Network-formation type energy storage hierarchical collaborative control method and system based on boundary and balance linkage

Abstract

This invention discloses a hierarchical collaborative control method and system for grid-type energy storage based on boundary and equilibrium linkage, relating to the field of grid-type energy storage systems. It includes: acquiring multi-source data from each battery cell and modeling the real-time terminal voltage of each cell; establishing an inconsistency coupling index based on the relative dispersion of each dimension of the multi-source data and the corresponding weighting coefficients; introducing the inconsistency coupling index as a nonlinear decay factor to calculate the dynamic power boundary; constructing a soft-repair potential function based on the dynamic power boundary, incorporating the soft-repair potential function as a penalty term into the upper-level multi-objective fitness function for optimization, and outputting the optimal power; subtracting the optimal power from the dynamic power boundary to obtain the power boundary margin; constructing triggering conditions for boundary and equilibrium linkage based on the power boundary margin and the inconsistency coupling index; if the triggering conditions are met, the lower layer performs energy reconstruction based on a bidirectional Cuk converter to achieve energy equilibrium; if the triggering conditions are not met, the equilibrium system enters a dormant state.

Description

Technical Field

[0001] This invention relates to the field of active support and battery safety management technology for grid-type energy storage systems, specifically to a hierarchical collaborative control method and system for grid-type energy storage based on boundary and equilibrium linkage. Background Technology

[0002] Grid-based energy storage often operates under high-frequency, high-dynamic charging and discharging conditions, which places extremely stringent demands on battery safety and system-level energy management. Existing energy storage charging, discharging, and balancing control strategies generally employ a decoupled architecture between "system-level power allocation" and "cell-level passive balancing," revealing underlying logical flaws when dealing with grid-based operating conditions. These flaws manifest in the following ways:

[0003] (1) The disconnect between charge / discharge power distribution and equalization optimization leads to battery cells reaching safety limits more quickly: Existing charge / discharge optimization typically treats the battery pack as a single ideal model, with optimization objectives focused solely on maximizing grid support benefits and minimizing overall aging costs. However, due to manufacturing tolerances and uneven temperature field distribution, there are inevitably differences in internal resistance, capacity, and temperature rise characteristics among battery cells. At the physical level, total charge / discharge power is the direct driving force behind the evolution of internal inconsistencies. The same charge / discharge current flowing through differentiated internal resistances will nonlinearly amplify the SOC difference, temperature difference, and polarization voltage drop difference between cells. After the inconsistency worsens, the "weakest cell" in the system will be the first to reach safety limits such as voltage and temperature, forcing the available power window of the entire battery pack to shrink passively. Existing upper-level optimization lacks awareness of this dynamic shrinkage process and is prone to outputting aggressive commands that exceed the actual safety boundaries, leading to overload, lithium plating, or even thermal runaway of weak cells.

[0004] (2) The equalization action and power optimization lack coordinated feedback action, resulting in blind equalization: The existing equalization control at the BMS level is mostly based on fixed SOC thresholds or voltage thresholds for independent triggering, which is a post-event remedial measure. Moreover, traditional equalization circuits, such as the switched capacitor method and the Buck-Boost method, have hardware defects such as discontinuous equalization current and large ripple. When the battery is in a highly polarized state, the forced equalization will further aggravate the electrochemical stress of weak cells due to transient impact current.

[0005] (3) Control commands may cause step shocks, which can damage battery life: When the system detects that a single cell has exceeded its limit, traditional methods often use direct cut-off or step derating hard cut-off strategies. Under conditions of severe internal inconsistency, the step change in macroscopic power will be converted into a sudden large current impact on each internal cell, resulting in violent fluctuations in polarization voltage. This not only aggravates the misjudgment of BMS data acquisition, but also causes irreversible mechanical and chemical damage to battery life.

[0006] In view of the above, this application is hereby submitted. Summary of the Invention

[0007] To break down the physical barriers between charge / discharge power optimization and micro-unit balancing in traditional energy storage systems, this invention provides a hierarchical collaborative control method and system for grid-type energy storage based on boundary and balancing linkage. The core approach is based on the total charge / discharge power determining the external energy exchange intensity and balancing control determining the internal energy redistribution. By constructing an inconsistency coupling index that integrates multi-dimensional differences, it directly maps this index to the system-level dynamic power feasible domain, establishing a linkage mechanism of "boundary contraction - lower-level targeted energy reconstruction - upper-level boundary dynamic release." Simultaneously, a bidirectional Cuk converter with continuous current and no impact is matched as the underlying execution unit to eliminate the limitations on grid-type energy storage output caused by battery life safety risks due to inconsistency, maximizing the usable capacity throughout the entire lifecycle.

[0008] This invention is achieved through the following technical solution:

[0009] In a first aspect, the present invention provides a hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage, the method comprising:

[0010] Acquire multi-source data for each battery cell and perform real-time terminal voltage modeling for each cell;

[0011] Based on the relative dispersion of each dimension of multi-source data and the corresponding weight coefficients, an inconsistency coupling index is established.

[0012] Inconsistent coupling index is introduced as a nonlinear attenuation factor to calculate the dynamic power boundary;

[0013] A soft-repair potential function is constructed based on the dynamic power boundary, and the soft-repair potential function is incorporated as a penalty term into the upper-level multi-objective fitness function for optimization, and the optimal power is output.

[0014] The power boundary margin is obtained by subtracting the dynamic power boundary from the optimal power; the triggering conditions for the linkage between the boundary and the equilibrium are constructed based on the power boundary margin and the inconsistency coupling index.

[0015] If the triggering condition is met, the lower layer performs energy reconstruction based on the bidirectional Cuk converter to achieve energy balance; if the triggering condition is not met, the system enters a balanced sleep state.

[0016] Furthermore, the method also includes:

[0017] In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, the continuous duty cycle is calculated and the actual energy transfer is performed by the bidirectional Cuk converter until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the balance sleeps.

[0018] Furthermore, based on the relative dispersion of each dimension of the multi-source data and the corresponding weight coefficients, an inconsistency coupling index is established, including:

[0019] Based on multi-source data, the relative dispersion of physical quantities in each dimension is calculated; the relative dispersion includes SOC relative dispersion, voltage relative dispersion, and temperature relative dispersion.

[0020] Calculate the variance of each physical quantity within a preset time window; calculate the weighting coefficient of each physical quantity based on the variance and the sensitivity adjustment factor of each dimension.

[0021] Based on the relative dispersion and weighting coefficients, an inconsistency coupling index is constructed.

[0022] Furthermore, the expression for the inconsistency coupling index is:

[0023] ;

[0024] In the formula, This is an inconsistency coupling index; the larger the value, the weaker the weakest unit. This represents the maximum SOC difference between individual cells within the battery pack. The average SOC of the battery pack; This represents the maximum terminal voltage difference within the battery pack. This represents the average terminal voltage of the battery pack. This represents the maximum temperature difference within the battery pack. This represents the average temperature of the battery pack.

[0025] Furthermore, the expression for the dynamic power boundary is:

[0026] ;

[0027] in, The dynamic power boundary is the safe boundary of the dynamic power that the battery system is allowed to output at the current moment. This refers to the rated output power of the battery system in its brand-new condition. This represents the lowest possible health state for the entire battery pack. This is the highest single-cell temperature in the entire battery pack; The basic health contraction coefficient; This is the boundary shrinkage function derived from the inconsistency coupling index and based on the Sigmoid function.

[0028] Furthermore, the expression for the upper-level multi-objective fitness function is:

[0029] ;

[0030] In the formula, For the battery system, a multi-objective fitness function is used to synthesize the system. For the power grid support revenue function, For power grid frequency deviation, Voltage deviation; The cost function is the battery aging cost. This is the soft repair potential function; Weighting coefficients for grid support revenue; This is a weighting factor for aging costs; The penalty weighting coefficients for the soft repair potential function.

[0031] Furthermore, the soft-fix potential function is incorporated as a penalty term into the upper-level multi-objective fitness function for optimization, including:

[0032] When testing power When the boundary is exceeded, the gradient of the upper-level multi-objective fitness function with respect to power is: ,in, This represents the gradient vector of the upper-level multi-objective fitness function in the power dimension. This represents the gradient of the fundamental objective function with respect to power, excluding the penalty term. For dynamic power boundaries; The continuous pullback force gradient generated by soft repair, this force is related to the depth of the over-limit. It is proportional to the speed of the vehicle, and achieves smooth deceleration during iterative optimization.

[0033] Furthermore, the triggering conditions include:

[0034] The first condition is insufficient power boundary margin: the power boundary margin is less than the margin dead zone threshold;

[0035] The second condition is that inconsistency exceeds the limit: the value of the inconsistency coupling index is greater than the trigger release threshold;

[0036] When both of the above conditions are met, the battery system determines that relying solely on upper-level power contraction is insufficient to ensure battery safety, and lower-level balancing must be initiated for energy reconfiguration.

[0037] Furthermore, in the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, the continuous duty cycle is calculated, and the actual energy transfer is performed by the bidirectional Cuk converter until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the balance enters sleep mode, including:

[0038] In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, and state variables are defined. Discrete state equations;

[0039] According to the discrete state equation, when the state variables When the value is equal to 1, the PI controller of the bidirectional Cuk converter calculates the continuous duty cycle based on the target SOC deviation and performs actual energy transfer.

[0040] After the actual energy transfer is completed, the value of the inconsistency coupling index decreases, leading to the expansion of the dynamic power boundary, until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the system enters a state of equilibrium dormancy.

[0041] Secondly, this invention provides a hierarchical collaborative control system for grid-type energy storage based on boundary and equilibrium linkage, the system comprising:

[0042] The acquisition unit is used to acquire multi-source data of each battery cell and perform real-time terminal voltage modeling of the cells.

[0043] The Inconsistency Coupling Index Establishment Unit is used to establish an inconsistency coupling index based on the relative dispersion of each dimension of multi-source data and the corresponding weight coefficients.

[0044] The dynamic power boundary calculation unit is used to calculate the dynamic power boundary by introducing the inconsistency coupling index as a nonlinear attenuation factor.

[0045] The upper-level power boundary optimization unit is used to construct a soft-repair potential function based on the dynamic power boundary, and incorporate the soft-repair potential function as a penalty term into the upper-level multi-objective fitness function for optimization, and output the optimal power.

[0046] The trigger condition construction unit is used to calculate the power boundary margin based on the optimal power and dynamic power boundary; and to construct the trigger conditions for boundary and equilibrium linkage based on the power boundary margin and inconsistency coupling index.

[0047] The lower-level energy reconfiguration unit is used to perform energy reconfiguration based on the bidirectional Cuk converter if the triggering condition is met, thereby achieving energy balance; otherwise, it will enter a balanced sleep state.

[0048] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0049] This invention relates to a hierarchical collaborative control method and system for grid-type energy storage based on boundary and equilibrium linkage. In order to break the physical barrier between the optimization of charging and discharging power and the equilibrium of micro-cells in traditional energy storage systems, this invention is based on the core route that the total charging and discharging power determines the intensity of external energy exchange and equilibrium control determines the internal energy redistribution. By constructing an inconsistency coupling index that integrates multi-dimensional differences, it directly maps it to the dynamic power feasible domain at the system level, and establishes a linkage mechanism of "boundary contraction - lower-level targeted energy reconstruction - upper-level boundary dynamic release". At the same time, it matches a bidirectional Cuk converter with continuous current and no impact as the underlying execution unit to eliminate the limitation of battery life safety risks caused by inconsistency on the output of grid-type energy storage, and maximize the usable capacity throughout the entire life cycle.

[0050] (1) Establish an explicit mathematical mapping between inconsistent evolution and power boundary: by deriving the boundary contraction function that incorporates the Sigmoid property. And strictly defined the relative dispersion. and adaptive weights The impact of the "barrel effect" on the system's output was precisely quantified from a mathematical perspective. The dynamic erosion process eliminates the subjectivity of boundary setting in existing technologies.

[0051] (2) A soft repair mechanism based on a continuously differentiable potential function is proposed: by constructing a quadratic soft repair potential function And derive its gradient This mathematically guarantees the continuity of the first derivative of the power curve sent to the PCS, completely solving the defect of system oscillation caused by the breakage of the Jacobian matrix due to traditional hard truncation.

[0052] (3) Deep hardware and software collaboration to achieve lossless energy reconfiguration: To meet the balancing needs when inconsistencies worsen, the physical characteristics of the bidirectional Cuk converter, which has continuous current and no transient impact, are precisely matched. With the hysteresis linkage mechanism, ineffective switching losses are eliminated, and the overall conversion efficiency is improved at the system level.

[0053] (4) Actively adapt to the dynamic response requirements of grid-based energy storage: through real-time updated dynamic power boundary and soft repair potential function This ensures that when the grid frequency / voltage fluctuates drastically, the upper-level optimization algorithm can quickly converge to the optimal power command within the safety boundary. At the same time, by monitoring the boundary margin ΔP, the lower-level equalization is triggered in advance, avoiding the cumulative damage of weak cells caused by high-frequency charging and discharging switching under grid-type operating conditions. Attached Figure Description

[0054] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0055] Figure 1 The flowchart shows the hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage of the present invention.

[0056] Figure 2 This is a detailed flowchart of the hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage of the present invention;

[0057] Figure 3 This is a control diagram showing the balance between modules in this invention.

[0058] Figure 4 This is a diagram illustrating the core mechanism of the hierarchical collaborative control of grid-type energy storage based on boundary and equilibrium linkage in this invention.

[0059] Figure 5 This is a structural block diagram of the grid-type energy storage hierarchical collaborative control system based on boundary and equilibrium linkage of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0061] Example 1

[0062] like Figure 1 and Figure 2 As shown, this invention relates to a hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage. Specifically, it involves a control method based on the inconsistency evolution mechanism of battery packs, which achieves deep linkage and collaboration between "upper-layer power optimization and lower-layer equilibrium release boundary" by constructing a dynamic power feasible domain for charging and discharging and a soft-repair potential function. This method includes:

[0063] Step 1: Obtain multi-source data for each battery cell and perform real-time terminal voltage modeling for each cell;

[0064] Specifically, the terminal voltage of each individual battery cell is collected. Operating current Surface temperature The ohmic resistance of a single element is obtained through online identification. and health status Based on the first-order RC equivalent circuit model, the voltage equation for a single unit is derived as follows:

[0065] ;

[0066] In the formula, For the first Individual in Real-time terminal voltage at any given moment; For the first Individual cells regarding their state of charge The open-circuit voltage function; for The total operating current flowing through the battery pack at all times; For the first The pure ohmic internal resistance of a single unit; For the first Individual in The polarization voltage drop at time t has dynamic characteristics that satisfy the differential equation: ,in For polarization internal resistance, It is a polarizing capacitor.

[0067] Step 2: Based on the relative dispersion of each dimension of the multi-source data and the corresponding weight coefficients, establish an inconsistency coupling index;

[0068] In this embodiment, step 2 specifically includes:

[0069] Based on multi-source data, the relative dispersion of physical quantities in each dimension is calculated; the relative dispersion includes SOC relative dispersion, voltage relative dispersion, and temperature relative dispersion.

[0070] Calculate the variance of each physical quantity within a preset time window; calculate the weighting coefficient of each physical quantity based on the variance and the sensitivity adjustment factor of each dimension.

[0071] Based on the relative dispersion and weighting coefficients, an inconsistency coupling index is constructed.

[0072] Specifically, the traditional single SOC range cannot truly reflect the degree of degradation at the electrochemical level. This invention derives an inconsistency coupling index based on the nonlinear influence of various physical quantities on battery aging and degradation.

[0073] Assume the number of individual cells in the battery pack is , define the first Each dimension of physical quantity ( The relative dispersions (corresponding to SOC, voltage, and temperature, respectively) are:

[0074] ;

[0075] In the formula, For the first The relative dispersion of dimensions; For the first The monomer in the first Measured values ​​of dimensions; For the whole group The monomer in the first The maximum value of the dimension; For the whole group The monomer in the first The minimum value of the dimension; For the whole group The monomer in the first The arithmetic mean of the dimensions.

[0076] It should be noted that the SOC calculation formula is as follows:

[0077]

[0078] in, Initial time The state of charge; This refers to the battery's rated capacity. for The charging and discharging current of the battery at all times.

[0079] Length is Define the variance of the j-th physical quantity within this window. for:

[0080] ;

[0081] Where M is the number of sampling points within the window, For the j-th physical quantity, the m-th sample value is... This is the arithmetic mean of the physical quantity within the window.

[0082] Considering the exponential effect of temperature on aging rate and the direct impact of polarization voltage on the safety boundary, the weighting coefficients of each physical quantity in the inconsistency coupling index are derived. :

[0083] ;

[0084] In the formula, For the first The normalized weight coefficients of the dimension satisfy ; For the first The variance of historical data for a physical quantity within a set time window reflects the degree of fluctuation of that parameter. For the first Sensitivity adjustment factor for dimension (temperature term) Value greater than SOC item ); To sum the index variables, iterate from 1 to 3.

[0085] Finally, the inconsistency coupling index is derived:

[0086] ;

[0087] In the formula, This is an inconsistency coupling index. The larger the value, the weaker the weakest unit (the unit with the lowest safety margin (i.e., the unit most likely to fail due to parameters such as voltage, temperature, and SOC exceeding their limits)). This represents the maximum SOC difference between individual cells within the battery pack. The average SOC of the battery pack; This represents the maximum terminal voltage difference within the battery pack. This represents the average terminal voltage of the battery pack. This represents the maximum temperature difference within the battery pack. This represents the average temperature of the battery pack.

[0088] It should be noted that, For step 1 The arithmetic mean was calculated. Although there are differences such as inconsistent internal resistance between individual cells, the average terminal voltage of this battery pack is... It is not for the purpose of accurately calculating the absolute remaining charge or total energy of the battery pack, but rather to use it as a reference point to calculate the relative dispersion of the maximum terminal voltage difference.

[0089] This is obtained by calculating the arithmetic mean of the SOC of all individual cells in the battery pack. The arithmetic mean of T for all cells in the battery pack is obtained.

[0090] The above refers to the situation when the battery system discharges at high power. Increased polarization voltage drop of monomers with high internal resistance and Joule fever A sharp rise led to and Exponential expansion, directly mapped to the above formula as The surge in [data] precisely quantifies the rate of degradation of the weakest monomer.

[0091] Step 3: Introduce the inconsistency coupling index as a nonlinear attenuation factor and calculate the dynamic power boundary;

[0092] Specifically, the maximum safe discharge power (i.e., dynamic power boundary) allowed by the battery system in real time is defined as follows: Based on the relationship between power and current The key is to find the weakest single entity (let's call it a single entity). Maximum permissible safe current .

[0093] weakest single entity The safety boundary constraints are:

[0094] ;

[0095] The inverse solution yields the maximum current that the weakest cell can withstand:

[0096] ;

[0097] In the formula, The weakest single entity The maximum continuous safe current that can be flowed; This is the absolute safe cutoff voltage for a single unit. This is a steady-state approximation formula and needs to be updated in real time during actual control. To take into account dynamic polarization effects and avoid overly optimistic safety boundaries.

[0098] At this point, considering the barrel effect, the entire safe current is clamped at... Map this current to a dynamic power boundary. and introduce As a nonlinear decay factor:

[0099] ;

[0100] in, The dynamic power boundary is the safe boundary of the dynamic power that the battery system is allowed to output at the current moment. This refers to the rated output power of the battery system in its brand-new condition. This represents the lowest possible health state for the entire battery pack. This is the highest single-cell temperature in the entire battery pack; The basic health contraction coefficient; This is the boundary shrinkage function derived from the inconsistency coupling index and based on the Sigmoid function.

[0101] Basic health contraction coefficient Defined as:

[0102] ;

[0103] in,

[0104] ;

[0105] symbol = max , Take 0.3; when >80%, .

[0106] The temperature-based shrinkage coefficient is defined as:

[0107] ;

[0108] In the formula: This refers to the battery's rated operating temperature. This refers to the maximum permissible operating temperature of the battery.

[0109] To avoid The boundary drops sharply just after exceeding the threshold; the boundary contraction function is derived based on the Sigmoid function. :

[0110] ;

[0111] In the formula, This is the boundary contraction function, i.e., the boundary contraction proportionality coefficient, with a value range of (0,1). Inconsistent safety warning thresholds; The kurtosis parameter determines the sensitivity of boundary contraction. ;when hour, The boundary does not shrink; when cross back, It exhibits an exponential decay.

[0112] Step 4, based on dynamic power boundary A soft-repair potential function is constructed and incorporated as a penalty term into the upper-level multi-objective fitness function F for optimization, outputting the optimal power. ;

[0113] Traditional optimization algorithms struggle when encountering dynamic power boundaries. Using hard truncation at times will cause the Jacobian matrix of control commands to be discontinuous, resulting in a step shock to the PCS.

[0114] This invention introduces the ideas of the Lagrange multiplier method and the potential field method to construct a continuously differentiable soft-repair potential function. :

[0115] ;

[0116] In the formula, This is the soft-repair potential function, i.e., the virtual potential energy penalty value generated by power exceeding the limit; This represents the charge / discharge power value currently being tested by the upper-level optimization algorithm. The potential field stiffness coefficient determines the pullback force after exceeding the limit. .

[0117] This soft repair potential function As a penalty term, it is incorporated into the upper-level multi-objective fitness function. middle:

[0118] ;

[0119] In the formula, For the battery system, a multi-objective fitness function is used to synthesize the system. For the power grid support revenue function, For power grid frequency deviation, Voltage deviation; The cost function is the battery aging cost. This is the soft repair potential function; Weighting coefficients for grid support revenue; This is a weighting factor for aging costs; The penalty weighting coefficients for the soft repair potential function.

[0120] When testing power When the boundary is exceeded, the gradient of the upper-level multi-objective fitness function with respect to power is:

[0121] ;

[0122] in, This represents the gradient vector of the upper-level multi-objective fitness function in the power dimension. This represents the gradient of the fundamental objective function with respect to power, excluding the penalty term. For dynamic power boundaries; This is a continuous pullback force gradient generated by soft repair. This force is related to the depth of the over-limit repair. It is proportional to the speed of the vehicle, and achieves smooth deceleration during iterative optimization.

[0123] Update the formula at the location of the heuristic algorithm. In this case, since the penalty force is proportional to the depth of the out-of-bounds error, the algorithm naturally slows down as it approaches the boundary, eliminating the abrupt change in higher-order derivatives caused by hard truncation.

[0124] Step 4 above is an optimization of step 3, where the dynamic power boundary was calculated. The traditional approach is to abruptly cut off the circuit at this boundary, which can damage the battery. Therefore, step 4, based on step 3, utilizes dynamic power boundaries. Constructing the soft repair potential function When the power of the optimization algorithm exceeds At that time, the soft repair potential function in step 4 This will generate a pull-back force proportional to the depth of the overshoot, resulting in a gradual change that is incorporated into subsequent optimization calculations. This helps to avoid step-like impacts.

[0125] Step 5, set the dynamic power boundary. With optimal power Subtraction yields the power boundary margin. According to power boundary margin Coupling metrics with inconsistency Establish triggering conditions for the linkage between boundaries and equilibrium;

[0126] Specifically, the upper-level optimizer (i.e., the upper-level multi-objective fitness function) outputs the optimal power. Then, the power boundary margin of the battery system is calculated. :

[0127] ;

[0128] In the formula, The optimal total charge and discharge power command output by the upper-level optimizer (i.e., the upper-level multi-objective fitness function) after soft-fix constraints is the optimal power. This refers to the remaining margin in the power space, i.e., the power boundary margin.

[0129] Set margin dead zone threshold Perform a linked trigger logic judgment:

[0130] like This indicates that the current inconsistency has not severely eroded the output space, and the balancing circuit remains dormant.

[0131] like This triggers the boundary release requirement. At this point, the system calculates that in order to... To restore to a safe level, it is necessary to The amount of reduction And send it down to the lower-level equalizer.

[0132] The trigger condition for the boundary and equilibrium linkage is that the following two conditions are met simultaneously:

[0133] (1) The first condition is insufficient power boundary margin: the power boundary margin is less than the margin dead zone threshold, i.e., ΔP < Where ΔP is the power boundary margin, The margin dead zone threshold, The value is between 0.1 and 0.2 times. ;

[0134] (2) The second condition is that the inconsistency exceeds the limit: the value of the inconsistency coupling index is greater than the trigger release threshold, i.e. > ,in This is an inconsistent coupling indicator. Trigger release threshold;

[0135] When both of the above conditions are met, the battery system determines that relying solely on upper-level power reduction is insufficient to ensure battery safety, and lower-level equalization must be initiated for energy reconfiguration. At this point, the target... Decrease Δ = - , Stop release threshold ( ).

[0136] Step 6: If the triggering condition is met, the lower layer performs energy reconstruction based on the bidirectional Cuk converter to achieve energy balance; if the triggering condition is not met, the system enters a balanced sleep state.

[0137] Specifically, after receiving a release request, the lower-level equalizer performs energy transfer using a bidirectional Cuk converter between adjacent cells, such as... Figure 3 As shown, Figure 3 The module group intra- / inter-module equalization control diagram is derived; its steady-state mathematical model in continuous conduction mode is derived, and the average equalization current on the output side is calculated. Duty cycle The relationship is:

[0138] ;

[0139] In the formula, The average equalization current transferred by the Cuk converter; the real-time terminal voltage of the high SOC cell (energy output side), defined as... ; The real-time terminal voltage of a low SOC cell (energy receiving side) is defined as... ; PWM duty cycle of high-frequency MOSFETs, value range ; The equivalent loop resistance, including the parasitic resistance of the line and inductance, is defined as... ,in: , These are the winding resistances of the inductors on the input and output sides of the Cuk converter, respectively. The equivalent series resistance of a Cuk capacitor; The resistance of the connecting lines.

[0140] Since both the input and output terminals of the Cuk circuit have inductors connected in series (let the output inductance be...), According to the principle that inductor current cannot change abruptly ( The rate of change of current is limited by the voltage across the inductor. Under the most demanding startup conditions, the inductor... The maximum voltage across the terminals is Therefore, the upper limit of the rate of change of current is:

[0141] ;

[0142] In the formula, The maximum rate of change of the equilibrium current; The maximum value of the terminal voltage of a high-SOC single unit; : The inductance of the filter inductor on the output side of the Cuk converter.

[0143] By properly designing the inductor parameters, ensure that the current ripple rate meets the system requirements:

[0144] ;

[0145] in This design ensures the current ripple rate to match the system's maximum allowable current change rate.

[0146] ;

[0147] In the formula, Current ripple rate; This represents the ripple amplitude of the output-side inductor current. This represents the average value of the output-side inductor current.

[0148] The above formulas are energy reconstruction models, representing how much current is transferred, proving that the start-up is safe and shock-free, and evaluating the smoothness of the current during the transfer process.

[0149] Step 7: In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, the continuous duty cycle is calculated and the actual energy transfer is performed by the bidirectional Cuk converter until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the balance sleeps.

[0150] Specifically, to prevent the battery system from Frequent oscillations near the boundary necessitate the introduction of a hysteresis-based mathematical logic judgment into the equilibrium enable signal:

[0151] Define state variables Discrete state equations:

[0152] ;

[0153] In the formula, for The balanced enable state of each control cycle (1 for on, 0 for sleep). Next control cycle; The upper limit threshold for inconsistency that triggers load balancing startup; The inconsistency threshold that triggers the balancing to stop.

[0154] when At that time, the PI controller of the bidirectional Cuk converter is based on the target SOC deviation. Calculate the continuous duty cycle:

[0155] ;

[0156] In the formula, No. The PWM duty cycle output per cycle; Proportional adjustment coefficient; Integral adjustment coefficient; No. Historical sequence of the difference between the short-term individual and the average SOC.

[0157] Energy transfer leads to The descent occurs according to the Sigmoid function in step 3. characteristic, Subsequently, it rebounded and expanded. Regain size, eventually Time-balanced hibernation, completing one "boundary contraction" cycle. Lower layer reconstruction The closed loop of "boundary release".

[0158] The above are as follows Figure 4 As shown, Figure 4 This is a diagram illustrating the core mechanism of hierarchical collaborative control of grid-type energy storage based on boundary and equilibrium linkage.

[0159] Example 2

[0160] like Figure 5 As shown, the difference between this embodiment and Embodiment 1 is that this embodiment provides a hierarchical collaborative control system for grid-type energy storage based on boundary and equilibrium linkage. This system corresponds one-to-one with the hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage in Embodiment 1. The system includes:

[0161] The acquisition unit is used to acquire multi-source data of each battery cell and perform real-time terminal voltage modeling of the cells.

[0162] The Inconsistency Coupling Index Establishment Unit is used to establish an inconsistency coupling index based on the relative dispersion of each dimension of multi-source data and the corresponding weight coefficients.

[0163] The dynamic power boundary calculation unit is used to calculate the dynamic power boundary by introducing the inconsistency coupling index as a nonlinear attenuation factor.

[0164] The upper-level power boundary optimization unit is used to construct a soft-repair potential function based on the dynamic power boundary, and incorporate the soft-repair potential function as a penalty term into the upper-level multi-objective fitness function for optimization, and output the optimal power.

[0165] The trigger condition construction unit is used to calculate the power boundary margin based on the optimal power and dynamic power boundary; and to construct the trigger conditions for boundary and equilibrium linkage based on the power boundary margin and inconsistency coupling index.

[0166] The lower-level energy reconfiguration unit is used to perform energy reconfiguration based on the bidirectional Cuk converter if the triggering condition is met, thereby achieving energy balance; otherwise, it will enter a balanced sleep state.

[0167] The execution process of each unit can be carried out according to the process steps of the hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage in Example 1. In this example, they will not be described in detail.

[0168] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0169] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0170] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0171] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0172] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A network-constructing type energy storage hierarchical collaborative control method based on boundary and balanced linkage, characterized in that, The method includes: Acquire multi-source data for each battery cell and perform real-time terminal voltage modeling for each cell; Based on the relative dispersion of each dimension of the multi-source data and the corresponding weight coefficients, an inconsistency coupling index is established. The inconsistency coupling index is introduced as a nonlinear attenuation factor to calculate the dynamic power boundary. A soft-repair potential function is constructed based on the dynamic power boundary, and the soft-repair potential function is incorporated as a penalty term into the upper-level multi-objective fitness function for optimization, and the optimal power is output. Subtract the optimal power from the dynamic power boundary to obtain the power boundary margin; construct the triggering conditions for boundary and equilibrium linkage based on the power boundary margin and the inconsistency coupling index; If the triggering condition is met, the lower layer performs energy reconstruction based on the bidirectional Cuk converter to achieve energy balance; if the triggering condition is not met, the energy balance goes into sleep mode.

2. The border and balance linkage-based network construction type energy storage hierarchical coordination control method according to claim 1, characterized in that, The method also includes: In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, the continuous duty cycle is calculated and the actual energy transfer is performed by the bidirectional Cuk converter until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the balance sleeps.

3. The border and balance linkage-based network construction type energy storage hierarchical coordination control method according to claim 1, characterized in that, Based on the relative dispersion of each dimension of the multi-source data and the corresponding weight coefficients, an inconsistency coupling index is established, including: Based on the multi-source data, the relative dispersion of each physical quantity is calculated; the relative dispersion includes SOC relative dispersion, voltage relative dispersion, and temperature relative dispersion. Calculate the variance of each physical quantity within a preset time window; calculate the weighting coefficient of each physical quantity based on the variance and the sensitivity adjustment factor of each dimension. Based on the relative dispersion and the weighting coefficients, an inconsistency coupling index is constructed.

4. The hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage according to claim 3, characterized in that, The expression for the inconsistency coupling index is: ； In the formula, This is an inconsistency coupling index; the larger the value, the weaker the weakest unit. This represents the maximum SOC difference between individual cells within the battery pack. The average SOC of the battery pack; This represents the maximum terminal voltage difference within the battery pack. This represents the average terminal voltage of the battery pack. This represents the maximum temperature difference within the battery pack. This represents the average temperature of the battery pack.

5. The border and balance linkage-based network construction type energy storage hierarchical coordination control method according to claim 1, characterized in that, The expression for the dynamic power boundary is: ； in, The dynamic power boundary is the safe boundary of the dynamic power that the battery system is allowed to output at the current moment. This refers to the rated output power of the battery system in its brand-new condition. This represents the lowest possible health state for the entire battery pack. This is the highest single-cell temperature in the entire battery pack; The basic health contraction coefficient; This is the boundary shrinkage function derived from the inconsistency coupling index and based on the Sigmoid function.

6. The hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage as described in claim 1, characterized in that, The expression for the upper-level multi-objective fitness function is: ； In the formula, For the battery system, a multi-objective fitness function is used to synthesize the system. For the power grid support revenue function, For power grid frequency deviation, Voltage deviation; The cost function is the battery aging cost function; This is the soft repair potential function; The weighting coefficient for grid support revenue; This is a weighting factor for aging costs; represents the penalty weighting coefficient for the soft repair potential function.

7. The border and balance linkage-based network construction type energy storage hierarchical coordination control method according to claim 1, characterized in that, The soft-repair potential function is incorporated as a penalty term into the upper-level multi-objective fitness function for optimization, including: When testing power When the boundary is exceeded, the gradient of the upper-level multi-objective fitness function with respect to power is: ,in, This is the gradient vector of the upper-layer multi-objective fitness function in the power dimension; This represents the gradient of the fundamental objective function with respect to power, excluding the penalty term. For dynamic power boundaries; The continuous pullback force gradient generated by soft repair, this force is related to the depth of the over-limit. It is proportional to the speed of the vehicle, and achieves smooth deceleration during iterative optimization.

8. The hierarchical collaborative control method for grid-type energy storage based on boundary and equilibrium linkage according to claim 1, characterized in that, The triggering conditions include: The first condition is insufficient power boundary margin: the power boundary margin is less than the margin dead zone threshold; The second condition is that inconsistency exceeds the limit: the value of the inconsistency coupling index is greater than the trigger release threshold; When both of the above conditions are met, the battery system determines that relying solely on upper-level power contraction is insufficient to ensure battery safety, and lower-level balancing must be initiated for energy reconfiguration.

9. The border and balance linkage-based network construction type energy storage hierarchical coordination control method according to claim 2, characterized in that, In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, the continuous duty cycle is calculated, and the actual energy transfer is performed by the bidirectional Cuk converter until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the balance goes into sleep mode, including: In the process of achieving energy balance, the balance enable signal is introduced into the hysteresis logic judgment, and state variables are defined. Discrete state equations; According to the discrete state equation, when the state variable When the value is equal to 1, the PI controller of the bidirectional Cuk converter calculates the continuous duty cycle based on the target SOC deviation and performs actual energy transfer. After the actual energy transfer is completed, the value of the inconsistency coupling index decreases, causing the dynamic power boundary to expand until the inconsistency coupling index is less than the inconsistency lower limit threshold, at which point the system enters a balanced dormancy state.

10. A network-structured energy storage hierarchical collaborative control system based on boundary and balanced linkage, characterized in that, The system includes: The acquisition unit is used to acquire multi-source data of each battery cell and perform real-time terminal voltage modeling of the cells. The inconsistency coupling index establishment unit is used to establish an inconsistency coupling index based on the relative dispersion of each dimension of the multi-source data and the corresponding weight coefficients. The dynamic power boundary calculation unit is used to incorporate the inconsistency coupling index as a nonlinear attenuation factor to calculate the dynamic power boundary. The upper-level power boundary optimization unit is used to construct a soft-repair potential function based on the dynamic power boundary, and to incorporate the soft-repair potential function as a penalty term into the upper-level multi-objective fitness function for optimization, and output the optimal power. The trigger condition construction unit is used to calculate the power boundary margin based on the optimal power and the dynamic power boundary; and to construct the trigger conditions for boundary and equilibrium linkage based on the power boundary margin and the inconsistency coupling index. The lower-level energy reconfiguration unit is used to perform energy reconfiguration based on the bidirectional Cuk converter to achieve energy balance if the triggering condition is met; otherwise, it will go into a balanced sleep state.

Images