Secondary battery energy-saving carbon-reducing adaptive charging regulation method and system

Abstract

The application discloses a kind of secondary battery energy-saving carbon-reducing adaptive charging regulation methods and systems, it is related to battery charging control technical field, comprising: S1, low-carbon charging environment construction;S2, carbon cost quantification;S3, digital twin prediction;S4, multi-objective dynamic weight optimization;S5, closed-loop learning correction.The application unifies charging loss, battery aging loss and heat dissipation energy consumption into integrated carbon cost, so that the optimization goal of charging strategy changes from the traditional minimum charging time or maximum charging efficiency to the minimum life cycle carbon emission.Different from the existing scheme that only converts carbon emission according to power consumption, the application includes real-time optimization framework the production end carbon emission brought by battery aging, solves the technical contradiction that "energy saving in charging process but battery premature scrapping increases carbon", realizes the leap of charging regulation from single-point energy saving to full-chain carbon reduction.

Description

Technical Field

[0001] This invention relates to the field of battery charging control technology, specifically to an energy-saving and carbon-reducing adaptive charging regulation method and system for secondary batteries. Background Technology

[0002] Charging control technology for secondary batteries (especially lithium-ion batteries) is a core component of battery management systems, directly impacting battery safety, cycle life, and energy efficiency. Traditional charging control methods primarily use battery state of charge (SOC) and terminal voltage as feedback parameters, employing strategies such as constant current-constant voltage (CC-CV), multi-stage constant current, or pulse charging to ensure charging safety and maximize cycle life. In recent years, with the large-scale application of electric vehicles and energy storage systems, research on charging strategies has gradually incorporated parameters such as battery state of health (SOH), internal resistance characteristics, and electrochemical models. Adaptive charging methods based on model predictive control, dynamic programming, and intelligent optimization algorithms have emerged, capable of adjusting the charging current at different aging stages to mitigate capacity decay. Simultaneously, some research has begun to focus on energy loss and temperature rise constraints during charging, incorporating charging efficiency into the optimization objective, forming a basic framework for multi-objective charging control.

[0003] However, existing charging control methods only assess carbon emissions at the level of converting charging energy consumption into carbon emissions. This means calculating carbon emissions by multiplying the energy consumed during charging by the grid's average carbon emission factor, without incorporating charging losses, battery aging losses, and heat dissipation energy consumption as a whole into the carbon cost accounting system. This single-point conversion method severs the inherent physical connection between "charging energy consumption—charging losses—battery aging—thermal management energy consumption," resulting in optimization targets only reflecting carbon emissions from charging, while ignoring the additional carbon emissions from manufacturing caused by accelerated battery aging due to improper charging strategies, and the carbon emissions from heat dissipation systems used to suppress temperature rise. Since battery production itself has a high carbon emission intensity, if charging strategies unilaterally pursue minimizing energy consumption while disregarding battery lifespan losses, the contradictory result of "carbon saving during charging but premature battery failure and increased carbon emissions" may occur.

[0004] In response to this problem, this application proposes an energy-saving and carbon-reducing adaptive charging control method and system for secondary batteries to solve the aforementioned issues. Summary of the Invention

[0005] The purpose of this invention is to provide an energy-saving and carbon-reducing adaptive charging control method and system for secondary batteries, in order to solve the problem that the battery production process itself has a high carbon emission intensity, and if the charging strategy unilaterally pursues the minimization of power consumption without considering battery life loss, it may lead to the contradictory result of "carbon saving during the charging stage and premature battery scrapping and carbon increase".

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] In a first aspect, this application provides an energy-saving and carbon-reducing adaptive charging control method for secondary batteries, comprising:

[0008] By acquiring real-time grid carbon emission factor data and combining it with battery charging state parameters, a low-carbon charging environment model characterizing the carbon emission intensity under the current charging environment is constructed.

[0009] Based on the aforementioned low-carbon charging environment model, the charging loss, battery aging loss, and heat dissipation energy consumption during the battery charging process are uniformly quantified into a comprehensive carbon cost, resulting in a carbon cost quantification value.

[0010] The carbon cost quantification value is input into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain a future state trend prediction sequence.

[0011] Based on the future state trend prediction sequence, multi-objective dynamic weight adjustment is performed among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target to generate the optimal charging current curve.

[0012] Charging is performed according to the optimal charging current curve, and actual charging process data is collected after charging is completed. The control strategy for the next charging is then corrected online based on the actual charging process data.

[0013] Furthermore, the step of acquiring real-time grid carbon emission factor data and combining it with battery charging state parameters to construct a low-carbon charging environment model characterizing the carbon emission intensity under the current charging environment includes:

[0014] Based on power grid area division and real-time power dispatch information, the real-time carbon emission factor of the power grid area where the current charging time is located is dynamically obtained.

[0015] Simultaneously, the battery's state of charge, state of health, and temperature are collected as the battery's charging state parameters;

[0016] By integrating the real-time carbon emission factor with the battery charging state parameters, a low-carbon charging environment model is constructed, with the charging start time and current battery state as inputs and the indirect carbon emission intensity corresponding to a unit of charging amount as output.

[0017] Furthermore, the process of fusing the real-time carbon emission factor with the battery charging state parameters to construct a low-carbon charging environment model, with the charging start time and current battery state as inputs and the indirect carbon emission intensity corresponding to a unit of charging amount as output, includes:

[0018] A predicted sequence of carbon emission factors for the power grid within a preset future time period is obtained. Combined with battery charging state parameters, a predictive extension form of the low-carbon charging environment model is constructed. This predictive extension form is used to output a sequence of estimated carbon emission intensity at multiple future times during the initial charging stage, providing a time-domain matched carbon emission benchmark for subsequent carbon cost quantification.

[0019] Furthermore, based on the aforementioned low-carbon charging environment model, the charging loss, battery aging loss, and heat dissipation energy consumption during the battery charging process are uniformly quantified into a comprehensive carbon cost, resulting in a quantified carbon cost value, including:

[0020] The carbon emissions converted from charging loss, aging loss, and heat dissipation energy consumption are calculated separately, and the three are weighted and summed to obtain the comprehensive carbon cost.

[0021] The carbon emissions converted from charging losses are determined by the charging conversion efficiency and the carbon emission intensity output by the low-carbon charging environment model. The carbon emissions converted from aging losses are determined by the cumulative impact of the charging current sequence on battery life decay. The carbon emissions converted from heat dissipation energy consumption are determined by the mapping relationship between the predicted temperature rise and the power of the heat dissipation system.

[0022] Furthermore, the overall carbon cost is determined in the following manner:

[0023] Define charging current sequence The carbon emission Closs, calculated from charging loss, is expressed as follows:

[0024]

[0025] in, This refers to the real-time power loss during charging. For charging conversion efficiency, Let be the carbon emission factor of the power grid at time t;

[0026] Carbon emissions converted from aging losses Represented as:

[0027]

[0028] in, This is a function for predicting the amount of battery life degradation based on a battery aging model.

[0029] Carbon emissions from heat dissipation energy consumption Represented as:

[0030]

[0031] in, This is a mapping function between heat dissipation power and temperature difference. To predict temperature, For ambient reference temperature;

[0032] Overall carbon cost:

[0033]

[0034] Wherein, α, β, and γ are the conversion coefficients that uniformly convert energy loss, lifespan decay, and heat dissipation energy consumption to carbon emission equivalents.

[0035] Furthermore, the step of inputting the quantified carbon cost value into a pre-built digital twin prediction model to predict the battery's temperature rise trend, polarization trend, and aging trend during future charging processes, thereby obtaining a future state trend prediction sequence, includes:

[0036] The digital twin prediction model is constructed based on the electrochemical equivalent circuit model and the thermal coupling model. Using the quantified carbon cost value and the current battery state as initial boundary conditions, the model prediction control framework iteratively extrapolates the battery internal temperature, polarization voltage increment and cumulative capacity decay rate at each moment within the future charging window, forming the future state trend prediction sequence indexed by time.

[0037] Furthermore, during the iterative deduction process, after each prediction step, the digital twin prediction model uses the prediction state of the current step as the input state for the next step to perform rolling updates. After completing a preset number of prediction steps, it determines whether the predicted temperature exceeds the safety threshold or whether the predicted polarization voltage exceeds the polarization limit. If they do, a safety constraint correction is triggered, and the corrected boundary conditions are re-substituted into the digital twin prediction model to continue the prediction deduction until the prediction of the entire charging window is completed, and a future state trend prediction sequence that satisfies the safety constraints is output.

[0038] Furthermore, the step of dynamically adjusting the weights of carbon emission reduction targets, lifespan maintenance targets, charging efficiency targets, and user demand targets based on the predicted future state trend sequence to generate an optimal charging current curve includes:

[0039] Based on the predicted future state trend sequence, normalized evaluation indices for each objective under the current charging strategy are calculated.

[0040] Based on the normalized evaluation index and corresponding dynamic weight coefficient of each objective, a comprehensive cost function is constructed, and the optimal charging current sequence is searched within the current constraint range with the minimization of the comprehensive cost function as the optimization objective.

[0041] The dynamic weighting coefficient is adjusted in real time based on the real-time carbon cost quantification value, the current health status of the battery, and the charging mode selected by the user.

[0042] Furthermore, the dynamic weighting coefficient is adjusted in real time based on the real-time carbon cost quantification value, the current battery health status, and the charging mode selected by the user, including:

[0043] When the real-time carbon cost quantification value is higher than the preset carbon cost threshold, the weight coefficient of the carbon emission reduction target will be automatically increased and the weight coefficient of the charging efficiency target will be reduced accordingly.

[0044] When the battery's current health status is lower than a preset health threshold, the weighting coefficient of the lifespan maintenance target is automatically increased;

[0045] The charging modes selected by the user include fast charging mode, long life mode and low carbon mode. The initial ratio of each dynamic weight coefficient is different for each mode. During the charging process, each dynamic weight coefficient is adjusted based on the initial ratio according to the real-time carbon cost quantification value and the current health status of the battery.

[0046] Secondly, this application provides a secondary battery energy-saving and carbon-reducing adaptive charging control system, comprising:

[0047] The low-carbon charging environment construction module is used to acquire real-time grid carbon emission factor data and, in combination with battery charging state parameters, construct a low-carbon charging environment model that characterizes the carbon emission intensity under the current charging environment.

[0048] The carbon cost quantification module is used to quantify the charging loss, battery aging loss and heat dissipation energy consumption during the battery charging process into a comprehensive carbon cost based on the low-carbon charging environment model, and obtain the carbon cost quantification value.

[0049] The digital twin prediction module is used to input the quantified carbon cost value into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain a future state trend prediction sequence.

[0050] The multi-objective dynamic weight optimization module is used to perform multi-objective dynamic weight adjustment among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target based on the future state trend prediction sequence, and generate the optimal charging current curve.

[0051] The closed-loop learning correction module is used to perform charging according to the optimal charging current curve, and to collect actual charging process data after charging is completed, and to make online corrections to the control strategy for the next charging based on the actual charging process data.

[0052] Compared with existing technologies, this invention provides an energy-saving and carbon-reducing adaptive charging control method and system for secondary batteries. This invention quantifies charging losses, battery aging losses, and heat dissipation energy consumption into a unified comprehensive carbon cost, shifting the optimization objective of charging strategies from minimizing charging time or maximizing charging efficiency to minimizing carbon emissions throughout the entire lifecycle. Unlike existing solutions that only calculate carbon emissions based on electricity consumption, this invention incorporates the allocation of carbon emissions from battery aging at the production end into the real-time optimization framework, resolving the technical contradiction of "energy saving during charging but premature battery failure leading to increased carbon emissions," and achieving a leap from single-point energy saving to full-chain carbon reduction in charging control.

[0053] This invention employs a closed-loop learning mechanism after charging to online correct the parameters of the digital twin prediction model based on feedback signals such as actual energy consumption, temperature rise, and capacity recovery, ensuring that the prediction model always closely reflects the current aging state of the battery. Unlike existing technologies that use open-loop prediction models and whose deviations gradually increase with battery aging, this invention enables the charging strategy to have adaptive adjustment capabilities throughout the battery's entire lifespan, maintaining control accuracy without the need for periodic offline calibration. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0055] Figure 1 A flowchart of an energy-saving and carbon-reducing adaptive charging control method for secondary batteries provided in an embodiment of the present invention;

[0056] Figure 2 This is a block diagram of a secondary battery energy-saving and carbon-reducing adaptive charging control system provided in an embodiment of the present invention. Detailed Implementation

[0057] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0058] As attached Figure 1 As shown:

[0059] Example 1:

[0060] An energy-saving and carbon-reducing adaptive charging control method for secondary batteries includes:

[0061] S1. Obtain real-time grid carbon emission factor data and, in conjunction with battery charging state parameters, construct a low-carbon charging environment model that characterizes the carbon emission intensity under the current charging environment.

[0062] In step S1, based on the grid area division and real-time power dispatch information, the real-time carbon emission factor of the grid area where the current charging time is located is dynamically obtained;

[0063] Simultaneously, the battery's state of charge, state of health, and temperature are collected as the battery's charging state parameters;

[0064] By fusing the real-time carbon emission factor with the battery charging state parameters, a low-carbon charging environment model is constructed with the charging start time and current battery state as inputs and the indirect carbon emission intensity corresponding to a unit of charging amount as output.

[0065] Specifically, the predicted sequence of carbon emission factors for the power grid within a preset future time period is obtained, and combined with the battery charging state parameters, a predictive extension form of the low-carbon charging environment model is constructed. The predictive extension form is used to output the estimated carbon emission intensity sequence for multiple future times at the beginning of the charging stage, providing a time-domain matching carbon emission benchmark for subsequent carbon cost quantification.

[0066] Furthermore, the grid carbon emission factor is a quantitative indicator of carbon emissions per kilowatt-hour of electricity production and consumption. Its value is directly determined by the power structure; a higher proportion of coal-fired power results in a larger carbon emission factor, while an increase in the proportion of clean energy such as wind and solar power leads to a lower carbon emission factor. Traditional carbon emission accounting generally uses the annual provincial average electricity carbon factor, which is difficult to accurately reflect the differences in carbon emissions across different regions, time periods, and users. Therefore, this step dynamically obtains the real-time carbon emission factor of the grid region where the current charging time is located, based on grid area division and real-time power dispatch information.

[0067] In practice, relying on the electricity carbon metering system, the electricity carbon emission factor is calculated in milliseconds and updated in minutes by installing electricity carbon meters at each node of the power system's source, grid, and load chain.

[0068] The carbon meter isolates the energy metering and carbon metering functions through a logic layer, combining the two into one.

[0069] Based on the power grid topology, real-time node power data is collected, and intelligent algorithms are deployed to generate high-precision dynamic carbon emission factors by combining dynamic factors such as fluctuations in renewable energy power generation. Simultaneously, the battery's state of charge (SOC), state of health (SOH), and temperature are collected as battery charging status parameters.

[0070] Based on this, the predicted sequence of carbon emission factors for the power grid in the future preset time period can be obtained. The carbon emission factor sequence of the power system at continuous time nodes can be decomposed into various feature components such as trend characteristics, periodic characteristics and typical event characteristics through time series decomposition method. Then, the predicted carbon emission factor values ​​for each future time node can be output by the prediction model.

[0071] By fusing real-time carbon emission factors, predicted carbon emission factor sequences, and battery charging state parameters, a low-carbon charging environment model is constructed. This model takes the charging start time and current battery state as input and the indirect carbon emission intensity corresponding to a unit of charging amount as output. The predictive extension of this model can output an estimated carbon emission intensity sequence for multiple future times during the initial charging stage, providing a time-domain matched carbon emission benchmark for subsequent carbon cost quantification.

[0072] S2. Based on the low-carbon charging environment model, the charging loss, battery aging loss and heat dissipation energy consumption during the battery charging process are uniformly quantified into a comprehensive carbon cost to obtain a carbon cost quantification value.

[0073] In step S2, the carbon emissions converted from charging loss, carbon emissions converted from aging loss, and carbon emissions converted from heat dissipation energy consumption are calculated respectively, and the three are weighted and summed to obtain the comprehensive carbon cost.

[0074] Wherein, the carbon emission amount converted from charging loss is determined by the charging conversion efficiency and the carbon emission intensity output by the low-carbon charging environment model; the carbon emission amount converted from aging loss is determined by the cumulative impact of the charging current sequence on battery life decay; and the carbon emission amount converted from heat dissipation energy consumption is determined by the mapping relationship between the predicted temperature rise and the power of the heat dissipation system.

[0075] Specifically, traditional battery charging carbon emission accounting only considers the direct carbon emissions from the electricity consumed from the grid during charging, ignoring the indirect carbon emissions from battery aging, charging losses, and heat dissipation. This step quantifies these three types of losses into a unified comprehensive carbon cost.

[0076] Specifically, the carbon emissions converted from the three types of losses are calculated separately, and the three are weighted and summed to obtain the comprehensive carbon cost.

[0077] The carbon emissions converted from charging losses are determined by the charging conversion efficiency and the carbon emission intensity output by the low-carbon charging environment model:

[0078] Define charging current sequence Below, the real-time power loss during charging is The charging conversion efficiency is The carbon emission factor of the power grid at time t is Then, the carbon emissions are converted from charging losses. Represented as:

[0079]

[0080] Where α is the conversion factor for converting energy loss to carbon emission equivalent. This integral is performed over the entire charging time window, accumulating the carbon emissions corresponding to the additional electrical energy consumption caused by insufficient charging efficiency at each moment.

[0081] The carbon emissions calculated from aging losses are determined by the cumulative effect of the charging current sequence on battery life degradation. Let... , , Here is the lifespan degradation prediction function based on the battery aging model, where In a charged state, For battery temperature, the carbon emissions are calculated based on aging losses. Represented as:

[0082]

[0083] Wherein, β is the conversion factor for converting lifetime decay to carbon emission equivalent.

[0084] The physical meaning of this function is:

[0085] The irreversible capacity decay that occurs during each battery charge is equivalent to shortening the battery's entire lifespan, thereby accelerating the carbon emission amortization during battery production, transportation, and recycling. Therefore, this aging loss needs to be quantified as carbon cost.

[0086] The carbon emission equivalent of heat dissipation energy consumption is determined by the mapping relationship between the predicted temperature rise and the power of the heat dissipation system. Let... Let be the mapping function between heat dissipation power and temperature difference, where To predict temperature, Using the ambient reference temperature, the carbon emissions are calculated based on the energy consumption for heat dissipation. Represented as:

[0087]

[0088] Wherein, γ is the conversion factor for converting heat dissipation energy consumption to carbon emission equivalent.

[0089] Overall carbon cost The sum of the three:

[0090]

[0091] This comprehensive carbon cost quantification serves as an input benchmark for subsequent digital twin predictions and multi-objective optimization, elevating the optimization objective of the charging strategy from "minimizing single energy consumption" to "minimizing carbon costs throughout the entire lifecycle."

[0092] S3. Input the quantified carbon cost value into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain the future state trend prediction sequence.

[0093] In step S3, the digital twin prediction model is constructed based on the electrochemical equivalent circuit model and the thermal coupling model. The carbon cost quantification value and the current battery state are used as the initial boundary conditions. The model prediction control framework is used to iteratively deduce the battery internal temperature, polarization voltage increment and cumulative capacity decay rate at each moment in the future charging window, forming the future state trend prediction sequence indexed by time.

[0094] Specifically, during the iterative deduction process, after each prediction step, the prediction state of the current step is used as the input state for the next step and the model is updated in a rolling manner. After each preset number of prediction steps are completed, it is determined whether the predicted temperature exceeds the safety threshold or whether the predicted polarization voltage exceeds the polarization limit. If they do, the safety constraint correction is triggered, and the corrected boundary conditions are re-substituted into the digital twin prediction model to continue the prediction deduction until the prediction of the entire charging window is completed, and the future state trend prediction sequence that meets the safety constraints is output.

[0095] Furthermore, the digital twin prediction model is constructed based on the electrochemical equivalent circuit model and the thermal coupling model.

[0096] The electrochemical equivalent circuit model adopts the Thevenin equivalent circuit architecture and simulates the dynamic characteristics of ohmic polarization and electrochemical polarization of the battery through an RC network.

[0097] The thermal coupling model describes the competitive relationship between internal heat generation and external heat dissipation within the battery. Internal heat generation includes ohmic heat, reaction heat, and polarization heat, while external heat dissipation is represented by convection, conduction, or radiation boundary terms. Both models achieve thermoelectric coupling through the influence of temperature on the battery's internal resistance and polarization parameters.

[0098] Using the quantified carbon cost and the current battery state (including SOC, SOH, temperature, internal resistance, etc.) as initial boundary conditions, a model predictive control framework is used to iteratively deduce the battery's internal temperature, polarization voltage increment, and cumulative capacity decay rate at each moment within the future charging window, forming a future state trend prediction sequence indexed by time.

[0099] During the iterative simulation, after each prediction step, the prediction state of the current step is used as the input state for the next step and the simulation is updated in a rolling manner.

[0100] This rolling update mechanism enables the model to continuously use the latest prediction results to correct the direction of subsequent inferences, avoiding the accumulation of errors caused by open-loop prediction.

[0101] After each preset number of prediction steps are completed, it is determined whether the predicted temperature exceeds the safety threshold or the predicted polarization voltage exceeds the polarization limit. If they do, a safety constraint correction is triggered, and the corrected boundary conditions are re-substituted into the digital twin prediction model to continue the prediction simulation.

[0102] This safety constraint correction mechanism ensures that the predicted sequence always stays within the battery's safe operating range until the prediction of the entire charging window is completed, and finally outputs a future state trend prediction sequence that meets the safety constraints.

[0103] Temperature prediction models based on digital twins can predict temperature rises several minutes in advance, with prediction errors kept within a low range.

[0104] S4. Based on the future state trend prediction sequence, perform multi-objective dynamic weight adjustment among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target to generate the optimal charging current curve.

[0105] In step S4, based on the future state trend prediction sequence, the normalized evaluation index of each objective under the current charging strategy is calculated respectively.

[0106] Based on the normalized evaluation index and corresponding dynamic weight coefficient of each objective, a comprehensive cost function is constructed, and the optimal charging current sequence is searched within the current constraint range with the minimization of the comprehensive cost function as the optimization objective.

[0107] The dynamic weighting coefficient is adjusted in real time based on the real-time carbon cost quantification value, the current health status of the battery, and the charging mode selected by the user.

[0108] Specifically, when the real-time carbon cost quantification value is higher than the preset carbon cost threshold, the weighting coefficient of the carbon emission reduction target will be automatically increased and the weighting coefficient of the charging efficiency target will be reduced accordingly.

[0109] When the battery's current health status is lower than a preset health threshold, the weighting coefficient of the lifespan maintenance target is automatically increased;

[0110] The charging modes selected by the user include fast charging mode, long life mode and low carbon mode. The initial ratio of each dynamic weight coefficient is different for each mode. During the charging process, each dynamic weight coefficient is adjusted based on the initial ratio according to the real-time carbon cost quantification value and the current health status of the battery.

[0111] Furthermore, the core of this step lies in unifying multiple often conflicting optimization objectives (carbon emission reduction requires low current to reduce losses and temperature rise, charging efficiency requires high current to shorten time, and lifespan maintenance requires moderate current to avoid accelerated aging) into an optimization framework through a dynamic weighting mechanism.

[0112] Specifically, based on the future state trend prediction sequence, normalized evaluation indicators for each objective under the current charging strategy are calculated.

[0113] The evaluation indicators for carbon emission reduction targets are determined based on the comprehensive carbon cost quantification value;

[0114] The evaluation metrics for lifetime maintenance targets are determined based on the predicted cumulative capacity decay rate;

[0115] The evaluation index for charging efficiency targets is determined based on the predicted charging completion time;

[0116] The evaluation metrics for user demand objectives are determined based on the expected charging time window corresponding to the user's selected charging mode (fast charging mode, long life mode, or low carbon mode).

[0117] Based on the normalized evaluation indexes and corresponding dynamic weight coefficients of each objective, a comprehensive cost function is constructed. The optimization objective is to minimize this comprehensive cost function and search for the optimal charging current sequence within the current constraint range (including the maximum allowable charging current, the minimum charging current, the current change rate limit, etc.).

[0118] The dynamic weighting coefficient adjustment mechanism is as follows: When the real-time carbon cost quantification value is higher than the preset carbon cost threshold, the weighting coefficient of the carbon emission reduction target is automatically increased and the weighting coefficient of the charging efficiency target is correspondingly decreased. The logic behind this adjustment is as follows:

[0119] High carbon costs mean that the carbon emission intensity of the power grid is high during the current charging period. At this time, charging power should be reduced first to reduce high carbon electricity consumption.

[0120] When the battery's current health status is below a preset health threshold, the weighting coefficient of the lifespan maintenance target is automatically increased—at this point, the battery is nearing the end of its lifespan, and overcharging will accelerate its failure and increase carbon emissions from battery replacement.

[0121] Different charging modes correspond to different initial ratios for each group of dynamic weight coefficients. During the charging process, each dynamic weight coefficient is adjusted based on the initial ratio according to the real-time carbon cost quantification value and the current health status of the battery.

[0122] S5. Perform charging according to the optimal charging current curve, and collect actual charging process data after charging is completed. Based on the actual charging process data, make online corrections to the control strategy for the next charging.

[0123] In step S5, the actual current, actual voltage, actual temperature, and actual energy consumption at each moment during the charging process are collected, and the actual capacity recovery value of the battery is measured after charging is completed; the actual current, actual voltage, actual temperature, and actual energy consumption are compared with the corresponding predicted values ​​in the future state trend prediction sequence, and the prediction deviation is calculated.

[0124] The prediction deviation and the actual capacity recovery value are used as feedback signals to update the internal parameters of the digital twin prediction model, and the updated digital twin prediction model is used as the prediction benchmark for the next charging regulation.

[0125] Furthermore, step S5 implements the closed-loop self-learning capability of the charging strategy. During the charging process according to the optimal charging current curve, the actual current, actual voltage, actual temperature, and actual energy consumption at each moment are collected, and the actual capacity recovery value of the battery is measured after charging is completed.

[0126] The actual measured values ​​are compared with the corresponding predicted values ​​in the future state trend prediction sequence output by the digital twin prediction model in step three, and the prediction deviations are calculated, including current tracking deviation, voltage prediction deviation, temperature prediction deviation and energy consumption prediction deviation.

[0127] The prediction bias and actual capacity recovery value are used as feedback signals to update the internal parameters of the digital twin prediction model.

[0128] The specific update method is to use the prediction bias correction model to correct the equivalent circuit parameters (such as ohmic internal resistance, polarization resistance, polarization capacitance, etc.) and thermal model parameters (such as thermal capacity, thermal resistance, etc.) to make the model more consistent with the current actual state of the battery.

[0129] By using actual capacity recovery values ​​to correct the attenuation coefficient of the aging model, the aging trend prediction can be made more accurate.

[0130] The updated digital twin prediction model is used as the prediction benchmark for the next charging adjustment, thereby realizing a complete closed loop of "prediction-execution-feedback-correction" and ensuring that the charging strategy remains optimal throughout the battery's entire life cycle.

[0131] As can be seen from the above, this invention quantifies charging loss, battery aging loss, and heat dissipation energy consumption into a unified comprehensive carbon cost, shifting the optimization objective of charging strategies from the traditional minimizing charging time or maximizing charging efficiency to minimizing carbon emissions throughout the entire lifecycle. Unlike existing schemes that only calculate carbon emissions based on electricity consumption, this invention incorporates the allocation of carbon emissions from battery aging at the production end into the real-time optimization framework, resolving the technical contradiction of "energy saving during charging but premature battery failure leading to increased carbon emissions," and achieving a leap from single-point energy saving to full-chain carbon reduction in charging regulation.

[0132] like Figure 2 As shown, in one embodiment, this application provides a secondary battery energy-saving and carbon-reducing adaptive charging control system, comprising:

[0133] The low-carbon charging environment construction module is used to acquire real-time grid carbon emission factor data and, in combination with battery charging state parameters, construct a low-carbon charging environment model that characterizes the carbon emission intensity under the current charging environment.

[0134] The carbon cost quantification module is used to quantify the charging loss, battery aging loss and heat dissipation energy consumption during the battery charging process into a comprehensive carbon cost based on the low-carbon charging environment model, and obtain the carbon cost quantification value.

[0135] The digital twin prediction module is used to input the quantified carbon cost value into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain a future state trend prediction sequence.

[0136] The multi-objective dynamic weight optimization module is used to perform multi-objective dynamic weight adjustment among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target based on the future state trend prediction sequence, and generate the optimal charging current curve.

[0137] The closed-loop learning correction module is used to perform charging according to the optimal charging current curve, and to collect actual charging process data after charging is completed, and to make online corrections to the control strategy for the next charging based on the actual charging process data.

[0138] As can be seen from the above, this invention, through a closed-loop learning mechanism after charging, corrects the parameters of the digital twin prediction model online based on feedback signals such as actual energy consumption, temperature rise, and capacity recovery, ensuring that the prediction model always closely matches the current aging state of the battery. Unlike existing technologies where the prediction model is used in an open loop and the deviation gradually increases with battery aging, this invention enables the charging strategy to have adaptive adjustment capabilities throughout the entire battery lifecycle, maintaining control accuracy without the need for periodic offline calibration.

[0139] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A secondary battery energy-saving and carbon-reducing adaptive charging control method, characterized in that, include: By acquiring real-time grid carbon emission factor data and combining it with battery charging state parameters, a low-carbon charging environment model characterizing the carbon emission intensity under the current charging environment is constructed. Based on the aforementioned low-carbon charging environment model, the charging loss, battery aging loss, and heat dissipation energy consumption during the battery charging process are uniformly quantified into a comprehensive carbon cost, resulting in a carbon cost quantification value. The carbon cost quantification value is input into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain a future state trend prediction sequence. Based on the future state trend prediction sequence, multi-objective dynamic weight adjustment is performed among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target to generate the optimal charging current curve. Charging is performed according to the optimal charging current curve, and actual charging process data is collected after charging is completed. The control strategy for the next charging is then corrected online based on the actual charging process data.

2. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 1, characterized in that, The process of acquiring real-time grid carbon emission factor data and combining it with battery charging state parameters to construct a low-carbon charging environment model characterizing the carbon emission intensity under the current charging environment includes: Based on power grid area division and real-time power dispatch information, the real-time carbon emission factor of the power grid area where the current charging time is located is dynamically obtained. Simultaneously, the battery's state of charge, state of health, and temperature are collected as the battery's charging state parameters; By integrating the real-time carbon emission factor with the battery charging state parameters, a low-carbon charging environment model is constructed, with the charging start time and current battery state as inputs and the indirect carbon emission intensity corresponding to a unit of charging amount as output.

3. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 2, characterized in that, The process of fusing the real-time carbon emission factor with the battery charging state parameters to construct a low-carbon charging environment model, with the charging start time and current battery state as inputs and the indirect carbon emission intensity corresponding to a unit of charging amount as output, includes: A predicted sequence of carbon emission factors for the power grid within a preset future time period is obtained. Combined with battery charging state parameters, a predictive extension form of the low-carbon charging environment model is constructed. This predictive extension form is used to output a sequence of estimated carbon emission intensity at multiple future times during the initial charging stage, providing a time-domain matched carbon emission benchmark for subsequent carbon cost quantification.

4. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 1, characterized in that, Based on the aforementioned low-carbon charging environment model, the charging loss, battery aging loss, and heat dissipation energy consumption during the battery charging process are uniformly quantified into a comprehensive carbon cost, resulting in a quantified carbon cost value, including: The carbon emissions converted from charging loss, aging loss, and heat dissipation energy consumption are calculated separately, and the three are weighted and summed to obtain the comprehensive carbon cost. The carbon emissions converted from charging losses are determined by the charging conversion efficiency and the carbon emission intensity output by the low-carbon charging environment model. The carbon emissions converted from aging losses are determined by the cumulative impact of the charging current sequence on battery life decay. The carbon emissions converted from heat dissipation energy consumption are determined by the mapping relationship between the predicted temperature rise and the power of the heat dissipation system.

5. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 4, characterized in that, The overall carbon cost is determined in the following ways: Define charging current sequence The carbon emission Closs, calculated from charging loss, is expressed as follows: in, This refers to the real-time power loss during charging. For charging conversion efficiency, Let be the carbon emission factor of the power grid at time t; Carbon emissions converted from aging losses Represented as: in, This is a function for predicting the amount of battery life degradation based on a battery aging model. Carbon emissions from heat dissipation energy consumption Represented as: in, This is a mapping function between heat dissipation power and temperature difference. To predict temperature, For ambient reference temperature; Overall carbon cost: Wherein, α, β, and γ are the conversion coefficients that uniformly convert energy loss, lifespan decay, and heat dissipation energy consumption to carbon emission equivalents.

6. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 1, characterized in that, The step involves inputting the quantified carbon cost value into a pre-built digital twin prediction model to predict the battery's temperature rise trend, polarization trend, and aging trend during future charging processes, resulting in a future state trend prediction sequence, including: The digital twin prediction model is constructed based on the electrochemical equivalent circuit model and the thermal coupling model. Using the quantified carbon cost value and the current battery state as initial boundary conditions, the model prediction control framework iteratively extrapolates the battery internal temperature, polarization voltage increment and cumulative capacity decay rate at each moment within the future charging window, forming the future state trend prediction sequence indexed by time.

7. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 6, characterized in that, During the iterative simulation process, the digital twin prediction model updates itself by using the prediction state of the current step as the input state for the next step after each prediction step. After completing a preset number of prediction steps, it determines whether the predicted temperature exceeds the safety threshold or whether the predicted polarization voltage exceeds the polarization limit. If they do, a safety constraint correction is triggered, and the corrected boundary conditions are re-introduced into the digital twin prediction model to continue the prediction simulation until the prediction of the entire charging window is completed. Finally, a future state trend prediction sequence that satisfies the safety constraints is output.

8. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 1, characterized in that, The step of generating an optimal charging current curve by dynamically adjusting the weights of carbon emission reduction targets, lifespan maintenance targets, charging efficiency targets, and user demand targets based on the predicted future state trend sequence includes: Based on the predicted future state trend sequence, normalized evaluation indices for each objective under the current charging strategy are calculated. Based on the normalized evaluation index and corresponding dynamic weight coefficient of each objective, a comprehensive cost function is constructed, and the optimal charging current sequence is searched within the current constraint range with the minimization of the comprehensive cost function as the optimization objective. The dynamic weighting coefficient is adjusted in real time based on the real-time carbon cost quantification value, the current health status of the battery, and the charging mode selected by the user.

9. The energy-saving and carbon-reducing adaptive charging control method for secondary batteries according to claim 8, characterized in that, The dynamic weighting coefficient is adjusted in real time based on the real-time carbon cost quantification value, the current battery health status, and the user-selected charging mode, including: When the real-time carbon cost quantification value is higher than the preset carbon cost threshold, the weight coefficient of the carbon emission reduction target will be automatically increased and the weight coefficient of the charging efficiency target will be reduced accordingly. When the battery's current health status is lower than a preset health threshold, the weighting coefficient of the lifespan maintenance target is automatically increased; The charging modes selected by the user include fast charging mode, long life mode and low carbon mode. The initial ratio of each dynamic weight coefficient is different for each mode. During the charging process, each dynamic weight coefficient is adjusted based on the initial ratio according to the real-time carbon cost quantification value and the current health status of the battery.

10. A secondary battery energy-saving and carbon-reducing adaptive charging control system, characterized in that, include: The low-carbon charging environment construction module is used to acquire real-time grid carbon emission factor data and, in combination with battery charging state parameters, construct a low-carbon charging environment model that characterizes the carbon emission intensity under the current charging environment. The carbon cost quantification module is used to quantify the charging loss, battery aging loss and heat dissipation energy consumption during the battery charging process into a comprehensive carbon cost based on the low-carbon charging environment model, and obtain the carbon cost quantification value. The digital twin prediction module is used to input the quantified carbon cost value into a pre-built digital twin prediction model to predict the temperature rise trend, polarization trend and aging trend of the battery during future charging process, and obtain a future state trend prediction sequence. The multi-objective dynamic weight optimization module is used to perform multi-objective dynamic weight adjustment among carbon emission reduction target, lifespan maintenance target, charging efficiency target and user demand target based on the future state trend prediction sequence, and generate the optimal charging current curve. The closed-loop learning correction module is used to perform charging according to the optimal charging current curve, and to collect actual charging process data after charging is completed, and to make online corrections to the control strategy for the next charging based on the actual charging process data.

Images