-precision, high-isolation voltage and Hall current sensors. During routine charge / discharge monitoring, the sampling frequency is no less than 10Hz to ensure the capture of dynamic processes. For electrochemical impedance spectroscopy (EIS) testing, the sampling frequency needs to be increased to above 1kHz to meet the requirements of AC signal analysis.

[0031] Individual cell voltage: The voltage of each individual cell in the energy storage battery cluster is acquired through a high-precision, multi-channel battery monitoring unit. The sampling frequency is synchronized with the total voltage, typically 10Hz. This is a key data source for assessing battery inconsistencies.

[0032] Thermal parameter acquisition: At least three thermocouples or digital temperature sensors are placed at key locations in the battery module (such as near the positive and negative terminals, or on the center surface of the cell) to monitor the battery surface temperature T_s1, T_s2, T_s3.

[0033] An ambient temperature sensor is placed in the heat dissipation channel or cabinet of the battery cluster to monitor the ambient temperature T_amb.

[0034] Synchronization and Calibration: All sensor data are synchronized using a unified time scale, and zero-point and range calibrations are performed periodically to eliminate system errors.

[0035] III. Performance Evaluation and Calculation Module This module is the system's "evaluation center," which processes the collected raw data and calculates key indicators reflecting battery performance.

[0036] Basic performance parameter calculation: Actual capacity C_actual: The integral of the discharge current I(t) over a complete cycle from full charge to discharge cutoff: C_actual = ∫|I(t)| dt (when I(t) < 0 during discharge). The integration interval is from SOC = 100% to SOC = 0% (or the manufacturer-defined cutoff condition).

[0037] State of Health (SOH_c): Based on the definition of capacity decay, SOH_c = (C_actual / C_rated) × 100%. Where C_rated is the nominal capacity of the battery at the time of manufacture.

[0038] Energy efficiency η: The ratio of discharge energy E_discharge to charge energy E_charge in one complete charge-discharge cycle: η = E_discharge / E_charge × 100%. Where E = ∫V(t)*I(t) dt.

[0039] Electrochemical impedance spectroscopy and internal resistance analysis: Periodically (e.g., weekly or monthly) or in case of performance anomalies, an EIS test subprocess is triggered. The system controls the charging and discharging equipment to inject a small-amplitude sinusoidal alternating current excitation signal I_ac(t) = I_0 * sin(2πft) containing multiple frequencies (e.g., 0.1Hz ~ 1000Hz) into the battery, while simultaneously acquiring the battery's voltage response U_ac(t).

[0040] By analyzing the amplitude ratio and phase difference of voltage and current at each frequency f, the impedance spectrum of the battery Z(f) = R(f) + jX(f) is obtained.

[0041] For quantitative analysis, the following methods are used: Figure 3 The second-order RC equivalent circuit model shown is fitted. This model includes the ohmic internal resistance R_ohm, the parallel R_ct / / C_dl circuit corresponding to the charge transfer process, and the parallel R_diff / / C_diff circuit corresponding to the lithium-ion diffusion process. The open-circuit voltage OCV is a function of SOC.

[0042] In the time domain, the voltage response U(t) of this model to a step current I can be expressed as: U(t) = OCV(SOC) + I * R_ohm + I * R_ct * [1 - exp(-t / (R_ct * C_dl))] + I * R_diff * [1 - exp(-t / (R_diff * C_diff))] By fitting EIS data using the nonlinear least squares method, parameters such as R_ohm, R_ct, C_dl, R_diff, and C_diff can be accurately identified. R_ohm primarily reflects the electrolyte and contact resistance, while R_ct reflects the resistance to electrochemical reaction kinetics. The sum of these two, R_dc = R_ohm + R_ct, can be approximated as the DC internal resistance and is an important parameter for evaluating battery SOH and calculating SOP. A significant increase in R_ohm is usually associated with electrolyte drying and interface deterioration; a significant increase in R_ct is associated with active material deactivation and excessive SEI film growth.

[0043] Power State of Operation (SOP) Estimation: SOP refers to the maximum power that a battery can continuously provide or receive within the next short time window (e.g., 10 seconds, 30 seconds) at the current moment (SOC, temperature, health status). This is crucial for real-time power scheduling in the industrial park.

[0044] This invention employs an estimation method based on internal resistance and voltage boundaries. The instantaneous SOP is limited by three factors: maximum charging current, maximum discharging current, and voltage safety window.

[0045] Power under current limitation: Limited by the maximum allowable current of the battery I_max, the power is P_I = V(t) * I_max.

[0046] Power under voltage limitations: Considering the internal resistance voltage drop, during a discharge with current I over a time interval Δt, the voltage will drop to V(t) - I * R_min, which cannot be lower than the minimum allowable voltage V_min. Therefore, the maximum discharge current I_dis_max_v = (V(t) - V_min) / R_min, corresponding to the discharge power P_dis_v = V(t) * I_dis_max_v. Similarly, the maximum charging current I_chg_max_v = (V_max - V(t)) / R_min, corresponding to the charging power P_chg_v = V(t) * I_chg_max_v. Here, R_min is a conservative internal resistance value, which can be taken as the currently identified R_dc value.

[0047] State of Charge (SOC) Limitation: Charge acceptance decreases at high SOC, and discharge capability decreases at low SOC. A SOC-related coefficient can be introduced to correct this.

[0048] Taking all the above factors into account, the SOP at time t (using discharge SOP as an example) can be estimated as follows: SOP_discharge(t) = min( P_I, P_dis_v, SOC(t) * P_rated ) Where P_rated is the rated power. The SOC(t) coefficient linearly decreases the maximum power at low SOC. The charging SOP is calculated similarly, using (1-SOC(t)) as the coefficient.

[0049] IV. Safety Early Warning and Diagnosis Module like Figure 4 As shown, this module is the system's "safety brain," performing real-time or near-real-time data analysis to issue early warnings before a failure occurs.

[0050] Online internal resistance identification and trend warning based on recursive least squares method: During the static or low-current constant-current phases of the charging and discharging process, online parameter identification can be performed using a first-order RC model (simplified model). The battery terminal voltage U(k) can be expressed in the discrete-time domain as: U(k) = OCV(k) + I(k)*R_ohm + U_p(k) + v(k) Where U_p(k) is the polarization voltage and v(k) is the measurement noise. The dynamics of the polarization voltage can be approximated as: U_p(k) = exp(-Δt / τ)*U_p(k-1) + R_p*(1-exp(-Δt / τ))*I(k-1), where τ = R_p*C_p.

[0051] Treating OCV as a quantity that changes slowly with SOC, it can be considered approximately constant within a short time window. The above equation is rearranged into linear parametric form: y(k) = φ(k)^T * θ(k) + v(k), where y(k)=U(k)-OCV_est, φ(k)=[I(k), U_p(k-1), I(k-1)], θ(k)=[R_ohm, exp(-Δt / τ), R_p*(1-exp(-Δt / τ))].

[0052] The parameter estimate θ_hat(k) is updated online using a recursive least squares method with a forgetting factor λ (0.95 < λ < 1). R_ohm(k) and R_p(k) can be calculated from θ_hat(k).

[0053] Warning logic: Calculate the average rate of change of internal resistance ΔR_rate over N consecutive sampling periods. Set a threshold θ_R (e.g., daily change rate of R_ohm exceeding 0.5%, daily change rate of R_p exceeding 1%). When ΔR_rate exceeds θ_R M times consecutively, a "accelerated aging of internal resistance" warning is triggered. This often indicates serious side reactions occurring inside the battery, such as explosive growth of the SEI film.

[0054] Thermal runaway prediction based on a thermo-electric coupling model: Thermal runaway is the most serious safety hazard in lithium-ion batteries. This invention uses a simplified thermo-electric coupling model to predict battery temperature rise in real time.

[0055] Heat generation rate model Q_gen: The total heat generation of the battery mainly includes joules Q_joule, reaction heat Q_reaction, and side reaction heat Q_side.

[0056] Q_gen = Q_joule + Q_reaction + Q_side Joule heating: Q_joule = I² * (R_ohm + R_ct). The internal resistance parameter is obtained from EIS or online identification.

[0057] Heat of reaction (reversible heat): Q_reaction = |I| * (T * ΔS / nF). Where ΔS is the entropy change coefficient of the battery reaction, a function of SOC and temperature, which can be obtained through experimental determination or by consulting literature.

[0058] Side reaction heat: mainly simulates SEI film decomposition, negative electrode and electrolyte reaction, etc. It is described using the Arrhenius equation: Q_side = A * exp(-E_a / (R_g * T)) * f(SOC). A is the pre-exponential factor, E_a is the activation energy, and f(SOC) describes the relationship between the side reaction rate and SOC (usually faster at higher SOCs).

[0059] Heat dissipation model: Assume that the battery dissipates heat through convection, and the heat dissipation rate Q_loss = h * A_s * (T - T_amb), where h is the convective heat transfer coefficient and A_s is the equivalent heat dissipation surface area of ​​the battery.

[0060] Heat balance equation and temperature prediction: m * C_p * dT / dt = Q_gen - Q_loss Where m is the battery mass and C_p is the specific heat capacity. This is a first-order differential equation. Given the current time T(k), I(k), SOC(k) and parameters, numerical methods such as the Euler method can be used to predict the temperature T_pre(k+1), T_pre(k+2), ... within a future time period Δt_pre.

[0061] Warning logic: If the predicted temperature T_pre exceeds the thermal runaway initiation temperature T_TR (e.g., 150°C) at any point within a future t_warn time (e.g., 300 seconds), the highest-level "thermal runaway risk" alarm is immediately triggered, and emergency measures such as isolation and forced cooling are recommended. This model can predict dangerous temperature rise trends in advance, even before the actual temperature sensor readings reach the dangerous value.

[0062] Comprehensive diagnosis of inconsistencies: Inconsistency in battery clusters is a key factor leading to capacity loss, accelerated aging, and even localized overcharging and over-discharging. This invention not only monitors voltage differences but also defines a comprehensive inconsistency index, UI, for quantitative evaluation.

[0063] like Figure 5 As shown, within a battery cluster, the average voltage V_avg, standard deviation σ_v, and range ΔV_max = max(V_i) - min(V_i) of all individual cells are calculated in real time. While the range may be affected by individual outliers, combining it with the standard deviation reflects the overall degree of dispersion. More importantly, the deterioration of inconsistency is a process; therefore, the time derivative d(σ_v) / dt is introduced to capture its changing trend.

[0064] Define the voltage inconsistency index U as: UI(t) = α * (σ_v(t) / V_avg(t)) + β * (ΔV_max(t) / V_avg(t)) + γ *|d(σ_v) / dt| Where α, β, and γ are weighting coefficients set based on experience, for example, α=0.4, β=0.4, γ=0.2. The absolute value of |d(σ_v) / dt| represents the degree of drastic change of interest, whether it increases or suddenly decreases (which may be accompanied by equilibration actions or malfunctions).

[0065] Dynamic threshold UI_th(t): The tolerance for inconsistency is related to the battery state. For new batteries or batteries with high SOH, UI_th can be set lower to ensure greater consistency. For aged batteries, UI_th can be appropriately relaxed. Furthermore, in the high and low SOC ranges, the voltage is more sensitive to changes in SOC, and UI_th should also be dynamically adjusted. It can be set as UI_th(t) = UI_base + k1*(1-SOH) + k2*|SOC-50%|, where UI_base is the base threshold, and k1 and k2 are adjustment coefficients.

[0066] When UI(t) > UI_th(t) and this condition persists for a period of time, the system triggers an "Inconsistency Deterioration" warning, prompting operations and maintenance personnel to perform load balancing maintenance or check for problematic individual units.

[0067] V. Data Management and Human-Computer Interaction Module This module serves as the interface between the system and users, as well as external systems.

[0068] Data Management: A time-series database is used to store all raw sampling data, intermediate calculation results, and early warning events. The data storage period is configurable to meet the needs of long-term traceability analysis.

[0069] Human-computer interaction: Provides a graphical software interface. Users can: Configure the testing scheme (select load curve, set test duration, trigger EIS test, etc.).

[0070] Monitor the testing process in real time, view the current, voltage, and temperature curves, as well as the real-time display of key indicators such as SOC, SOH, and SOP.

[0071] View the list of security alerts and historical events.

[0072] Generate a comprehensive test report with one click. The report includes: capacity test results and SOH curve, EIS test spectrum and fitting parameter table, energy efficiency analysis, historical SOP statistics, internal resistance change trend graph, inconsistency evolution graph, all warnings triggered in this test and analysis conclusions, summary of the current status of the battery system and maintenance recommendations.

[0073] System Communication: Supports standard communication protocols (such as Modbus TCP, MQTT, IEC 61850, etc.), and can report key status information of the battery system, such as available capacity = C_actual, current maximum available charging / discharging power = SOP(t), overall SOH, fault / early warning status, etc., to the energy management system of industrial and commercial parks in real time, providing direct data support for optimized scheduling.

[0074] Example 2: A testing method using a lithium-ion battery testing system for power storage Based on the above system, the detection method provided by the present invention includes the following steps: S1: Configure testing parameters. Users can select a historical typical daily load curve of the target energy storage power station's industrial park, or define a custom test curve, through the user interface. They can also set the total test duration (e.g., 72-hour cyclic test), safety threshold parameters, report generation options, etc.

[0075] S2: Start the operating condition simulation and test. Based on the configuration in S1, the system generates the I_sim(t) command from the battery operating condition simulation module to control the connected charging and discharging equipment to apply excitation to the battery cluster under test. At the same time, the data acquisition and sensing module is powered on and begins high-frequency data acquisition.

[0076] S3: Synchronous Data Acquisition. Throughout the test, the data acquisition module operates continuously, sending voltage, current, and temperature data with precise timestamps to the data processing bus.

[0077] S4: Performance evaluation calculation. The performance evaluation and calculation module receives data streams in real time.

[0078] SOC estimation is performed (usually using ampere-hour integration combined with OCV correction method); At the end of a complete charge-discharge cycle, C_actual and SOH_c are automatically calculated; Execute the EIS test sub-process periodically (or manually) to perform impedance spectrum analysis and parameter identification; Based on the current SOC, voltage, temperature, and identified internal resistance, the SOP value is calculated and updated in real time.

[0079] S5: Security Early Warning and Diagnosis. The security early warning and diagnosis module runs in parallel with S4.

[0080] Using the recursive least squares method, the internal resistance is continuously identified online in the background, and its changing trend is monitored. Real-time acquisition of current, voltage, and temperature data; running a thermo-electric coupling model to predict short-term temperature rise. The voltage inconsistency index (UI) of the battery cluster is calculated once per second. Based on the built-in algorithm logic, it determines whether an internal resistance change warning, thermal runaway risk warning, or inconsistency deterioration warning has been triggered. Once triggered, an alarm immediately pops up on the human-machine interface and the event is recorded. S6: Data management, report generation, and information reporting.

[0081] Throughout the entire testing process, all data was stored in a time-series database; After the test is completed (or during the test), the user can instruct the system to generate a predefined comprehensive test report; Throughout the entire test and after its completion, the system proactively reports the key state parameters (SOH, SOP, health status) of the battery system to the park's energy management system via the communication interface at set intervals.

[0082] Through the above steps, this invention completes a full testing process from simulated operating conditions and comprehensive evaluation to advanced diagnosis, realizing a deep, accurate, and forward-looking "physical examination" of lithium-ion battery systems for power storage.

Claims

1. A lithium ion battery detection system for electric energy storage, characterized by: Comprising a battery working condition simulation module for generating simulated charge and discharge control instructions including charging, discharging, standing, and power fluctuation according to preset or imported industrial park load curves; a data acquisition and sensing module for real-time acquisition of electrical, thermal, and environmental parameters of battery monomers or battery clusters; a performance evaluation and calculation module for calculating battery capacity, internal resistance, energy efficiency, health state, and power state parameters based on the acquired data; a safety warning and diagnosis module for warning and diagnosing battery thermal runaway, inconsistency, and potential faults through multi-feature fusion analysis and model prediction; a data management and human-computer interaction module for storing and managing detection data and providing a visual interface for configuring detection schemes, displaying evaluation results, and providing warning information.

2. The lithium-ion battery detection system for electrical energy storage of claim 1, wherein: The battery working condition simulation module generates a standard charge and discharge curve according to historical operation data and superimposes a random power disturbance term ΔP(t), which follows a normal distribution N(0, σ²), where σ is a standard deviation set according to the load fluctuation characteristics of the park. The calculation formula of the simulated charge and discharge current instruction I_sim(t) is: I_sim(t) = I_base(t) + ΔP(t) / V_avg, where I_base(t) is the basic current determined according to the standard charge and discharge curve, and V_avg is the average voltage of the battery cluster.

3. The lithium ion battery detection system for power energy storage according to claim 1, characterized in that: In the performance evaluation and calculation module, the calculation formula of the battery health state SOH_c is: SOH_c = (C_actual / C_rated) × 100%, where C_actual is the actual measured complete charge and discharge capacity, and C_rated is the nominal capacity of the battery; at the same time, the calculation formula of the battery power state SOP(t) is: SOP(t) = min(SOC(t) * V(t) / R_min, (1 - SOC(t)) * V(t) / R_min, P_max), where SOC(t) is the state of charge at the current time, V(t) is the current voltage, R_min is the allowable minimum direct current internal resistance, and P_max is the rated maximum power of the battery.

4. The system for detecting lithium-ion battery for energy storage of claim 3, wherein: The performance evaluation and calculation module obtains the alternating current internal resistance spectrum of the battery through electrochemical impedance spectroscopy testing and adopts an equivalent circuit model fitting, wherein the voltage response U(t) expression of the second-order RC equivalent circuit model is: U(t) = OCV(SOC) + I(t) * R_ohm + I(t) * R_ct * (1 - exp(-t / (R_ct * C_dl))) + I(t) * R_diff * (1 - exp(-t / (R_diff * C_diff))), Wherein, OCV(SOC) is the open circuit voltage, which is a function of SOC; R_ohm is the ohmic resistance; R_ct and C_dl are the charge transfer resistance and double-layer capacitance, respectively; R_diff and C_diff are the diffusion resistance and diffusion capacitance, respectively; the sum of R_ohm and R_ct obtained by parameter identification is taken as the direct current resistance R_dc, which is used for auxiliary evaluation of SOH and SOP.

5. The lithium-ion battery detection system for electrical energy storage of claim 1, wherein: The safety warning and diagnosis module uses the recursive least squares method to identify the ohmic resistance R_ohm and polarization resistance R_p of the battery online, and triggers a warning when the resistance change rate AR_rate exceeds the threshold value θ_R in the continuous N sampling periods, and the calculation formula of AR_rate is: AR_rate = |R(k) - R(k-N)| / (R(k-N) * N), Wherein, R(k) is the resistance (R_ohm or R_p) identified at the kth sampling time.

6. The system for detecting lithium-ion battery for energy storage of claim 5, wherein: The safety warning and diagnosis module constructs a thermal runaway warning model based on a thermal-electric coupling model, and the calculation formula of the battery heat generation rate Q_gen is: Q_gen = I² * (R_ohm + R_ct) + |I| * (T * ΔS / nF) + A * exp(-E_a / (R_g* T)) * f(SOC), Wherein, I is the current; T is the battery temperature; ΔS is the reaction entropy; n is the number of electron transfer; F is the Faraday constant; A is the pre-exponential factor; E_a is the activation energy; R_g is the ideal gas constant; f(SOC) is a function related to SOC; the real-time solution of the heat balance equation m * C_p * dT / dt = Q_gen - h * A_s * (T - T_amb) predicts the battery temperature rising trend, wherein m and C_p are the battery mass and specific heat capacity, h and A_s are the heat dissipation coefficient and surface area, and T_amb is the ambient temperature; when the predicted temperature T_pre exceeds the safety threshold T_safe within a set time, an advanced warning is triggered.

7. The lithium-ion battery detection system for energy storage applications of claim 1, wherein: The safety warning and diagnosis module calculates the standard deviation σ_v and the range ΔV_max of the voltage of each single battery in the battery cluster, and analyzes the change trend in the time dimension, and defines the voltage inconsistency index UI as: UI = α * (σ_v / V_avg) + β * (ΔV_max / V_avg) + γ * |d(σ_v) / dt| Wherein, α, β, and γ are weighting coefficients, and V_avg is the average value of the single battery voltage; when UI exceeds the dynamic threshold UI_th(t), it is determined that the battery cluster has a serious inconsistency risk, and the dynamic threshold UI_th(t) is related to the SOH and average SOC of the battery.

8. The lithium-ion battery detection system for electrical energy storage of any one of claims 1-7, wherein: The sampling frequency of the data acquisition and sensing module is not less than 1 kHz during electrochemical impedance spectroscopy testing, and not less than 10 Hz during regular charge and discharge monitoring; the collected electrical parameters at least include total voltage, total current, and single battery voltage, and the thermal parameters at least include the surface temperatures of at least three key points and the ambient temperature.

9. The lithium-ion battery detection system for energy storage applications of claim 1, wherein: The data management and human-computer interaction module can generate a comprehensive detection report containing SOH attenuation curve, internal resistance change trend, inconsistency analysis atlas and thermal runaway risk assessment report, and support data communication with the park energy management system to report the available capacity and maximum callable power of the battery system.

Images