An outdoor energy storage battery state monitoring fault diagnosis method and system

Abstract

The application discloses an outdoor energy storage battery state monitoring fault diagnosis method and system, which comprises the following steps: three micro electrochemical parameters of an ideal healthy battery are extracted in advance to construct a three-dimensional electrochemical manifold space and generate an intrinsic lithium intercalation and extraction steady state domain; end voltage, charging and discharging current and surface temperature are collected in real time to reverse real-time micro parameters and generate concentration polarization state points and evolution trajectories; a temperature correction factor and a Nernst equation are used to dynamically calculate a lithium precipitation critical surface and construct a dynamic lithium precipitation safety domain; the position combination relationship of the state points relative to the double domains is calculated, the battery micro physical state is accurately divided into a uniform intercalation and extraction zone, a concentration polarization accumulation zone, a local lithium precipitation boundary zone and a crystal lattice stress rupture zone, and the corresponding maintenance strategy, duty cycle pulse relaxation, dynamic de-rating current limiting or deep discharge and fault logic isolation instructions are executed. The application realizes high-precision closed-loop diagnosis of the battery based on micro manifold topology.

Description

Technical Field

[0001] This invention relates to the field of energy storage battery condition monitoring and fault pre-diagnosis technology, and in particular to an outdoor energy storage battery condition monitoring and fault diagnosis method and system. Background Technology

[0002] With the large-scale deployment of outdoor energy storage systems, monitoring the condition of batteries and diagnosing faults in complex climatic environments has become a core challenge in ensuring the safe operation of the entire energy storage power station.

[0003] Currently, traditional battery state monitoring methods mainly rely on simple threshold determinations based on macroscopic external parameters such as terminal voltage and surface temperature. The fundamental flaw of this approach lies in the severe hysteresis of macroscopic parameters, which fails to reveal the complex microscopic electrochemical evolution processes within the battery, particularly the difficulty in accurately quantifying internal concentration polarization. This results in existing systems being extremely insensitive to early microscopic faults, posing a significant risk of undetected and hidden failures.

[0004] Secondly, in the face of drastically fluctuating outdoor temperatures, existing technologies generally employ rigid static safety alarm thresholds. In fact, battery polarization deteriorates rapidly in low-temperature environments, easily leading to localized lithium plating. Fixed static boundaries cannot adapt to this electrochemical and physical law, not only easily causing excessive idleness and waste of battery charging and discharging performance at room temperature, but also potentially causing serious safety accidents due to the failure of defense mechanisms under harsh low-temperature conditions.

[0005] Furthermore, existing battery management systems lack refined standards for classifying microstates and tiered active intervention methods. When polarization accumulates or early lithium plating tends to occur inside the battery, traditional systems can only passively trigger shutdown when the state deteriorates to a dangerous limit. They cannot take flexible closed-loop control measures such as pulse relaxation or dynamic derating before irreversible lattice structure destruction occurs, making it difficult to actively guide the dangerous state trajectory back to the safe region, thus failing to cut off the incipient path of thermal runaway at its root.

[0006] Therefore, there is an urgent need in this field for a novel battery state diagnosis method that can quantify polarization phenomena with high precision at the microscopic level, possess a dynamic boundary contraction mechanism that adapts to varying temperature environments, and implement a step-by-step closed-loop intervention strategy to completely solve the technical bottlenecks of early hidden danger underreporting and passive defense. Summary of the Invention

[0007] To address the shortcomings of existing technologies, a fault diagnosis method for outdoor energy storage battery state monitoring includes: The baseline state data of an ideal healthy battery under standard cycle conditions is obtained in advance. Three microscopic electrochemical parameters are extracted from the data: lithium ion concentration on the negative electrode solid surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface. A three-dimensional electrochemical manifold space is constructed using the three microscopic electrochemical parameters as the coordinate basis. The continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time are projected into the three-dimensional electrochemical manifold space to generate the reference trajectory envelope volume composed of continuous state points, and this volume is defined as the intrinsic lithium insertion / extraction steady-state domain. The current battery's terminal voltage, charging and discharging current, and surface temperature are collected in real time. The three microscopic electrochemical parameters are inverted in real time through a preset nonlinear state observer and projected onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point. Then, the state space evolution trajectory is updated by combining the current battery's historical concentration polarization state points. Based on the temperature correction factor obtained from the surface temperature mapping and the Nernst equation, the current lithium plating critical surface is calculated in the three-dimensional electrochemical manifold space. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain. Calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / deintercalation steady-state domain and the dynamic lithium plating safety domain, and divide the battery microphysical state in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region. When the location is determined to be in the uniform insertion / deintercalation region, the current charge / discharge strategy is maintained; when the location is determined to be in the concentration polarization accumulation region, a duty cycle pulse relaxation instruction is generated and executed; when the location is determined to be in the local lithium plating boundary region, the overpotential overshoot across the lithium plating critical surface is calculated, and a dynamic derating current limiting instruction is generated and executed based on the overpotential overshoot; when the location is determined to be in the lattice stress fracture region, a deep discharge and fault logic isolation instruction is generated and executed.

[0008] According to a preferred embodiment, the three microscopic electrochemical parameters are retrieved in real time using a preset nonlinear state observer, including: The terminal voltage, charging and discharging current, and surface temperature are substituted as excitation vectors into a pre-set single-particle reduced-order electrochemical model with electrolyte diffusion dynamics. The partial differential equations are then solved to obtain the theoretical internal state prior estimation matrix and the theoretical terminal voltage. Calculate the dynamic residual between the theoretical terminal voltage and the terminal voltage, and use the extended Kalman filter algorithm to calculate the observer gain matrix at the current time step; Based on the observer gain matrix, the theoretical internal state prior estimation matrix is ​​corrected in a closed loop, and the posterior optimal state estimation matrix is ​​output. The concentration values ​​of the outermost spatial nodes of the negative electrode are extracted from the posterior optimal state estimation matrix as the lithium ion concentration on the solid surface of the negative electrode. The concentration difference rate between the two ends of the porous electrode is extracted as the lithium ion concentration gradient in the liquid phase. Based on the interphase potential difference and the stripped ohmic voltage drop resolved from the posterior optimal state estimation matrix, the overpotential of the solid-liquid interface is calculated.

[0009] According to a preferred embodiment, constructing a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as a coordinate basis, and projecting it onto the three-dimensional electrochemical manifold space includes: The obtained lithium ion concentration on the negative electrode solid surface, the lithium ion concentration gradient in the liquid phase, and the overpotential at the solid-liquid interface were subjected to extreme value normalization. The three normalized microscopic electrochemical parameters are configured as three mutually orthogonal linearly independent basis vectors in the topological space, thereby constructing an absolute coordinate system for the three-dimensional electrochemical manifold space; The current instantaneous scalar values ​​of the three normalized microscopic electrochemical parameters are mapped to three-dimensional spatial coordinates in the absolute coordinate system to locate the unique real-time concentration polarization state point.

[0010] According to a preferred embodiment, the three-dimensional solid space enclosed by the lithium plating critical surface constitutes a dynamic lithium plating safety domain, and its dynamic boundary contraction mechanism is as follows: The electrolyte ionic conductivity constant and solid-phase diffusion coefficient in the single-particle reduced-order electrochemical model are corrected online and in real time by the temperature correction factor; the temperature correction factor characterizes the current temperature decay. As the temperature characterized by the temperature correction factor continues to decrease, the corrected electrolyte ionic conductivity constant and the solid-phase diffusion coefficient exhibit nonlinear decay, and the absolute value of the lithium plating initiation overpotential calculated after substituting into the Nernst equation decreases dynamically accordingly. The critical surface for lithium plating undergoes a three-dimensional geometric centripetal contraction toward the origin of the coordinate system in the three-dimensional electrochemical manifold space based on the reduced absolute value of the lithium plating initiation overpotential, resulting in a strictly monotonically decreasing envelope volume of the dynamic lithium plating safety domain.

[0011] According to a preferred embodiment, the real-time division of the battery's microphysical state into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region includes: If the real-time concentration polarization state point is located within both the intrinsic lithium intercalation / deintercalation steady-state region and the dynamic lithium plating safety region, it is determined to be located in the uniform intercalation / deintercalation region. If the real-time concentration polarization state point is located inside the dynamic lithium plating safety domain and simultaneously outside the intrinsic lithium intercalation / deintercalation steady-state domain, then it is determined to be located in the concentration polarization accumulation region. If the real-time concentration polarization state point is located inside the intrinsic lithium insertion / extraction steady-state domain but outside the dynamic lithium plating safety domain, it is determined to be located in the local lithium plating boundary region. If the real-time concentration polarization state point is located outside both the intrinsic lithium insertion / extraction steady-state domain and the dynamic lithium plating safety domain, it is determined to be located in the lattice stress fracture region.

[0012] According to a preferred embodiment, the generation and execution of the duty cycle pulse relaxation instruction includes: Calculate the shortest Euclidean distance from the real-time concentration polarization state point to the outer boundary of the intrinsic lithium insertion / deintercalation steady-state domain, and solve for the derivative vector of the state space evolution trajectory to obtain the polarization divergence direction. The shortest Euclidean distance and the polarization divergence direction are input into a preset fuzzy logic reasoning system that establishes a mapping relationship between distance, direction and zero current intermittent time constant, and the zero current intermittent time constant required to force the internal concentration polarization to subside is calculated. Based on the zero-current intermittent time constant, the duty cycle of the pulse width modulation signal of the underlying battery management main control chip is adjusted, and a relaxation period is forcibly inserted into the continuous charge and discharge command to cut off the charge transfer reaction.

[0013] According to a preferred embodiment, calculating the overpotential overshoot across the lithium plating critical surface includes: Within the three-dimensional electrochemical manifold space, the state space evolution trajectory is extracted and updated in real time. Solve for the geometric intersection of the extension direction of the tangent of the state space evolution trajectory and the critical surface of lithium plating at the current moment in space, and define the interface overpotential at this intersection as the safety critical clamping threshold. The absolute difference between the solid-liquid interface overpotential and the safety critical clamping threshold at the current instant of the real-time concentration polarization state point is calculated as the overpotential overshoot, which characterizes the degree of thermodynamic overshoot.

[0014] According to a preferred embodiment, generating and executing a dynamic rate throttling instruction includes: The overpotential overshoot is used as an error feedback signal input into the closed-loop adaptive proportional-integral-derivative controller to calculate the charge and discharge power compensation term that needs to be reduced. Based on the charging and discharging power compensation term, the physical threshold of the maximum allowable charging current of the current battery is reduced proportionally. The updated maximum allowable charging current physical threshold is sent to the main controller of the battery pack to cut off the power, forcing the evolution trajectory of the real-time concentration polarization state point in the next time step to be restricted back to the dynamic lithium plating safety domain, thereby achieving closed-loop safety traction.

[0015] According to a preferred embodiment, generating and executing the deep discharge and fault logic isolation instructions includes: A forced deep discharge request is generated to control the continuous discharge of faulty battery cells that have been determined to have lattice stress cracking, until the terminal voltage of the faulty battery cell drops below the preset thermodynamic absolute safe potential lower limit. The allowable charge / discharge power threshold of the faulty battery cell is forcibly set to zero, and subsequent charge / discharge control commands for the faulty battery cell are blocked to achieve fault isolation at the pure logic level.

[0016] This invention provides an outdoor energy storage battery state monitoring and fault diagnosis system, comprising: The manifold space construction module is used to pre-acquire the baseline state data of an ideal healthy battery under standard cycle conditions, extract three microscopic electrochemical parameters from them: lithium ion concentration on the negative electrode solid phase surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface, and construct a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as the coordinate basis. The steady-state domain definition module is used to project the continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time into the three-dimensional electrochemical manifold space, generate the reference trajectory envelope volume composed of continuous state points, and define it as the intrinsic lithium insertion / extraction steady-state domain. The state inversion and trajectory generation module is used to collect the current battery's terminal voltage, charging and discharging current and surface temperature in real time. It uses a preset nonlinear state observer to invert the three microscopic electrochemical parameters in real time and projects them onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point. Then, it combines the current battery's historical concentration polarization state points to update and generate the state space evolution trajectory. The dynamic safety domain calculation module is used to calculate the current lithium plating critical surface in the three-dimensional electrochemical manifold space based on the temperature correction factor obtained by the surface temperature mapping and the Nernst equation. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain. The state space mapping partitioning module is used to calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / deintercalation steady-state domain and the dynamic lithium plating safety domain, and to divide the battery micro-physical state in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region. The fault diagnosis and safety control execution module is used to maintain the current charge / discharge strategy when it is determined to be in the uniform insertion / deintercalation region; generate and execute a duty cycle pulse relaxation instruction when it is determined to be in the concentration polarization accumulation region; calculate the overpotential overshoot across the lithium plating critical surface when it is determined to be in the local lithium plating boundary region, and generate and execute a dynamic derating current limiting instruction based on the overpotential overshoot; and generate and execute a deep discharge and fault logic isolation instruction when it is determined to be in the lattice stress fracture region.

[0017] The present invention has the following beneficial effects: 1. This invention breaks through the limitations of traditional outdoor energy storage systems that rely on macroscopic parameters for early warning. It maps the inverted microscopic electrochemical parameters to a three-dimensional manifold topological space, quantifying the concentration polarization phenomenon inside the battery with high precision. This significantly improves monitoring sensitivity, provides microscopic evolutionary evidence for fault diagnosis, and fundamentally eliminates early-stage hidden dangers.

[0018] 2. This invention innovatively introduces a dynamic boundary contraction mechanism that decays with temperature. For outdoor variable-temperature environments, a temperature correction factor is used to drive the centripetal contraction of the lithium plating critical surface, aligning with the law of intensified polarization at low temperatures. This mechanism precisely locks the safety boundary at each temperature, avoiding performance limitations at room temperature while ensuring safety in extremely cold conditions.

[0019] 3. This invention constructs a refined micro-state classification criterion, matched with a tiered active safety strategy. When polarization accumulation or lithium plating tendencies occur, intervention is carried out by generating pulse relaxation commands or lowering the charging limit through a closed loop. This control closed loop can forcibly pull the state trajectory back to the safe region before lattice fracture, completely nipping thermal runaway in the bud. Attached Figure Description

[0020] Figure 1 A structural block diagram of an outdoor energy storage battery state monitoring and fault diagnosis system is provided as an exemplary embodiment. Figure 2 A flowchart of an outdoor energy storage battery state monitoring and fault diagnosis method provided as an exemplary embodiment. Detailed Implementation

[0021] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0022] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular forms “a,” “the,” and “the” used in this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0023] It should be understood that although the terms first, second, third, etc., may be used in this invention to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first information may also be referred to as second information without departing from the scope of this invention, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0024] See Figure 1 The outdoor energy storage battery status monitoring and fault diagnosis system of the present invention may include: The manifold space construction module is used to pre-acquire the baseline state data of an ideal healthy battery under standard cycle conditions, extract three microscopic electrochemical parameters from them: lithium ion concentration on the negative electrode solid phase surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface, and construct a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as the coordinate basis. The steady-state domain definition module is used to project the continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time into the three-dimensional electrochemical manifold space, generate the reference trajectory envelope volume composed of continuous state points, and define it as the intrinsic lithium insertion / extraction steady-state domain. The state inversion and trajectory generation module is used to collect the current battery's terminal voltage, charging and discharging current and surface temperature in real time. It uses a preset nonlinear state observer to invert the three microscopic electrochemical parameters in real time and projects them onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point. Then, it combines the current battery's historical concentration polarization state points to update and generate the state space evolution trajectory. The dynamic safety domain calculation module is used to calculate the current lithium plating critical surface in the three-dimensional electrochemical manifold space based on the temperature correction factor obtained by the surface temperature mapping and the Nernst equation. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain. The state space mapping partitioning module is used to calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / deintercalation steady-state domain and the dynamic lithium plating safety domain, and to divide the battery micro-physical state in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region. The fault diagnosis and safety control execution module is used to maintain the current charge / discharge strategy when it is determined to be in the uniform insertion / deintercalation region; generate and execute a duty cycle pulse relaxation instruction when it is determined to be in the concentration polarization accumulation region; calculate the overpotential overshoot across the lithium plating critical surface when it is determined to be in the local lithium plating boundary region, and generate and execute a dynamic derating current limiting instruction based on the overpotential overshoot; and generate and execute a deep discharge and fault logic isolation instruction when it is determined to be in the lattice stress fracture region.

[0025] See Figure 2 The method for fault diagnosis of outdoor energy storage battery status monitoring according to the present invention may include: S1. Pre-acquire the baseline state data of an ideal healthy battery under standard cycle conditions, extract three microscopic electrochemical parameters from it: lithium ion concentration on the negative electrode solid phase surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface, and construct a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as the coordinate basis.

[0026] "Baseline state data" is strictly isolated from the real-time data of the battery under test. Specifically, it refers to historical test datasets recorded during the manufacturing process or in a laboratory environment for cells from the same batch with a state of health (SOH) ≥ 98%, under standard cycle conditions (such as CC-CV conditions: charging at 1C constant current to the cutoff voltage and then charging at constant voltage to 0.05C in a 25°C constant temperature chamber). This data serves as an absolute health reference, free from any interference from degradation or faults.

[0027] Those skilled in the art will understand that the "extraction from baseline state data" described in this invention is essentially an open-loop forward calculation and targeted spatial node addressing process based on ideal health data. Since the baseline state data originates from a healthy battery at the factory stage, the system has absolute confidence in the physical model parameters at this time, eliminating the need for a feedback correction mechanism. The specific extraction and deduction process is as follows: 1. Offline excitation and open-loop forward solution of multidimensional state matrix: The continuous reference state data sequence (measured terminal voltage, charge / discharge current, and surface temperature) recorded under standard cyclic conditions was used as the excitation vector and imported into an offline single-particle reduced-order electrochemical model (SPMe) with electrolyte diffusion dynamics. The time and spatial domains were discretized into a mesh using the finite difference method. The solid / liquid phase partial differential equations in the model were solved by forward difference in a separate loop, thereby generating a "prior global state matrix" containing all physical quantities of spatial nodes at each time step.

[0028] 2. Targeted extraction of lithium-ion concentration from the negative electrode solid surface: According to the partial differential equation of Fick's second law for solid-state applications in spherical coordinates, the solid-state lithium-ion concentration exhibits a radially non-uniform distribution. Within the aforementioned "global state matrix" of the porous negative electrode domain, the system precisely locates the spatial discrete grid of the microscopic spherical particles. The instantaneous concentration solution at the radial coordinates (i.e., the outermost solid-liquid interface surface of the particles) is extracted from this grid, thus establishing the current lithium-ion concentration at the negative electrode solid-state surface.

[0029] 3. Targeted extraction using a liquid-phase lithium-ion concentration gradient: Based on the electrolyte liquid phase mass conservation equation based on concentrated solution theory, the system extracts the local extreme values ​​of electrolyte lithium-ion concentration on both sides of the porous electrode (negative electrode current collector side and positive electrode separator side) along the one-dimensional thickness direction (x-axis) of the cell in the "global state matrix". A first-order spatial difference operation is performed to extract the concentration difference rate across the porous electrode as the liquid phase lithium-ion concentration gradient, which is used to accurately characterize the macroscopic liquid phase concentration polarization driving force.

[0030] 4. Targeted extraction of solid-liquid interfacial overpotential: The system extracts the solid-phase matrix potential from the global state matrix and subtracts the liquid-phase pore potential to obtain the "interphase potential difference". Then, the lithium ion concentration on the negative electrode solid surface extracted in step 2 is substituted into the open circuit potential function to calculate the local equilibrium potential. Finally, the local equilibrium potential is subtracted from the interphase potential difference, and the pure ohmic polarization voltage drop is stripped. The remaining net voltage value is extracted as the solid-liquid interface overpotential that drives charge transfer.

[0031] In one embodiment, constructing a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as a coordinate basis, and projecting it onto the three-dimensional electrochemical manifold space, includes: The obtained lithium ion concentration on the negative electrode solid surface, the lithium ion concentration gradient in the liquid phase, and the overpotential at the solid-liquid interface were subjected to extreme value normalization. The three normalized microscopic electrochemical parameters are configured as three mutually orthogonal linearly independent basis vectors in the topological space, thereby constructing an absolute coordinate system for the three-dimensional electrochemical manifold space; The current instantaneous scalar values ​​of the three normalized microscopic electrochemical parameters are mapped to three-dimensional spatial coordinates in the absolute coordinate system to locate the unique real-time concentration polarization state point.

[0032] Traditional battery management typically uses linear lookup tables based on one-dimensional macroscopic features such as voltage and current. The "manifold space" defined in this invention is a geometric abstraction of a nonlinear dynamic state. It does not refer to a real three-dimensional physical space, but rather, through extremum normalization, forces the three heterogeneous physical quantities—surface concentration, concentration gradient, and interface overpotential—to be configured as three mutually orthogonal basis vectors.

[0033] Preferably, the lithium-ion concentration on the solid-phase surface of the negative electrode represents the aggregation density or degree of lithium-ion congestion at the outermost solid-liquid interface of the negative electrode active material particles (such as graphite spheres). During charging, lithium ions move from the electrolyte to the surface of the negative electrode particles and then diffuse into the particle interior through solid-phase diffusion. If the charging current is too high or the temperature is too low, causing solid-phase diffusion to slow down, the lithium ions on the surface will not have enough time to diffuse inward and will accumulate in large quantities on the particle surface, resulting in a rapid increase in the lithium-ion concentration on the solid-phase surface of the negative electrode. When this concentration approaches the theoretical maximum saturation concentration of the negative electrode material, it means that the "cargo space" on the particle surface is full, and subsequent lithium ions will not be able to successfully insert into the crystal lattice, thus easily undergoing a side reaction on the surface to gain electrons and be reduced to metallic lithium (i.e., the material basis of lithium plating).

[0034] The liquid-phase lithium-ion concentration gradient characterizes the degree of spatial non-uniformity of lithium-ion distribution in the electrolyte along the entire thickness of the battery cell (spanning the positive electrode, separator, and porous negative electrode), i.e., the severity of concentration polarization. During battery operation, one electrode consumes lithium-ions from the electrolyte, while the other releases them. If the charge / discharge rate (reaction consumption rate) exceeds the liquid-phase diffusion rate of ions in the electrolyte, it will cause ion depletion at one end and ion accumulation at the other, forming a huge concentration gradient. An excessively large liquid-phase concentration gradient generates significant concentration polarization resistance, severely hindering reaction kinetics. Under extreme conditions (such as extremely cold environments causing a sudden increase in electrolyte viscosity and a sharp decrease in ionic conductivity), a large concentration gradient is the core reason why the terminal voltage instantly reaches the cutoff voltage and the usable battery capacity drops sharply; it also indirectly worsens the potential distribution in local areas.

[0035] The overpotential at the solid-liquid interface characterizes the net kinetic driving force at the solid-liquid interface that drives the charge transfer reaction (lithium insertion or delithiation) of lithium ions, i.e., the degree to which the actual interphase potential difference deviates from the thermodynamic equilibrium potential. According to the Butler-Folmer equation in electrochemistry, a "push" is required for the reaction to proceed, and this push is the overpotential. To maintain the set charge / discharge current, the system must generate a corresponding overpotential. It is the direct thermodynamic criterion for determining whether lithium plating has occurred in the battery. During charging, the equilibrium potential of a normal lithium insertion reaction is higher than the equilibrium potential for lithium metal deposition. However, the more severe the polarization, the lower the interface overpotential of the negative electrode will be. Once this overpotential falls below the thermodynamic critical point for the onset of lithium metal deposition, thermodynamic conditions will allow lithium metal to crystallize and precipitate directly at the solid-liquid interface. Crossing this red line means that lithium plating is occurring.

[0036] S2. Project the continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time into the three-dimensional electrochemical manifold space to generate the reference trajectory envelope volume composed of continuous state points, and define it as the intrinsic lithium insertion / extraction steady-state domain.

[0037] Historical test data of an ideal healthy battery was acquired under standard laboratory isothermal and cyclic conditions as baseline state data. The three microscopic electrochemical parameters mentioned above were extracted from this data, and a single, globally consistent "three-dimensional electrochemical manifold space" was constructed using this data as a coordinate basis. Continuous sampling values ​​of these three variables over time from the baseline state data were projected into this space. Due to the support of three independent dimensions, continuous state points form a smooth "baseline evolution trajectory" in space. The maximum envelope volume occupied by this trajectory in three-dimensional space is physically defined as the "intrinsic lithium insertion / extraction steady-state domain."

[0038] In one embodiment, the intrinsic lithium insertion / extraction steady-state domain represents the optimal operating range of the battery under standard healthy conditions, where the microscopic electrochemical reactions (mass diffusion and charge transfer) are fully coordinated and there is no accumulation of malignant polarization.

[0039] It is a "statically fixed reference volume" in the three-dimensional coordinate system. The system uses it as an "absolute scale" to measure real-time polarization deviation, and as a "target area" to which the state point is expected to be pulled back when executing pulse relaxation commands.

[0040] S3. Real-time acquisition of the current battery's terminal voltage, charging / discharging current, and surface temperature; inversion of the three microscopic electrochemical parameters in real time using a preset nonlinear state observer; projection onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point; and then updating the state space evolution trajectory by combining the current battery's historical concentration polarization state points.

[0041] In one embodiment, the three microscopic electrochemical parameters are retrieved in real time using a pre-set nonlinear state observer, including: The terminal voltage, charging and discharging current, and surface temperature are substituted as excitation vectors into a pre-set single-particle reduced-order electrochemical model with electrolyte diffusion dynamics. The partial differential equations are then solved to obtain the theoretical internal state prior estimation matrix and the theoretical terminal voltage. Calculate the dynamic residual between the theoretical terminal voltage and the terminal voltage, and use the extended Kalman filter algorithm to calculate the observer gain matrix at the current time step; Based on the observer gain matrix, the theoretical internal state prior estimation matrix is ​​corrected in a closed loop, and the posterior optimal state estimation matrix is ​​output. The concentration values ​​of the outermost spatial nodes of the negative electrode are extracted from the posterior optimal state estimation matrix as the lithium ion concentration on the solid surface of the negative electrode. The concentration difference rate between the two ends of the porous electrode is extracted as the lithium ion concentration gradient in the liquid phase. Based on the interphase potential difference and the stripped ohmic voltage drop resolved from the posterior optimal state estimation matrix, the overpotential of the solid-liquid interface is calculated.

[0042] In this embodiment, unlike the offline extraction in step S1, under actual outdoor operating conditions, due to the time-varying degradation of individual battery cells (aging leading to increased internal resistance and capacity decay), continuing to use the fixed single-particle reduced-order electrochemical model (SPMe) with electrolyte diffusion kinetics for open-loop calculation will result in severe model drift errors. Therefore, the "inversion of the three real-time microscopic electrochemical parameters" mentioned in step S3 is essentially a closed-loop feedback reconstruction process based on extended Kalman filtering (EKF). Although the algorithm architecture of the inversion process has changed, its spatial addressing nodes in the state matrix remain absolutely homologous and symmetrical with those in step S1, ensuring that the real-time state point and the reference steady-state domain are in the same absolutely consistent three-dimensional electrochemical manifold coordinate system. The specific inversion mechanism is as follows: 1. Equation of State Prediction and Dynamic Residual Generation: The real-time acquired terminal voltage, charge / discharge current, and surface temperature are used as excitation vectors and substituted into a pre-set single-particle reduced-order electrochemical model with electrolyte diffusion kinetics. It is important to explain in detail that in this nonlinear state observation system, the excitation vectors have a clear physical division of labor: the charge / discharge current and surface temperature primarily serve as system control inputs, driving the partial differential equations within the model to solve in time-step increments, thereby obtaining the "theoretical internal state prior estimation matrix" for the current time step, and mapping it to the theoretical terminal voltage; while the terminal voltage in the excitation vector serves as a reference benchmark for the measurement equations, used to subtract and compare with the previously calculated theoretical terminal voltage, calculating and generating the "dynamic residual" characterizing the systematic deviation of the current model due to battery aging or environmental disturbances.

[0043] 2. Calculation of the observer gain matrix and state closed-loop correction: The dynamic residuals described above, combined with the Jacobian matrix of the system state transition process and the system noise / observation noise covariance matrix, are substituted into the extended Kalman filter algorithm to calculate the observer gain matrix at the current time step, which has global convergence. Based on the observer gain matrix, the "theoretical internal state prior estimation matrix" is corrected by closed-loop negative feedback (i.e., the gain matrix is ​​multiplied by the dynamic residuals to form a compensation term, which is then superimposed on the prior matrix), forcibly pulling back the drift error, thereby outputting a "posterior optimal state estimation matrix" with extremely high physical fidelity.

[0044] 3. Homogeneous Targeted Spatial Feature Stripping: After obtaining the reconstructed "posterior optimal state estimation matrix", the system executes the spatial grid addressing logic completely consistent with step S1 for feature parsing: Directional Extraction: The concentration value of the outermost spatial node of the negative electrode is directionally extracted from the posterior optimal state estimation matrix as the real-time lithium ion concentration on the solid surface of the negative electrode; the concentration difference rate between the two ends of the porous electrode in the matrix is ​​extracted as the real-time liquid phase lithium ion concentration gradient; Based on the interphase potential difference and the stripped ohmic voltage drop (i.e., explicitly subtracting the voltage drop generated by the dynamic internal resistance) parsed from the posterior optimal state estimation matrix, the real-time solid-liquid interface overpotential is calculated.

[0045] In one embodiment, the real-time concentration polarization state point represents the "current state," signifying the severity of the overall micro-polarization within the battery at that instant. It is a specific "coordinate point" in a three-dimensional coordinate system. The system determines which control strategy to apply at that moment by calculating which geometric region (steady-state, accumulation, or fracture zone) this "point" falls into.

[0046] State-space evolution trajectory: Representing the "development trend," it signifies the accumulation process and direction of battery polarization states over a period of time (whether polarization intensifies or subsides). In a three-dimensional coordinate system, it is a "space curve" connecting consecutive past state points. By solving for the "derivative" and "tangent" of this curve, the system can predict in advance which safety threshold the battery is about to cross, thus achieving early prevention (feedforward control) rather than waiting for a fault to occur before taking remedial action.

[0047] S4. Based on the temperature correction factor obtained from the surface temperature mapping and the Nernst equation, the current lithium plating critical surface is calculated in the three-dimensional electrochemical manifold space. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain.

[0048] In one embodiment, the three-dimensional solid space enclosed by the lithium plating critical surface constitutes a dynamic lithium plating safety domain, and its dynamic boundary contraction mechanism is as follows: The electrolyte ionic conductivity constant and solid-phase diffusion coefficient in the single-particle reduced-order electrochemical model are corrected online and in real time by the temperature correction factor; the temperature correction factor characterizes the current temperature decay. As the temperature characterized by the temperature correction factor continues to decrease, the corrected electrolyte ionic conductivity constant and the solid-phase diffusion coefficient exhibit nonlinear decay, and the absolute value of the lithium plating initiation overpotential calculated after substituting into the Nernst equation decreases dynamically accordingly. The critical surface for lithium plating undergoes a three-dimensional geometric centripetal contraction toward the origin of the coordinate system in the three-dimensional electrochemical manifold space based on the reduced absolute value of the lithium plating initiation overpotential, resulting in a strictly monotonically decreasing envelope volume of the dynamic lithium plating safety domain.

[0049] Optionally, those skilled in the art should understand that the "dynamic lithium plating safety domain" is not a fixed, static lookup table threshold, but a three-dimensional topological subset with clear physical and geometric meaning that can adaptively evolve with the environment. The implementation principle and physical process of its dynamic boundary shrinkage are as follows: 1. Physical Conduction Layer 1: The system first acquires the battery surface temperature and maps it to a temperature correction factor characterizing the temperature decay. The main control chip substitutes this temperature correction factor online into a pre-set single-particle reduced-order electrochemical model (SPMe) with electrolyte diffusion kinetics. Using thermodynamic relationships such as the Arrhenius empirical formula, it corrects the two core microscopic transport constants within the model in real time: the electrolyte ionic conductivity constant and the solid-phase diffusion coefficient. As the ambient temperature continues to decrease, these two parameters inevitably exhibit exponential nonlinear decay.

[0050] 2. Second Layer of Physical Conduction: After obtaining the decayed microscopic kinetic parameters, the system uses them as variables to recalculate the thermodynamic boundary in the Nernst equation describing the electrochemical equilibrium potential. Under standard conditions at room temperature, the initial overpotential driving irreversible deposition of lithium metal on the negative electrode surface typically has a certain margin (manifested as a large negative value, e.g., -50mV). However, when the electrolyte ionic conductivity and solid-phase diffusion coefficient decay nonlinearly, the internal polarization resistance of the system increases sharply. The results of substituting into the Nernst equation show that at low temperatures, due to the drastic consumption of mass transfer polarization, only a very small net driving force is needed to force the solid-liquid interface potential to fall below the thermodynamic red line for lithium metal deposition.

[0051] Therefore, the absolute value of the calculated lithium plating initiation overpotential decreases dynamically (e.g., rapidly narrowing from |-50mV| at room temperature to |-10mV| at extreme cold). This decrease precisely quantifies the physical fact that the battery's "safety buffer margin" against micro-polarization is significantly compressed at low temperatures.

[0052] 3. Third Layer of Topological Geometric Mapping: The core innovation of this invention lies in the self-consistent mapping of the aforementioned one-dimensional thermodynamic threshold to a globally unified three-dimensional electrochemical manifold absolute coordinate system. In this three-dimensional coordinate system, the origin physically represents the battery in an absolutely ideal resting state without any polarization accumulation. The aforementioned calculated "absolute value of the lithium plating initiation overpotential" directly constitutes the geometric boundary condition (i.e., the lithium plating critical surface) for determining whether a state point has crossed the boundary in this three-dimensional space. As the absolute value of this initiation overpotential decreases, the lithium plating critical surface must retreat geometrically and topologically towards the direction of zero polarization, that is, directly undergo three-dimensional geometric centripetal contraction towards the origin of the coordinate system of the three-dimensional electrochemical manifold space. Since the critical surface, as an outer shell, continuously collapses inward, the envelope volume of the three-dimensional solid space (i.e., the dynamic lithium plating safety domain) enclosed by this surface strictly exhibits a monotonically decreasing evolution trend.

[0053] In one embodiment, the dynamic lithium plating safety domain characterizes the "real-time survival red line," representing the ultimate survival limit where, thermodynamically, lithium plating will absolutely not occur after considering the current ambient temperature. In a three-dimensional coordinate system, it is a volume that "dynamically contracts centripetally" as the temperature decreases. The system treats this as a high-priority hard constraint boundary; once the state point approaches or falls outside this volume, it immediately triggers forced derating or logic isolation.

[0054] S5. Calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / extraction steady-state domain and the dynamic lithium plating safety domain, and divide the battery microphysical state in real time into a uniform insertion / extraction region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region.

[0055] In one embodiment, the battery microphysical state is divided in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region, including: If the real-time concentration polarization state point is located within both the intrinsic lithium intercalation / deintercalation steady-state region and the dynamic lithium plating safety region, it is determined to be located in the uniform intercalation / deintercalation region. If the real-time concentration polarization state point is located inside the dynamic lithium plating safety domain and simultaneously outside the intrinsic lithium intercalation / deintercalation steady-state domain, then it is determined to be located in the concentration polarization accumulation region. If the real-time concentration polarization state point is located inside the intrinsic lithium insertion / extraction steady-state domain but outside the dynamic lithium plating safety domain, it is determined to be located in the local lithium plating boundary region. If the real-time concentration polarization state point is located outside both the intrinsic lithium insertion / extraction steady-state domain and the dynamic lithium plating safety domain, it is determined to be located in the lattice stress fracture region.

[0056] The preferred embodiment of the present invention discloses in detail a real-time partitioning mechanism for microscopic physical states based on spatial topological set relationships.

[0057] Those skilled in the art will understand that in traditional battery management systems, the independent threshold determination of multiple parameters can easily lead to logical deadlocks or circular arguments under complex operating conditions. This invention relies on the aforementioned established single, globally consistent three-dimensional electrochemical manifold absolute coordinate system, treating the "intrinsic lithium insertion / extraction steady-state domain" and the "dynamic lithium plating safety domain" as two independent three-dimensional geometric entities. A spatial point calculation function is used to calculate in real time the geometric spatial inclusion relationship between the "real-time concentration polarization state point" and the aforementioned two volume domains.

[0058] This determination mechanism strictly follows the MECE (mutually exclusive, mutually exclusive) Boolean logic operation principle, precisely dividing the battery's microscopic physical state into four absolutely mutually exclusive subsets. Its specific implementation process and physical characterization are as follows: 1. A uniform intercalation / deintercalation region indicates that the battery is in its optimal thermodynamic and kinetic comfort zone. At this point, the lithium-ion concentration on the negative electrode surface is not saturated, the liquid phase concentration gradient is small (mass transport is smooth), and the overpotential at the solid-liquid interface is ample. The internal electrochemical reaction exhibits a highly uniform intercalation / deintercalation state, with no risk of malignant polarization accumulation or lithium plating, falling within an absolutely safe range.

[0059] 2. The concentration polarization accumulation region is characterized by conditions typically occurring during high-rate continuous charge-discharge operations at room temperature. The overflow of the steady-state region from the state point indicates that due to the excessively rapid reaction rate, lithium ions are experiencing "traffic congestion" in the solid or liquid phase, and concentration polarization is accumulating rapidly. However, because the ambient temperature is normal and the thermodynamic safety threshold (safety domain boundary) is relatively wide, this degree of polarization has not yet fallen below the critical threshold for lithium metal deposition. This stage is considered a "sub-health warning stage," requiring the insertion of relaxation time to smooth out concentration polarization.

[0060] 3. The local lithium plating boundary region is the core diagnostic logic of this invention for dealing with extremely cold conditions. In conventional understanding, a state point within the steady-state domain is generally considered safe. However, under extremely cold or sudden cooling conditions, as described in step S4, the "dynamic lithium plating safety domain" will drastically shrink towards the origin of the coordinate system. This shrinkage causes legitimate state points that were originally within the steady-state domain to suddenly be "exposed" outside the shrunken safety domain. Even if the charging and discharging current is not exceeded and the polarization deviation appears normal (not outside the steady-state domain), the extreme cold significantly lowers the thermodynamic lithium plating threshold, and metallic lithium already possesses the thermodynamic conditions for crystallization at local microscopic interfaces. This topological determination extremely accurately captures the hidden lithium plating fault caused by threshold drift at low temperatures, triggering subsequent closed-loop derating and current limiting intervention.

[0061] 4. Lattice Stress Fracture Region Characterization: This region indicates that the battery has entered the most dangerous stage of failure deterioration. The state point has simultaneously deviated from the health baseline and breached the safety threshold. At this point, not only has severe macroscopic lithium plating occurred, but also a huge concentration gradient stress has formed on the surface of the negative electrode particles due to extreme congestion and over-intercalation of lithium ions. When this stress exceeds the yield limit of the graphite lattice, it will lead to micromechanical fracture of the particles, large-area tearing of the SEI film, and continuous side reactions at the new interface. This state represents irreversible physical damage, and the highest priority logic isolation must be implemented immediately to break the thermal runaway cascade.

[0062] S6. When it is determined that the location is in the uniform insertion / deintercalation region, maintain the current charge / discharge strategy; when it is determined that the location is in the concentration polarization accumulation region, generate and execute a duty cycle pulse relaxation instruction; when it is determined that the location is in the local lithium plating boundary region, calculate the overpotential overshoot across the lithium plating critical surface, and generate and execute a dynamic derating current limiting instruction based on the overpotential overshoot; when it is determined that the location is in the lattice stress fracture region, generate and execute a deep discharge and fault logic isolation instruction.

[0063] In one embodiment, generating and executing a duty cycle pulse relaxation instruction includes: Calculate the shortest Euclidean distance from the real-time concentration polarization state point to the outer boundary of the intrinsic lithium insertion / deintercalation steady-state domain, and solve for the derivative vector of the state space evolution trajectory to obtain the polarization divergence direction. The shortest Euclidean distance and the polarization divergence direction are input into a preset fuzzy logic reasoning system that establishes a mapping relationship between distance, direction and zero current intermittent time constant, and the zero current intermittent time constant required to force the internal concentration polarization to subside is calculated. Based on the zero-current intermittent time constant, the duty cycle of the pulse width modulation signal of the underlying battery management main control chip is adjusted, and a relaxation period is forcibly inserted into the continuous charge and discharge command to cut off the charge transfer reaction.

[0064] In one embodiment, calculating the overpotential overshoot across the lithium plating critical surface includes: Within the three-dimensional electrochemical manifold space, the state space evolution trajectory is extracted and updated in real time. Solve for the geometric intersection of the extension direction of the tangent of the state space evolution trajectory and the critical surface of lithium plating at the current moment in space, and define the interface overpotential at this intersection as the safety critical clamping threshold. The absolute difference between the solid-liquid interface overpotential and the safety critical clamping threshold at the current instant of the real-time concentration polarization state point is calculated as the overpotential overshoot, which characterizes the degree of thermodynamic overshoot.

[0065] In one embodiment, generating and executing a dynamic rate reduction and limiting instruction includes: The overpotential overshoot is used as an error feedback signal input into the closed-loop adaptive proportional-integral-derivative controller to calculate the charge and discharge power compensation term that needs to be reduced. Based on the charging and discharging power compensation term, the physical threshold of the maximum allowable charging current of the current battery is reduced proportionally. The updated maximum allowable charging current physical threshold is sent to the main controller of the battery pack to cut off the power, forcing the evolution trajectory of the real-time concentration polarization state point in the next time step to be restricted back to the dynamic lithium plating safety domain, thereby achieving closed-loop safety traction.

[0066] In one embodiment, generating and executing deep discharge and fault logic isolation instructions includes: A forced deep discharge request is generated to control the continuous discharge of faulty battery cells that have been determined to have lattice stress cracking, until the terminal voltage of the faulty battery cell drops below the preset thermodynamic absolute safe potential lower limit. The allowable charge / discharge power threshold of the faulty battery cell is forcibly set to zero, and subsequent charge / discharge control commands for the faulty battery cell are blocked to achieve fault isolation at the pure logic level.

[0067] In one embodiment, when the state point is located in the uniform insertion / deintercalation region, the current charge / discharge strategy is maintained, the battery is in the optimal thermodynamic and kinetic equilibrium region, there is no risk of polarization, the system does not intervene, and it maintains normal operation.

[0068] When the state point is located in the concentration polarization accumulation region, a duty cycle pulse relaxation command is generated and executed. Polarization begins to accumulate but has not yet reached the lithium plating threshold, requiring flexible mitigation of concentration polarization. The shortest Euclidean distance from the current "state point" to the steady-state domain boundary (representing the polarization deviation) is calculated, and the polarization divergence direction (representing the deterioration trend) is obtained by differentiating the "evolution trajectory". These two values ​​are input into a fuzzy logic inference system to directly calculate the required zero-current intermittent time constant. Based on this, the underlying PWM duty cycle is modified, and a zero-current "rest period" of this duration is forcibly inserted into the continuous charge / discharge messages, allowing the congested lithium ions to naturally diffuse and mitigate according to Fick's law.

[0069] When the state point is located in the local lithium plating boundary region, a dynamic derating and current limiting command is generated and executed. Since the state point has approached or exceeded the contracted safety red line, the reaction rate must be forcibly suppressed. The geometric intersection of the "tangent" of the historical evolution trajectory and the current "lithium plating critical surface" in three-dimensional space is determined, and the overpotential at this intersection is extracted as the safety critical clamping threshold. The overpotential overshoot is obtained by subtracting this threshold from the current actual overpotential. This overshoot is used as an error input to the PID controller to calculate the power compensation term to be reduced, and accordingly, the physical threshold of the current battery's maximum allowable charging current is lowered. Through forced current reduction, the polarization trajectory of the next time step is forced back into the dynamic safety domain.

[0070] When the state point is located in the lattice stress fracture region, a deep discharge and fault logic isolation instruction is generated and executed. At this point, irreversible microscopic damage occurs, posing a risk of thermal runaway, necessitating complete melting. First, the discharge dissipation circuit is triggered, forcing the faulty battery to discharge and releasing the terminal voltage below the thermodynamic absolute safe potential lower limit, thus removing the energy source for its violent reaction. Subsequently, at the software level, the charge / discharge power threshold (SOP) of the faulty cell is permanently set to zero, and related control messages are masked, achieving complete isolation at the pure logic level and preventing the spread of thermal runaway.

[0071] This invention breaks through the limitations of traditional outdoor energy storage systems that rely on macroscopic parameters for early warning. It maps the inverted microscopic electrochemical parameters to a three-dimensional manifold topological space, quantifying the concentration polarization phenomenon inside the battery with high precision. This significantly improves monitoring sensitivity, provides microscopic evolutionary evidence for fault diagnosis, and fundamentally eliminates early-stage hidden dangers. It innovatively introduces a dynamic boundary contraction mechanism that decays with temperature. For outdoor variable-temperature environments, a temperature correction factor drives the centripetal contraction of the lithium plating critical surface, conforming to the law of intensified polarization at low temperatures. This mechanism accurately locks the safety boundary at each temperature, avoiding performance limitations at room temperature while ensuring safety in extremely cold conditions. A refined microscopic state classification criterion is constructed, matched with a tiered active safety strategy. When polarization accumulation or lithium plating tendency occurs, intervention is carried out by generating pulse relaxation commands or lowering the charging limit in a closed loop. This control closed loop can forcibly pull the state trajectory back to the safe region before lattice fracture, completely nipping thermal runaway in the bud.

[0072] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0073] The present invention discloses a non-transitory computer-readable storage medium storing computer instructions, which, when executed by a processor, cause the processor to perform the above-described method.

[0074] Those skilled in the art will understand that all or part of the steps in the above methods can be implemented by a program instructing related hardware (e.g., processor, FPGA, ASIC, etc.), and the program can be stored in a readable storage medium, such as a read-only memory, a disk, or an optical disk. All or part of the steps in the above embodiments can also be implemented using one or more integrated circuits. Accordingly, each module in the above embodiments can be implemented in hardware, such as by using integrated circuits to implement its corresponding function, or it can be implemented as a software functional module, such as by a processor executing a program / instruction stored in memory to implement its corresponding function. The embodiments of the present invention are not limited to any particular combination of hardware and software.

[0075] Furthermore, the functional units in the various embodiments of this document can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0076] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this article. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0077] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for fault diagnosis of outdoor energy storage battery status monitoring, characterized in that, include: The baseline state data of an ideal healthy battery under standard cycle conditions is obtained in advance. Three microscopic electrochemical parameters are extracted from the data: lithium ion concentration on the negative electrode solid phase surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface. A three-dimensional electrochemical manifold space is constructed using the three microscopic electrochemical parameters as the coordinate basis. The continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time are projected into the three-dimensional electrochemical manifold space to generate the reference trajectory envelope volume composed of continuous state points, and this volume is defined as the intrinsic lithium insertion / extraction steady-state domain. The current battery's terminal voltage, charging and discharging current, and surface temperature are collected in real time. The three microscopic electrochemical parameters are inverted in real time through a preset nonlinear state observer and projected onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point. Then, the state space evolution trajectory is updated by combining the current battery's historical concentration polarization state points. Based on the temperature correction factor obtained from the surface temperature mapping and the Nernst equation, the current lithium plating critical surface is calculated in the three-dimensional electrochemical manifold space. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain. Calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / deintercalation steady-state domain and the dynamic lithium plating safety domain, and divide the battery microphysical state in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region. When the location is determined to be in the uniform insertion / deintercalation region, the current charge / discharge strategy is maintained; when the location is determined to be in the concentration polarization accumulation region, a duty cycle pulse relaxation instruction is generated and executed; when the location is determined to be in the local lithium plating boundary region, the overpotential overshoot across the lithium plating critical surface is calculated, and a dynamic derating current limiting instruction is generated and executed based on the overpotential overshoot; when the location is determined to be in the lattice stress fracture region, a deep discharge and fault logic isolation instruction is generated and executed.

2. The method according to claim 1, characterized in that, The three microscopic electrochemical parameters, including those mentioned above, are retrieved in real time using a pre-set nonlinear state observer: The terminal voltage, charging and discharging current, and surface temperature are substituted as excitation vectors into a pre-set single-particle reduced-order electrochemical model with electrolyte diffusion dynamics. The partial differential equations are then solved to obtain the theoretical internal state prior estimation matrix and the theoretical terminal voltage. Calculate the dynamic residual between the theoretical terminal voltage and the terminal voltage, and use the extended Kalman filter algorithm to calculate the observer gain matrix at the current time step; Based on the observer gain matrix, the theoretical internal state prior estimation matrix is ​​corrected by closed-loop method, and the posterior optimal state estimation matrix is ​​output. The concentration values ​​of the outermost spatial nodes of the negative electrode are extracted from the posterior optimal state estimation matrix as the lithium ion concentration on the solid surface of the negative electrode. The concentration difference rate between the two ends of the porous electrode is extracted as the lithium ion concentration gradient in the liquid phase. Based on the interphase potential difference and the stripped ohmic voltage drop resolved from the posterior optimal state estimation matrix, the overpotential of the solid-liquid interface is calculated.

3. The method according to claim 2, characterized in that, Constructing a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as a coordinate basis, and projecting it onto the three-dimensional electrochemical manifold space, includes: The obtained lithium ion concentration on the negative electrode solid surface, the lithium ion concentration gradient in the liquid phase, and the overpotential at the solid-liquid interface were subjected to extreme value normalization. The three normalized microscopic electrochemical parameters are configured as three mutually orthogonal linearly independent basis vectors in the topological space, thereby constructing an absolute coordinate system for the three-dimensional electrochemical manifold space; The current instantaneous scalar values ​​of the three normalized microscopic electrochemical parameters are mapped to three-dimensional spatial coordinates in the absolute coordinate system to locate the unique real-time concentration polarization state point.

4. The method according to claim 3, characterized in that, The three-dimensional solid space enclosed by the lithium plating critical surface constitutes a dynamic lithium plating safety domain, and its dynamic boundary contraction mechanism is as follows: The electrolyte ionic conductivity constant and solid-phase diffusion coefficient in the single-particle reduced-order electrochemical model are corrected online and in real time by the temperature correction factor; the temperature correction factor characterizes the current temperature decay. As the temperature characterized by the temperature correction factor continues to decrease, the corrected electrolyte ionic conductivity constant and the solid-phase diffusion coefficient exhibit nonlinear decay, and the absolute value of the lithium plating initiation overpotential calculated after substituting into the Nernst equation decreases dynamically accordingly. The critical surface for lithium plating undergoes a three-dimensional geometric centripetal contraction toward the origin of the coordinate system in the three-dimensional electrochemical manifold space based on the reduced absolute value of the lithium plating initiation overpotential, resulting in a strictly monotonically decreasing envelope volume of the dynamic lithium plating safety domain.

5. The method according to claim 4, characterized in that, The method of dividing the battery's microscopic physical state into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region in real time includes: If the real-time concentration polarization state point is located within both the intrinsic lithium intercalation / deintercalation steady-state region and the dynamic lithium plating safety region, it is determined to be located in the uniform intercalation / deintercalation region. If the real-time concentration polarization state point is located inside the dynamic lithium plating safety domain and simultaneously outside the intrinsic lithium intercalation / deintercalation steady-state domain, then it is determined to be located in the concentration polarization accumulation region. If the real-time concentration polarization state point is located inside the intrinsic lithium insertion / extraction steady-state domain but outside the dynamic lithium plating safety domain, it is determined to be located in the local lithium plating boundary region. If the real-time concentration polarization state point is located outside both the intrinsic lithium insertion / extraction steady-state domain and the dynamic lithium plating safety domain, it is determined to be located in the lattice stress fracture region.

6. The method according to claim 1 or 5, characterized in that, The generation and execution of the duty cycle pulse relaxation instruction includes: Calculate the shortest Euclidean distance from the real-time concentration polarization state point to the outer boundary of the intrinsic lithium insertion / deintercalation steady-state domain, and solve for the derivative vector of the state space evolution trajectory to obtain the polarization divergence direction. The shortest Euclidean distance and the polarization divergence direction are input into a preset fuzzy logic reasoning system that establishes a mapping relationship between distance, direction and zero current intermittent time constant, and the zero current intermittent time constant required to force the internal concentration polarization to subside is calculated. Based on the zero-current intermittent time constant, the duty cycle of the pulse width modulation signal of the underlying battery management main control chip is adjusted, and a relaxation period is forcibly inserted into the continuous charge and discharge command to cut off the charge transfer reaction.

7. The method according to claim 1 or 5, characterized in that, The calculation of the overpotential overshoot across the lithium plating critical surface includes: Within the three-dimensional electrochemical manifold space, the state space evolution trajectory is extracted and updated in real time. Solve for the geometric intersection of the extension direction of the tangent of the state space evolution trajectory and the critical surface of lithium plating at the current moment in space, and define the interface overpotential at this intersection as the safety critical clamping threshold. The absolute difference between the solid-liquid interface overpotential and the safety critical clamping threshold at the current instant of the real-time concentration polarization state point is calculated as the overpotential overshoot, which characterizes the degree of thermodynamic overshoot.

8. The method according to claim 7, characterized in that, Generate and execute dynamic rate reduction and limiting instructions, including: The overpotential overshoot is used as an error feedback signal input into the closed-loop adaptive proportional-integral-derivative controller to calculate the charge and discharge power compensation term that needs to be reduced. Based on the charging and discharging power compensation term, the physical threshold of the maximum allowable charging current of the current battery is reduced proportionally. The updated maximum allowable charging current physical threshold is sent to the main controller of the battery pack to cut off the power, forcing the evolution trajectory of the real-time concentration polarization state point in the next time step to be restricted back to the dynamic lithium plating safety domain, thereby achieving closed-loop safety traction.

9. The method according to claim 1 or 5, characterized in that, The generation and execution of deep discharge and fault logic isolation instructions include: A forced deep discharge request is generated to control the continuous discharge of faulty battery cells that have been determined to have lattice stress cracking, until the terminal voltage of the faulty battery cell drops below the preset thermodynamic absolute safe potential lower limit. The allowable charge / discharge power threshold of the faulty battery cell is forcibly set to zero, and subsequent charge / discharge control commands for the faulty battery cell are blocked to achieve fault isolation at the pure logic level.

10. An outdoor energy storage battery state monitoring and fault diagnosis system, characterized in that, include: The manifold space construction module is used to pre-acquire the baseline state data of an ideal healthy battery under standard cycle conditions, extract three microscopic electrochemical parameters from them: lithium ion concentration on the negative electrode solid phase surface, lithium ion concentration gradient in the liquid phase, and overpotential at the solid-liquid interface, and construct a three-dimensional electrochemical manifold space using the three microscopic electrochemical parameters as the coordinate basis. The steady-state domain definition module is used to project the continuous sampled values ​​of the three microscopic electrochemical parameters in the reference state data as a function of time into the three-dimensional electrochemical manifold space, generate the reference trajectory envelope volume composed of continuous state points, and define it as the intrinsic lithium insertion / extraction steady-state domain. The state inversion and trajectory generation module is used to collect the current battery's terminal voltage, charging and discharging current and surface temperature in real time. It uses a preset nonlinear state observer to invert the three microscopic electrochemical parameters in real time and projects them onto the three-dimensional electrochemical manifold space to generate a single real-time concentration polarization state point. Then, it combines the current battery's historical concentration polarization state points to update and generate the state space evolution trajectory. The dynamic safety domain calculation module is used to calculate the current lithium plating critical surface in the three-dimensional electrochemical manifold space based on the temperature correction factor obtained by the surface temperature mapping and the Nernst equation. The three-dimensional solid space enclosed by the lithium plating critical surface constitutes the dynamic lithium plating safety domain. The state space mapping partitioning module is used to calculate the positional combination relationship between the real-time concentration polarization state point and the intrinsic lithium insertion / deintercalation steady-state domain and the dynamic lithium plating safety domain, and to divide the battery micro-physical state in real time into a uniform insertion / deintercalation region, a concentration polarization accumulation region, a local lithium plating boundary region, and a lattice stress fracture region. The fault diagnosis and safety control execution module is used to maintain the current charge and discharge strategy when it is determined that the uniform insertion / distraction region is located. When it is determined that the location is in the concentration polarization accumulation region, a duty cycle pulse relaxation instruction is generated and executed; when it is determined that the location is in the local lithium plating boundary region, the overpotential overshoot across the lithium plating critical surface is calculated, and a dynamic derating current limiting instruction is generated and executed based on the overpotential overshoot; when it is determined that the location is in the lattice stress fracture region, a deep discharge and fault logic isolation instruction is generated and executed.

Images