Low-voltage energy storage system at extremely low temperature

Abstract

The invention relates to the technical field of battery management or energy storage system control, in particular to a low-voltage energy storage system at extremely low temperature. The system comprises a multi-dimensional state sensing unit, a joint estimation unit, a dynamic risk quantification unit and an adaptive decision-making unit. The system recognizes temperature data pollution by resolving a temperature gradient index; the method is characterized in that the core temperature is calculated according to a pollution mark and a minimum reading, SOH damage and SOC are quantified, and then a comprehensive risk index and a safety distance are calculated; when the model collapses or the safety distance is insufficient, the system starts self-heating or rejects discharging in a self-adaptive manner; according to the method, active identification of sensor data pollution is realized, misjudgment caused by dependence on virtual high data can be avoided, and the conservative property and safety of estimation are ensured.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of battery management or energy storage system control, in particular to a low-voltage energy storage system under extreme low temperature. BACKGROUND

[0002] In the operation of an energy storage system under an extreme low temperature environment, a battery management system (BMS) faces severe challenges; temperature sensors deployed are easily disturbed by external asymmetric heat shocks, leading to data pollution and false high readings; traditional BMS algorithms lack the ability to perceive the credibility of data, and are prone to optimistic misjudgment relying on such data; low temperature can cause instantaneous capacity loss of the battery cell, and the traditional SOC estimation does not correct this effect, resulting in systematic errors in SOC estimation and failure to accurately reflect the real available power; under the dual challenges of data uncertainty and state estimation failure, the BMS is difficult to make safe discharge or heating decisions, which may lead to the operation of the system in a dangerous boundary, threatening the safety and long-term life of the system; therefore, how to accurately estimate the state of the system and make safe adaptive decisions under extreme low temperature and sensor uncertainty is a technical problem to be solved. SUMMARY

[0003] To solve the above technical problems, the application provides a low-voltage energy storage system under extreme low temperature, and in particular, the technical solution of the application comprises:

[0004] A multi-dimensional state perception unit is configured to collect an original data set and calculate a temperature gradient index; compare the temperature gradient index with a preset normal temperature gradient threshold to obtain a temperature data pollution flag; and combine the temperature data pollution flag, the mean value and the lowest reading of the original data set to calculate an estimated core temperature;

[0005] A joint estimation unit is configured to quantize an irreversible fractional damage of SOH according to the estimated core temperature and the collected discharge current, update a current SOH fraction, correct a dynamic available capacity reference based on the current SOH fraction and the estimated core temperature, and estimate a current SOC fraction based on the dynamic available capacity reference;

[0006] A dynamic risk quantization unit is configured to generate a power risk index according to the current SOC fraction, generate a temperature risk index according to the estimated core temperature, generate a model collapse risk index according to the temperature data pollution flag, and combine the power risk index, the temperature risk index and the model collapse risk index to obtain a comprehensive risk index by weighted summation; calculate a quantitative safety distance index according to the comprehensive risk index; and calculate a safety distance change rate of the quantitative safety distance index;

[0007] The adaptive decision unit is configured to output a start self-heating decision signal when the model collapse risk index is greater than zero and the estimated core temperature is less than a preset upper temperature limit; otherwise, output a discharge rejection decision signal when the quantified safety distance index is less than a preset first safety threshold or the safety distance change rate is less than a preset first change rate threshold; otherwise, output a discharge permission decision signal.

[0008] Preferably, the estimation process of the estimated core temperature comprises:

[0009] When the temperature data pollution flag is zero, the mean of the original data set is set as the estimated core temperature;

[0010] When the temperature data pollution flag is one, the lowest reading of the original data set is set as the estimated core temperature.

[0011] Preferably, the quantification process of the SOH irreversible fractional impairment comprises:

[0012] Obtaining the charge amount in the current time step;

[0013] And based on the estimated core temperature, the preset reference temperature, the charge amount, and the preset temperature-sensitive coefficient and the basic capacity impairment rate, the SOH irreversible fractional impairment is calculated.

[0014] Preferably, the correction process of the dynamic available capacity benchmark comprises:

[0015] Based on the estimated core temperature, a preset lookup table or a fitting function is queried to obtain a low-temperature capacity reduction factor;

[0016] The nominal capacity, the current SOH fraction, and the low-temperature capacity reduction factor are multiplied to obtain the dynamic available capacity benchmark.

[0017] Preferably, the calculation process of the power risk index comprises:

[0018] The difference between the preset SOC safety boundary and the power locking boundary is divided by the difference between the current SOC fraction and the power locking boundary to obtain a ratio;

[0019] And the ratio is subjected to exponential operation of a preset SOC risk-sensitive coefficient to obtain the power risk index.

[0020] Preferably, the calculation process of the temperature risk index comprises:

[0021] The difference between the preset temperature safety boundary and the preset temperature locking boundary is set as a numerator;

[0022] The difference between the estimated core temperature and the temperature locking boundary is set as a denominator;

[0023] The ratio is obtained by dividing the numerator by the denominator, and the temperature risk index is obtained by performing exponential operation on the ratio with a preset temperature risk sensitivity coefficient.

[0024] Preferably, the calculation process of the model collapse risk index comprises:

[0025] The temperature data pollution flag is multiplied by a preset pollution risk factor to obtain the model collapse risk index.

[0026] Preferably, the quantitative safety distance index is obtained by reciprocal calculation of the comprehensive risk index;

[0027] The safety distance change rate is calculated by first-order difference of the quantitative safety distance index within a preset time step.

[0028] Compared with the prior art, the present application has the following beneficial effects:

[0029] 1. The system can effectively identify sensor data pollution caused by external thermal shock by calculating the temperature gradient index and comparing it with the threshold value; when identifying pollution, the system automatically uses the lowest temperature reading as the estimated core temperature, avoiding optimistic misjudgment caused by relying on false high data, and ensuring the conservativeness and safety of subsequent estimation;

[0030] 2. The system combines the estimated core temperature with the health state score to dynamically correct the available capacity benchmark; this correction takes into account the instantaneous capacity loss of the battery caused by low temperature, solving the problem of systematic error in traditional SOC estimation at low temperature due to uncorrected capacity, so that the estimated SOC score can accurately reflect the real available power;

[0031] 3. The system innovatively quantifies the sensor data pollution state as a model collapse risk index, and performs weighted summation of the power risk and temperature risk to obtain a comprehensive risk; this multi-dimensional risk quantification method enables the system to first perceive the confidence of its estimation results, providing a unified quantitative basis for formulating robust decisions;

[0032] 4. The system introduces the quantitative safety distance index and its change rate, so that the decision unit can predict the dynamic development trend of the risk; when detecting data pollution, the system preferentially starts self-heating to restore data reliability; when the data is reliable, the system intervenes in advance according to the safety distance and its change rate to refuse high-risk discharge, realizing closed-loop adaptive safety management under uncertainty. BRIEF DESCRIPTION OF DRAWINGS

[0033] The present application will be further explained in conjunction with the drawings and examples:

[0034] Figure 1 is a structural diagram of the system of the present application. DETAILED DESCRIPTION

[0035] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with specific examples.

[0036] Example 1:

[0037] Please refer to Figure 1 The low-pressure energy storage system under extreme low temperature comprises:

[0038] The multi-dimensional state sensing unit is used to collect the original data set and calculate the temperature gradient index; compare the temperature gradient index with the preset normal temperature gradient threshold to obtain the temperature data pollution flag; and combine the temperature data pollution flag, the mean value and the lowest reading of the original data set to calculate the estimated core temperature.

[0039] The joint estimation unit is used to quantize the irreversible fractional damage of SOH according to the estimated core temperature and the collected discharge current, update the current SOH fraction, correct the dynamic available capacity reference based on the current SOH fraction and the estimated core temperature, and estimate the current SOC fraction based on the dynamic available capacity reference.

[0040] The dynamic risk quantization unit is used to generate the power risk index according to the current SOC fraction, generate the temperature risk index according to the estimated core temperature, generate the model collapse risk index according to the temperature data pollution flag, and combine the power risk index, the temperature risk index and the model collapse risk index to obtain the comprehensive risk index by weighted summation; calculate the quantitative safety distance index according to the comprehensive risk index; and calculate the safety distance change rate of the quantitative safety distance index.

[0041] The adaptive decision unit is used to output a start self-heating decision signal when the model collapse risk index is greater than zero and the estimated core temperature is less than the preset upper temperature limit; otherwise, output a discharge rejection decision signal when the quantitative safety distance index is less than the preset first safety threshold or the safety distance change rate is less than the preset first change rate threshold; otherwise, output a discharge permission decision signal.

[0042] The low-pressure energy storage system under extreme low temperature of the present embodiment comprises an energy storage body and a set of algorithm-based battery management system control unit, which integrates a multi-dimensional state sensing unit, a joint estimation unit, a dynamic risk quantization unit and an adaptive decision unit to form a closed-loop management system for coping with extreme low temperature challenges; the extreme low temperature in the present application refers to an environment temperature lower than -20°C, for example, a temperature environment in the interval of -40°C to -20°C.

[0043] Multi-dimensional state sensing unit

[0044] The unit aims to perceive the temperature distribution state inside the energy storage system in real time, and identify data pollution of sensor readings in an extremely low-temperature environment to generate a reliable core temperature as input for subsequent estimation and decision-making;

[0045] The unit collects the original data set through a multi-point NTC temperature sensor array deployed inside the energy storage system ;

[0046] Temperature gradient index The unit calculates the temperature gradient index , which represents the uniformity and symmetry of the system thermal state; defined as the difference between the highest reading and the lowest reading in the original data set ;

[0047] ;

[0048] wherein, and are real-time measurement data from NTC sensors; the index is designed to distinguish between normal internal heat dissipation and external asymmetric heat shock;

[0049] Temperature data pollution flag The unit compares the temperature gradient index with the preset normal temperature gradient threshold to obtain the temperature data pollution flag ; is a binary flag used to quickly identify the state of data being polluted by an external asymmetric heat source, resulting in an unreliable state;

[0050] ;

[0051] wherein, is set based on modeling of the system thermal physical state, representing the maximum normal temperature difference that can be generated by internal self-heating in an extreme low-temperature environment; is obtained offline by running different self-heating strategies and recording the statistical extreme value of under standard -40°C non-interference conditions;

[0052] Estimation of core temperature The unit calculates the estimated core temperature by combining the temperature data pollution flag , the mean value of the original data set , and the lowest reading ; see the embodiments for specific calculation logic;

[0053] Joint estimation unit

[0054] This unit aims to achieve a health status fraction of the energy storage system under high and low temperature coupling environments. and fraction of state of charge Dynamics, coupling, and accurate estimation;

[0055] SOH Update: This unit is based on estimated core temperature. and the collected discharge current Quantifying irreversible fractional damage of SOH The score is then subtracted from the SOH score of the previous period to update the current health status score. That is, the SOH fraction This enables modeling of the cumulative damage caused by low temperature and current over long-term lifespan.

[0056] Dynamic available capacity baseline Correction: This unit incorporates the current SOH score. With estimated core temperature The dynamic available capacity baseline is obtained by correction. ; As the denominator of the ampere-hour integral, it solves the systematic error caused by instantaneous capacity loss in traditional SOC estimation under low temperature environment;

[0057] SOC estimation: This unit uses the ampere-hour integral method, based on a dynamic available capacity baseline. As the current total capacity, and combined with the collected discharge current By integrating, the fraction of the current state of charge is estimated. That is, SOC score ;

[0058] Dynamic Risk Quantification Unit

[0059] This unit aims to quantify the multiple failure risks of energy storage systems into a single risk indicator. And use this indicator to establish the concept of safety margin;

[0060] Itemized risk generation: This unit is based on the current SOC score. Generate a power risk index Based on the estimated core temperature Generate a temperature risk index Pollution indicators based on temperature data Generate model collapse risk index ;

[0061] Overall Risk Index This unit incorporates the electricity risk index. temperature risk index model collapse risk index , and the integrated risk index is obtained by weighted summation ;

[0062] ;

[0063] wherein, is linearly weighted by three dimensionless sub-item risks, which is used to quantify the dangerous degree of the system distance from the multi-dimensional failure boundary; the risk weight coefficient is trained by offline executing a deep reinforcement learning model to maximize the long-term SOH score as the objective function, so as to determine its optimal relative weight;

[0064] Safety distance quantification: this unit calculates the quantitative safety distance index according to the integrated risk index , wherein is the inverse of ;

[0065] Safety distance change rate: this unit calculates the safety distance change rate of the quantitative safety distance index, which represents the dynamic change trend of the risk;

[0066] Adaptive decision unit

[0067] This unit aims to execute the adaptive control logic of risk hedging based on the quantified multi-dimensional risk and safety distance;

[0068] Rule 1: when the model collapse risk index is greater than zero and the estimated core temperature is less than the preset upper limit of temperature , output the start self-heating decision signal ; is the preset upper limit of battery tolerance temperature, which is usually set according to the safety specification book of the battery cell manufacturer; this rule aims to actively eliminate the data pollution source and restore the data credibility;

[0069] Rule 2: otherwise, when the quantitative safety distance index is less than the preset first safety threshold or the safety distance change rate is less than the preset first change rate threshold , output the discharge rejection decision signal ; and The values of and are determined by the long-term SOH score protection target and the allowed discharge decision signal The demand is balanced, and the offline optimization model is obtained; is a negative threshold value indicating a safe distance is being reduced at an unacceptable rate;

[0070] Rule 3: Otherwise, output the discharge permission decision signal ;

[0071] The system avoids optimistic misjudgment based on false high sensor readings through identification of data pollution , and forcibly adopts a conservative core temperature ; based on joint dynamic correction of state of health SOH and state of charge SOC , eliminating systematic errors in SOC estimation at low temperatures; finally, by integrating the model collapse risk index into the comprehensive risk model , and based on the safety distance and its rate of change , risk hedging is performed to ensure that the system can continue to operate reliably and safely under the dual challenges of extreme low temperature and sensor uncertainty, and maximize the long-term life of the battery.

[0072] Embodiment 2:

[0073] The estimation process of the estimated core temperature includes:

[0074] When the temperature data pollution flag is zero, set the mean of the original data set as the estimated core temperature;

[0075] When the temperature data pollution flag is one, set the lowest reading of the original data set as the estimated core temperature.

[0076] This embodiment is a specific refinement of the multi-dimensional state perception unit in Embodiment 1 to estimate the estimated core temperature ; this estimation logic uses the temperature data pollution flag as a logic switch to address the problem of false high sensor data caused by external asymmetric thermal shock, ensuring is a conservative estimate closest to the true thermal state inside the cell;

[0077] Logic 1: When the temperature data pollution flag is zero, indicating that the sensor data is reliable; at this time, set the mean of the original data set as the estimated core temperature ; is the arithmetic mean of the original data set , which most accurately represents the average thermal state of the entire energy storage system under reliable data conditions;

[0078] Logic two: when temperature data pollution flag is 1, indicating that the sensor data is not reliable; at this time, the system takes the most conservative strategy, setting the minimum reading of the original data set as the estimated core temperature ; is the minimum value in the core temperature

[0079] This estimation logic can be expressed in the form of weighted sum:

[0080] ;

[0081] Through the flag bit, the core temperature is conservatively switched, effectively avoiding the situation where the system makes dangerous discharge decisions based on false optimistic temperature when the temperature reading is artificially high due to external asymmetric heat shock; this conservative estimation under data pollution greatly improves the safety margin of the system in extreme low temperature environments.

[0082] Embodiment 3:

[0083] The quantification process of SOH irreversible fractional damage includes:

[0084] Obtain the charge amount in the current time step;

[0085] and based on the estimated core temperature, the preset reference temperature, the charge amount, and the preset temperature sensitivity coefficient and the basic capacity damage rate, the SOH irreversible fractional damage is calculated.

[0086] This embodiment is a specific refinement of the joint estimation unit in Embodiment 1 to quantify SOH irreversible fractional damage , aiming to establish a damage accumulation model that couples the core temperature and the charge amount ;

[0087] Obtain the charge amount in the current time step: by integrating the collected discharge current in the time step , the charge amount is calculated; is the absolute value of the charge amount flowing through the battery, which is the excitation factor causing SOH irreversible damage;

[0088] Calculate the SOH irreversible fractional damage : based on the estimated core temperature , the preset reference temperature , the charge amount and preset temperature sensitivity coefficient and basic capacity damage rate The irreversible fractional damage of SOH was calculated. ;

[0089] ;

[0090] in, It is a dimensionless loss of SOH fraction; It is a preset reference temperature; in this embodiment, the reference temperature is... Specifically refers to the ambient temperature at which the nominal capacity is specified in the battery cell manufacturer's specifications, such as 25°C;

[0091] and It was pre-calibrated through accelerated life testing; this calibration process was based on a set of experimental data. ;in, To maintain a constant temperature SOH loss measured when the load reaches a specific fraction of capacity loss. This corresponds to the cumulative charge amount; and The model was fitted using nonlinear least squares regression analysis on the experimental dataset to ensure optimal fit on the calibration dataset.

[0092] The term is the temperature acceleration factor of damage, which accurately simulates the exponential amplification effect of low temperature on irreversible damage to the battery cell.

[0093] This quantification process innovatively establishes the core temperature. Long-term damage from SOH The coupled model accurately captures the nonlinear, exponential amplification effect of extreme low temperatures on battery SOH damage by introducing an exponential temperature acceleration factor, providing a high-precision and high-reliability long-term health status benchmark for subsequent life protection decisions.

[0094] Example 4:

[0095] The process of revising the dynamic available capacity baseline includes:

[0096] Based on the estimated core temperature, a pre-defined lookup table or fitting function is queried to obtain the low-temperature capacity reduction factor;

[0097] The dynamic available capacity baseline is obtained by multiplying the nominal capacity, the current SOH fraction, and the low-temperature capacity reduction factor.

[0098] This embodiment modifies the dynamic available capacity benchmark of the joint estimation unit in Embodiment 1. Specific details;

[0099] Obtain the low-temperature capacity reduction factor Based on estimated core temperature Query the preset lookup table or fitting function to obtain the low-temperature capacity reduction factor. ; It is dimensionless, referring to the core temperature. The instantaneous discharge capacity of a battery cell is the fractional ratio of its nominal capacity to its instantaneous discharge capacity at different constant temperatures. Measuring the instantaneous discharge capacity of the battery cell Capacity relative to standard temperature And obtain, that is This is an inherent characteristic of system performance;

[0100] Calculate the dynamic available capacity baseline : nominal capacity Current SOH score Low-temperature capacity reduction factor Multiply to obtain the dynamic available capacity baseline. ;

[0101] ;

[0102] in, Simultaneously represented as a score of long-term health status and represent transient reversible damage at low temperatures. Factor correction;

[0103] This correction process dynamically adjusts the denominator of the SOC estimate by coupling long-term and instantaneous factors. This solves the problem of instantaneous capacity loss caused by neglecting low temperatures in traditional SOC estimation, and greatly improves the current SOC score. The accuracy of the estimation ensures that the power displayed by the BMS at extremely low temperatures is the actual usable power.

[0104] Example 5:

[0105] The calculation process for the electricity risk index includes:

[0106] The ratio is obtained by dividing the difference between the preset SOC safety boundary and the power lock boundary by the difference between the current SOC score and the power lock boundary.

[0107] The ratio is then used to perform an exponential calculation based on a preset SOC risk sensitivity coefficient to obtain the electricity risk index.

[0108] Electricity Risk Index Characterizes the degree of danger the system is from the power lock boundary; the difference between the preset SOC safety boundary and the power lock boundary is divided by the difference between the current SOC score and the power lock boundary to obtain the ratio;

[0109] The ratio is then used to calculate the exponent of the preset SOC risk sensitivity coefficient to obtain the electricity risk index;

[0110] Calculate the ratio: The preset SOC safety boundary... With respect to the preset power lock boundary The difference is set as the numerator; the current SOC score is... With power lock boundary The difference is set as the denominator; the ratio is obtained by dividing the numerator by the denominator.

[0111] Exponential calculation: and then apply a preset SOC risk sensitivity coefficient to the ratio. The electricity risk index is obtained through index calculation. ;

[0112] ;

[0113] in, It is a dimensionless coefficient greater than 1; The value is optimized offline, so that... exist near To meet specific risk response time targets in a timely manner, ensuring that the decision-making system intervenes in advance; It is the preset SOC safety boundary, representing the power level at which the system should begin to avoid risks;

[0114] Through index The calculation ensures that the current SOC score is... Approaching the battery lockout boundary At that time, the power risk index It exhibits non-linear and rapid growth, forming a strong risk warning signal to prevent sudden system failure under extremely low power conditions.

[0115] Example 6:

[0116] The calculation process for the temperature risk index includes:

[0117] The difference between the preset temperature safety boundary and the preset temperature lockout boundary is set as a molecule;

[0118] The difference between the estimated core temperature and the temperature lockout boundary is set as the denominator;

[0119] Divide the numerator by the denominator to obtain the ratio, and then perform an exponential calculation on the ratio using a preset temperature risk sensitivity coefficient to obtain the temperature risk index.

[0120] Temperature risk index Characterizes the degree of danger the system faces from the temperature lockout boundary;

[0121] Setting molecules: setting the preset temperature safety boundary With respect to the preset temperature lock-in boundary The difference is set as the molecule. ; It is the lowest temperature threshold at which the battery cell can operate safely. It is the upper boundary temperature at which the performance of the battery cell begins to decline significantly. Both are usually set based on the physical characteristics and safety limits of the battery cell.

[0122] Define the denominator: the estimated core temperature. Temperature lockout boundary The difference is set as the denominator. ;

[0123] Calculation and exponential operation: Divide the numerator by the denominator to obtain the ratio, and then apply the ratio to a preset temperature risk sensitivity coefficient. The temperature risk index is obtained through index calculation. ;

[0124] ;

[0125] in, It is a dimensionless coefficient, and its value is obtained through offline optimization. exist near The specific risk response time objective must be met.

[0126] This risk model achieves core temperature. Temperature risk The exponential effect; when estimating core temperature Approaching the temperature lockout boundary When, the denominator Approaching zero, leading to temperature risks. Rapid dispersion accurately simulates the exponential suppression effect of low temperature on the cell's discharge capability, forcing the system to take protective measures before the critical temperature point.

[0127] Example 7:

[0128] The calculation process for the model collapse risk index includes:

[0129] The model collapse risk index is obtained by multiplying the temperature data contamination marker with the preset contamination risk factor.

[0130] Model crash risk index Characterizes the confidence level of the BMS algorithm in the data input;

[0131] Calculate the risk index of model collapse : Contamination of temperature data With preset pollution risk factors Multiplying them together yields the model collapse risk index. ;

[0132] ;

[0133] in, It is a dimensionless, pre-defined high-value constant; The value is set high enough to ensure Greater than the weighted sum of all other risk items, i.e., when hour, Becoming a comprehensive risk The primary factor forces the system to enter the highest level of alert.

[0134] The introduction of this index will contaminate temperature data. Integrating it into the decision-making system solves the problem that traditional BMS algorithms cannot detect a decline in their own estimation confidence; through high-value factors This ensures that once data corruption is detected, the system will immediately implement risk hedging strategies, thus guaranteeing the robustness of the decision-making system.

[0135] Example 8:

[0136] The quantitative safety distance index is obtained by solving the inverse of the comprehensive risk index;

[0137] The rate of change of safe distance is calculated by quantifying the safe distance index using the first difference over a preset time step.

[0138] This embodiment calculates the quantified safety distance index using the dynamic risk quantification unit in Embodiment 1. and its rate of change Specific details;

[0139] Quantitative safety distance index Solution:

[0140] Quantitative safety distance index Through comprehensive risk index The reciprocal is obtained by solving;

[0141] ;

[0142] in, yes The reciprocal of quantifies the safety margin of the system from the multidimensional failure boundary;

[0143] Rate of change of safe distance Calculation:

[0144] Rate of change of safe distance By quantifying the safe distance index At the preset time step The first-order difference within is calculated;

[0145] ;

[0146] in, The preset time step should be matched with the decision cycle of the BMS control unit; for example, it can be set to 100 milliseconds. Characterizing safety margin The dynamic trend of change is a key input for making risk hedging decisions;

[0147] By integrating risk Converted into quantified safe distance This provides a clear concept of margin for decision-making; rate of change of safety distance. The introduction of this technology enables the decision-making system to sense whether danger is rapidly approaching, thereby allowing it to intervene in advance and implement emergency hedging strategies, avoiding a passive response mode.

[0148] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A low pressure energy storage system at extreme low temperatures, characterized in that, The method comprises the following steps: A multi-dimensional state perception unit is used to collect a raw data set and calculate a temperature gradient index; The temperature gradient index is compared with a preset normal temperature gradient threshold to obtain a temperature data pollution flag; and the estimated core temperature is calculated based on the temperature data pollution flag, the mean value of the raw data set, and the minimum reading. A joint estimation unit is used to quantify the irreversible fractional damage of SOH based on the estimated core temperature and the collected discharge current, update the current SOH fraction, and correct the dynamic available capacity reference based on the current SOH fraction and the estimated core temperature. The current SOC fraction is estimated based on the dynamic available capacity reference. A dynamic risk quantification unit is used to generate an electric quantity risk index based on the current SOC fraction. A temperature risk index is generated based on the estimated core temperature. A model collapse risk index is generated based on the temperature data pollution flag. The electric quantity risk index, the temperature risk index, and the model collapse risk index are combined and weighted to obtain a comprehensive risk index. The quantitative safety distance index is calculated based on the comprehensive risk index. The safety distance change rate of the quantitative safety distance index is calculated. An adaptive decision unit is used to output a start self-heating decision signal when the model collapse risk index is greater than zero and the estimated core temperature is less than a preset temperature upper limit. Otherwise, a discharge rejection decision signal is output when the quantitative safety distance index is less than a preset first safety threshold or the safety distance change rate is less than a preset first change rate threshold. Otherwise, a discharge permission decision signal is output.

2. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The calculation process of the estimated core temperature comprises: When the temperature data pollution flag is zero, the mean value of the raw data set is set as the estimated core temperature. When the temperature data pollution flag is one, the minimum reading of the raw data set is set as the estimated core temperature.

3. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The quantitative process of the irreversible fractional damage of SOH comprises: The charge quantity in the current time step is obtained. The irreversible fractional damage of SOH is calculated based on the estimated core temperature, a preset reference temperature, the charge quantity, and a preset temperature sensitivity coefficient and a basic capacity damage rate.

4. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The correction process of the dynamic available capacity reference comprises: The low-temperature capacity reduction factor is obtained by querying a preset lookup table or a fitting function based on the estimated core temperature. The dynamic available capacity reference is obtained by multiplying the nominal capacity, the current SOH fraction, and the low-temperature capacity reduction factor.

5. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The calculation process of the electric quantity risk index comprises: The difference between the preset SOC safety boundary and the electric quantity locking boundary is divided by the difference between the current SOC fraction and the electric quantity locking boundary to obtain a ratio. The ratio is subjected to exponential operation of a preset SOC risk sensitivity coefficient to obtain the electric quantity risk index.

6. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The calculation process of the temperature risk index comprises: The difference between the preset temperature safety boundary and the preset temperature locking boundary is set as the numerator. The difference between the estimated core temperature and the temperature locking boundary is set as the denominator. The ratio obtained by dividing the numerator by the denominator is subjected to exponential operation of a preset temperature risk sensitivity coefficient to obtain the temperature risk index.

7. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The calculation process of the model collapse risk index comprises: The model collapse risk index is obtained by multiplying the temperature data pollution flag and a preset pollution risk factor.

8. The low pressure energy storage system at extreme low temperature of claim 1, wherein, The quantitative safety distance index is calculated by reciprocal of the comprehensive risk index; The safety distance change rate is calculated by first-order difference of the quantitative safety distance index in preset time steps.

Images