A method for coordinated fast frequency regulation control of a thermal power unit and a flywheel energy storage array

Abstract

This invention discloses a collaborative fast frequency regulation control method for thermal power units and flywheel energy storage arrays. Addressing the problems existing in the prior art, this invention includes: real-time acquisition of grid frequency, AGC commands, thermal power unit operating parameters, and the state of charge of each flywheel; online evaluation of the thermal power unit's AGC response margin; decomposition of AGC power deviation into high-frequency and low-frequency components, and dynamic determination of the power allocation coefficient based on the real-time adjustable margin of the thermal power unit and the flywheel array state; optimization of power allocation within the flywheel array using priority call to appropriately charged units, over-limit protection, and equalization control strategies; construction of a three-layer multi-timescale collaborative architecture comprising millisecond-level flywheel autonomous primary frequency regulation, second-level joint AGC frequency regulation, and minute-level charge recovery; real-time monitoring of the thermal power regulation rate and dynamic compensation and correction through the flywheel; and establishment of a comprehensive frequency regulation performance evaluation index for adaptive optimization. This invention can effectively improve the frequency regulation response speed, accuracy, and economy of the thermal power-flywheel combined system.

Description

Technical Field

[0001] This invention belongs to the field of power system frequency regulation control technology, specifically relating to a collaborative fast frequency regulation control method for thermal power units and flywheel energy storage arrays, applicable to automatic power generation control frequency regulation scenarios of coal-fired power units coupled with flywheel energy storage arrays. Background Technology

[0002] With the rapid growth of installed capacity of new energy power generation, grid frequency fluctuations are becoming more frequent and their amplitudes are increasing, placing higher demands on the response speed, regulation accuracy, and sustainability of frequency regulation resources. Thermal power units, as the main resource for grid frequency regulation, have inherent limitations in their response characteristics: the response time of turbine regulating valves is on the order of seconds, and the response time of boiler thermal inertia is on the order of minutes. During large-scale load changes, there are response delays and reverse regulation issues. When the grid frequency changes abruptly, thermal power units often struggle to provide sufficient power support in the immediate event, affecting grid frequency stability.

[0003] Flywheel energy storage, as a physical energy storage technology with high power density and long cycle life, boasts significant advantages such as millisecond-level response speed, a cycle life exceeding 10 million cycles, and charge / discharge efficiency exceeding 87%–90%. It is particularly adept at high-frequency, instantaneous, and precise charge / discharge. Combining flywheel energy storage with thermal power units to form a "thermal power + flywheel" coordinated frequency regulation mode can effectively compensate for the slow response of thermal power units, achieving complementary advantages.

[0004] Since putting our energy storage frequency regulation power station into operation in 2019, we have accumulated rich experience in the joint frequency regulation operation of thermal power and energy storage. Our existing patented technology reserves cover multiple areas such as primary frequency regulation time-sharing control and DCS system transformation, laying a technical foundation for achieving more efficient collaborative control.

[0005] However, existing technical solutions still have the following shortcomings in actual operation:

[0006] (1) Coarse power allocation strategy: Existing methods mostly adopt a fixed ratio allocation strategy based on high and low frequency decomposition (such as the intrinsic mode function decomposition method), which allocates high frequency components to the flywheel and low frequency components to the thermal power unit. However, in actual operation, the real-time adjustable capacity of the thermal power unit is limited by the current operating conditions (deep adjustment state, load segment, coal quality changes, etc.), and the fixed allocation strategy is difficult to adapt to the dynamically changing operating constraints.

[0007] (2) Inadequate SOC (State of Charge) management strategy: During long-term, high-frequency charging and discharging, flywheel energy storage arrays are prone to situations where the SOC of a single flywheel deviates from the reasonable operating range. Existing methods mostly consider SOC management from the overall array level, lacking fine-grained control over the differentiated SOC states of each unit within the array.

[0008] (3) Insufficient consideration of the constraints of the thermal power unit: The existing method takes into account the dynamic characteristics of the thermal power unit less and does not fully consider factors such as the load regulation rate limit, regulation dead zone, and main steam pressure constraint of the unit, which leads to regulation deviation when the frequency regulation command exceeds the actual executable range of the unit.

[0009] (4) The multi-time scale coordination mechanism is imperfect: the timing connection and power superposition between primary frequency regulation (second-level response) and AGC frequency regulation (minute-level response) lack effective coordination, which can easily cause the flywheel to exit prematurely or the thermal power unit to be over-regulated. Summary of the Invention

[0010] The purpose of this invention is to overcome the shortcomings of the prior art and provide a collaborative and rapid frequency regulation control method for thermal power units and flywheel energy storage arrays. Through dynamic power allocation, multi-time-scale collaborative control, adaptive SOC equalization management, and real-time evaluation of thermal power unit AGC response margin, this method achieves efficient collaborative frequency regulation between thermal power and flywheel energy storage arrays, significantly improving the overall frequency regulation performance of the unit.

[0011] To achieve the above objectives, the present invention provides a coordinated fast frequency modulation control method for thermal power units and flywheel energy storage arrays, comprising the following steps:

[0012] Step 1: System Parameter Acquisition and Status Awareness

[0013] The following parameters are collected in real time:

[0014] The power grid frequency signal f and the frequency change rate df / dt (obtained via PMU device);

[0015] AGC scheduling instruction P AGC (Received via RTU device);

[0016] Real-time operating data of thermal power units, including: actual unit output P unit Main steam pressure P ms Load change rate limit R limit Current AGC response margin ΔP unit (i.e., the adjustable capacity between the unit's current load point and the boundary of the adjustable range allowed by the AGC command);

[0017] Real-time SOC value of each flywheel energy storage unit in the flywheel energy storage array i (i=1, 2, ..., n, where n is the number of flywheels), chargeable / dischargeable power limit P charge-i and P discharge-i Real-time rotational speed ω i .

[0018] Step 2: Online assessment of AGC response margin for thermal power units

[0019] Based on the current operating status of the thermal power unit, the online computer group's AGC response margin ΔP unit :

[0020] ΔP unit =min(ΔP upper ΔP lower )

[0021] in:

[0022] ΔP upper =P max -P unit P indicates the unit's upward adjustable margin. max This is the maximum allowable output of the unit under the current operating conditions (limited by main steam pressure, coal feed rate, environmental emissions, etc.).

[0023] ΔP lower =P unit -P min P indicates the downward adjustment margin of the unit. min This is the minimum allowable output of the unit under the current operating conditions (constrained by factors such as stable fuel load and deep adjustment capability).

[0024] Simultaneously considering the unit load change rate limit R limit The actual dynamic margin ΔP that the computer group can respond to in the next control cycle Δt. dyn =R limit ×Δt.

[0025] ΔP unit and ΔP dyn As a key constraint parameter for subsequent power allocation.

[0026] Step 3: Real-time decomposition and allocation of frequency modulation power demand

[0027] The power deviation ΔP between AGC scheduling instructions and the actual output of the unit cmd = P AGC - P unit A multi-timescale decomposition method is adopted:

[0028] (1) ΔP cmd Decomposed into high-frequency fluctuation components ΔP HF and low-frequency trend component ΔP LF ;

[0029] (2) Based on the real-time AGC response margin ΔP of the thermal power unit unit Total adjustable power capacity P of flywheel array fly_total Dynamically determine the power allocation coefficient α:

[0030] α = f(ΔP unit Pfly_total SOC avg ,df / dt)

[0031] in:

[0032] SOC avg The average SOC value of the flywheel array;

[0033] f(·) is a dynamic allocation function based on fuzzy logic or adaptive weight coefficients;

[0034] (3) Generate the target power command P for the flywheel array fly_cmd = α × ΔP cmd + β × ΔP HF Generate the target power command P for thermal power units unit_cmd = (1−α) × ΔP cmd + P unit_base Where β is the high-frequency compensation coefficient, Pu nit_base It contributes to the current foundation of the unit.

[0035] Step 4: Optimize power allocation within the flywheel array

[0036] The total target power command P of the flywheel array fly_cmd The energy is allocated to each flywheel energy storage unit, following the principles of the allocation strategy:

[0037] (1) Prioritize the use of appropriate SOC cells: Set the ideal working range of SOC [SOC] low SOC high (e.g., 40% to 60%), priority is given to assigning charging / discharging tasks to flywheel units with SOC in this range;

[0038] (2) SOC over-limit protection: When the SOC of a flywheel unit approaches the upper limit (> When the system needs charging, limit or prohibit the unit from continuing to charge; when the SOC is close to the lower limit (> Furthermore, when the system needs to discharge, the discharge of this unit should be restricted or prohibited.

[0039] (3) Balanced control: By introducing the SOC equalization factor, the power distribution weight of each flywheel unit is dynamically adjusted so that the SOC of each unit in the array gradually becomes consistent, avoiding overuse of a single unit;

[0040] (4) Response speed matching: High-frequency, small-amplitude power fluctuations are preferentially allocated to the units with higher speed and faster dynamic response in the flywheel array to achieve "fast to fast, slow to slow" matching control.

[0041] Step 5: Multi-timescale collaborative control strategy

[0042] Construct a three-tier collaborative control architecture:

[0043] First layer (millisecond level): The flywheel energy storage array autonomously responds to the primary frequency regulation demand based on the grid frequency deviation and frequency change rate. When the grid frequency exceeds the dead zone, it immediately releases or absorbs power with a response time of no more than 50 milliseconds.

[0044] Second level (second level): The control system adjusts the power distribution ratio between the flywheel array and the thermal power unit in real time according to the changes in AGC commands and the response margin of the thermal power unit. The flywheel bears the rapidly changing power components, and the thermal power unit gradually takes over.

[0045] The third layer (minute level): Based on the SOC status of the flywheel array, the thermal power unit is coordinated to perform SOC recovery adjustment. When the flywheel SOC deviates from the ideal range, the thermal power unit provides reverse power support to gradually bring the flywheel SOC back to the target range.

[0046] The three layers achieve a smooth transition through time window sliding average and status feedback, avoiding control command conflicts.

[0047] Step Six: Real-time Compensation and Correction of AGC Performance in Thermal Power Units

[0048] During the execution of AGC commands by thermal power units, the actual rate of change of unit output dP is monitored in real time. unit / dt, and the rate of change of the instruction dP cmd Comparison with / dt:

[0049] (1) When dP unit / dt <dP cmd When / dt, it indicates that the thermal power unit is responding lagging. The control system automatically increases the output power of the flywheel array for dynamic compensation to make up for the adjustment gap of the thermal power unit.

[0050] (2) When dP_unit / dt>dP_cmd / dt, it indicates that the thermal power unit may have overshoot. The control system should reduce the output of the flywheel array or adjust its charging / discharging direction to prevent the system power from overshooting.

[0051] (3) When the AGC response margin ΔP of the thermal power unit unit When the power regulation depth of the flywheel array is less than the preset threshold, the control system automatically reduces the power regulation depth of the flywheel array to prevent over-discharge or over-charge of the flywheel and ensure safe operation of the system.

[0052] Step 7: Closed-loop evaluation and adaptive optimization of coordinated frequency modulation performance

[0053] Establish a comprehensive evaluation index K for frequency modulation performance p :

[0054] K p = w1 × Kdelay + w2 × Ka ccuracy + w3 × K correlation

[0055] in:

[0056] K delay For response delay metrics (average time from command to power response);

[0057] K accuracy To adjust the accuracy index (the average deviation between the actual output and the target output);

[0058] K correlation The correlation index (correlation coefficient between the actual output curve and the command curve);

[0059] w1, w2, and w3 are weighting coefficients, which are dynamically adjusted according to the power grid dispatch assessment rules.

[0060] Real-time calculation of K by the control system p Value, when K p When the values ​​are below the preset target, the power distribution coefficient α and the SOC equalization parameter are automatically optimized and adjusted to achieve adaptive iterative optimization of the control strategy.

[0061] Compared with the prior art, the beneficial effects of the present invention are:

[0062] 1. This invention solves the problem of poor adaptability of fixed ratio power allocation strategies by online assessment of the AGC response margin and dynamic power allocation of thermal power units, effectively improving frequency regulation response speed and adjustment accuracy. Referring to measured data from similar projects at Ningxia Lingwu Company, the overall primary frequency regulation qualification rate of the unit can be increased from the pre-modification level to over 92%, and the AGC comprehensive ramp-up performance index K-value is improved by approximately 30%.

[0063] 2. This invention avoids overuse and frequent overcharging and over-discharging of individual flywheel units through the SOC equalization management strategy inside the flywheel array, so that the number of charge and discharge cycles of each unit is evenly distributed, which can give full play to the advantage of the flywheel's energy storage cycle life of more than 10 million cycles.

[0064] 3. This invention fully considers the load change rate limit, main steam pressure constraint and deep adjustment capability limit of thermal power units, and avoids frequency regulation commands exceeding the actual execution capability of the unit, which may lead to regulation deviation or equipment safety hazards.

[0065] 4. This invention achieves seamless integration of millisecond-level (flywheel), second-level (joint regulation), and minute-level (SOC recovery) through a multi-timescale three-layer collaborative control architecture, which can effectively suppress grid frequency fluctuations and improve grid frequency stability in scenarios with a high proportion of new energy sources.

[0066] 5. This invention, through continuous optimization of frequency regulation performance indicators, can improve the frequency regulation revenue of generating units in the power ancillary services market. The flywheel accelerates charging and decelerates discharging, fully leveraging the performance advantages of high-frequency, instantaneous, precise charging and discharging, thereby enhancing the unit's competitiveness in the grid frequency regulation market. Attached Figure Description

[0067] Figure 1 This is a diagram illustrating the overall architecture of the collaborative rapid frequency modulation control system for thermal power units and flywheel energy storage arrays according to the present invention.

[0068] Figure 2 This is a flowchart of the online evaluation process for the AGC response margin of thermal power units according to the present invention;

[0069] Figure 3 This is a schematic diagram of the multi-timescale three-layer collaborative control architecture of the present invention;

[0070] Figure 4 This is a flowchart of the internal power optimization allocation strategy for the flywheel array of the present invention;

[0071] Figure 5 This is a flowchart of the closed-loop evaluation and adaptive optimization of frequency modulation performance of the present invention. Detailed Implementation

[0072] The present invention will be further described below with reference to the embodiments and accompanying drawings.

[0073] Example: A 600MW supercritical coal-fired power unit coupled with a 6×4.5MW / 1.125MWh flywheel energy storage array

[0074] Taking a 600MW supercritical coal-fired power unit coupled with 6 fully magnetically levitated flywheel energy storage arrays with a single unit capacity of 4.5MW / 1.125MWh as an example (the total capacity of the flywheel array is 27MW / 6.75MWh, the charge-discharge cycle efficiency is greater than 87%, the single full-power discharge time is 15 minutes, and the cycle life is greater than 10 million cycles), the specific implementation process of the present invention is explained.

[0075] Step 1: System Configuration and Parameter Initialization

[0076] Configure a PMU device to collect power grid frequency signals with a sampling period of 20ms;

[0077] Configure the RTU device to receive AGC scheduling commands, with a communication cycle of 2 seconds;

[0078] Configure a flywheel array management system to collect the SOC (calculated as the square ratio of rotational speed) and charge / discharge power limit of each flywheel unit in real time;

[0079] Configure the DCS system interface of the thermal power unit to collect the actual output of the unit, main steam pressure, and load change rate limit (in this embodiment, it is set to 3% / min of the rated output of the unit, i.e., 18MW / min).

[0080] Initialize control parameters: ideal operating range of SOC [40%, 60%], upper and lower limits of SOC [20%, 80%], and AGC response margin alarm threshold of 5MW for thermal power units.

[0081] Step 2: Real-time Coordinated Frequency Modulation Process

[0082] Assuming the grid frequency suddenly drops by 0.08Hz (exceeding the primary frequency regulation dead zone ±0.033Hz), the AGC simultaneously issues a load increase command, targeting an additional 15MW of generating power. The control system performs the following operations:

[0083] First-level response (millisecond level): The flywheel array detects a frequency drop and immediately responds to the frequency adjustment demand. The six flywheels automatically discharge according to the SOC balancing strategy, releasing approximately 3MW of power within 50ms to quickly support the recovery of the grid frequency.

[0084] Second-level response (seconds): The control system calculates the power deviation ΔP. cmd =15MW, combined with the current AGC upward response margin ΔP of the thermal power unit unit =30MW (current unit output is 450MW, P) max =480MW), dynamically calculate the power allocation coefficient α=0.6 (the flywheel bears 60% of the AGC incremental demand), and generate the flywheel target command P. fly_cmd = 9MW, with a target command increment of 6MW for thermal power units. The flywheel array management system allocates 9MW of power to 6 flywheels, prioritizing flywheel units with a SOC between 40% and 60%, and reducing or suspending the discharge of units with excessively high or low SOC.

[0085] Third-level response (minute-level): The thermal power unit gradually increases its output. Once the actual output reaches the target value, the flywheel array gradually reduces its discharge power. Simultaneously, it determines whether SOC recovery is needed based on the SOC status. If the SOC of a certain flywheel drops to 30% after discharge, below the lower limit of the ideal range, the control system automatically sends a fine-tuning command to the thermal power unit, which provides a small amount of reverse power to gradually restore the flywheel SOC to above 40%.

[0086] Step 3: Dynamic Compensation and Correction

[0087] During the load increase process of the thermal power unit, the control system continuously monitors the actual rate of change of the unit's output. In this embodiment, the unit's load change rate is limited to 18MW / min, and the AGC requires an additional 15MW, with a theoretical minimum completion time of approximately 50 seconds. If, in the initial stage of load increase (the first 10 seconds), the unit's actual output only increases by 2MW, while the theoretical increase should be 3MW, the control system automatically increases the flywheel array discharge power by 1MW for dynamic compensation, ensuring that the total power response tracks the AGC command. As the unit's output gradually catches up, the flywheel compensation automatically decreases, achieving a smooth handover.

[0088] Step 4: Performance Evaluation and Optimization

[0089] The control system calculates the comprehensive evaluation index K of frequency modulation performance every 5 minutes. p Assume that during a certain assessment period: response delay K delay =0.85, Adjustment precision K accuracy =0.90, Relevance K correlation =0.88, overall K p =0.87. The control system determines K. p If the value is below the target value of 0.90, the fuzzy control rule for automatically adjusting the power distribution coefficient α increases the participation ratio of the flywheel in the initial stage of AGC response, thereby improving the response speed. Simultaneously, the SOC balancing weight is optimized, and K is gradually adjusted in subsequent control cycles. p Increase to above the target value.

[0090] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for coordinated rapid frequency modulation control of a thermal power unit and a flywheel energy storage array, characterized in that, Includes the following steps: Step 1: Real-time acquisition of power grid frequency signals, AGC dispatch commands, real-time operating parameters of thermal power units, and the state of charge and power limits of each flywheel unit in the flywheel energy storage array; Step 2: Based on the current operating conditions of the thermal power unit, evaluate the unit's AGC response margin online, including the computer group's upward adjustable margin, downward adjustable margin, and dynamic response margin within the control cycle; Step 3: Decompose the power deviation between the AGC scheduling command and the actual output of the thermal power unit into high-frequency fluctuation components and low-frequency trend components. Based on the real-time AGC response margin of the thermal power unit, the total adjustable power capacity of the flywheel array, and the average state of charge, dynamically determine the power allocation coefficient and generate the target power command for the flywheel array and the target power command for the thermal power unit. Step 4: Optimize the allocation of the target power command of the flywheel array among the flywheel units. The allocation strategy includes: prioritizing the use of flywheel units with a state of charge within a preset ideal range, providing charge and discharge limit protection for flywheel units with a state of charge exceeding the limit, and dynamically adjusting the power allocation weight of each flywheel unit based on the state of charge equalization factor. Step 5: Construct a multi-timescale collaborative control architecture, including: millisecond-level autonomous response of the flywheel energy storage array to primary frequency regulation requirements, second-level joint response of the thermal power unit and flywheel array to AGC frequency regulation commands, and minute-level recovery regulation based on the state of charge of the flywheel array. Smooth transition between timescales is achieved through state feedback. Step 6: During the execution of AGC commands by the thermal power unit, the actual output change rate of the unit is monitored in real time. When the actual output change rate of the unit is less than the command change rate, the output power of the flywheel array is automatically increased for dynamic compensation. When the actual output change rate of the unit is greater than the command change rate, the output of the flywheel array is automatically adjusted to suppress power overshoot. Step 7: Establish a comprehensive frequency modulation performance evaluation index that includes response delay, regulation accuracy, and correlation. Based on this index, adaptively optimize and adjust the power distribution coefficient and state-of-charge equalization parameters.

2. The method according to claim 1, characterized in that, The online assessment of the AGC response margin of thermal power units in step two specifically includes: Obtain the current actual output P of the thermal power unit unit The maximum allowable output P under the current operating conditions max and minimum output P min and the unit load change rate limit R limit ; Calculate the static upward adjustable margin ΔP upper =P max -P unit Static downward adjustable margin ΔPl ower =P unit -P min The smaller of the two values ​​is taken as the static response margin ΔP. unit =min(ΔP upper ΔP lower ); Calculate the dynamic response margin ΔP within the control period Δt. dyn =R limit ×Δt; Static response margin ΔP unit and dynamic response margin ΔP dyn As a constraint, it is used for the dynamic determination of the power allocation coefficient in step three.

3. The method according to claim 1, characterized in that, The process of dynamically determining the power allocation coefficient in step three adopts an allocation function based on fuzzy logic or an adaptive weighting function. The input variables of this allocation function include at least: the real-time AGC response margin ΔP of the thermal power unit. unit Total adjustable power capacity P of the flywheel array fly_total Average State of Charge (SOC) of Flywheel Array avg And the rate of change of grid frequency df / dt.

4. The method according to claim 1, characterized in that, In step four, the flywheel units with a state of charge within a preset ideal range are prioritized. This means that the ideal working range of the state of charge is set to 40% to 60%. When the flywheel array needs to discharge, the discharge power is preferentially allocated to the flywheel units with a state of charge higher than 40% and not exceeding the limit. When charging is required, the charging power is preferentially allocated to the flywheel units with a state of charge lower than 60% and not exceeding the limit.

5. The method according to claim 1, characterized in that, In step four, dynamically adjusting the power allocation weight of each flywheel unit based on the state of charge balance factor means that when calculating the allocation weight of each flywheel unit, a balance compensation amount that is positively correlated with the degree of deviation of the current state of charge from the average state of charge of the array is introduced, so that the allocation result tends to prioritize the discharge of units with a high state of charge and prioritize the charging of units with a low state of charge, so as to gradually realize that the state of charge of each unit in the array tends to be consistent.

6. The method according to claim 1, characterized in that, Step four also includes a response speed matching strategy, specifically: prioritizing the allocation of high-frequency, small-amplitude power fluctuation components to flywheel units in the flywheel array that currently have higher rotational speeds and faster dynamic response speeds.

7. The method according to claim 1, characterized in that, The multi-timescale collaborative control architecture in step five is specifically implemented as follows: In the first-level control, the flywheel energy storage array monitors the grid frequency deviation. When the frequency deviation exceeds the preset primary frequency regulation dead zone, the flywheel array responds autonomously and outputs power within 50 milliseconds to provide primary frequency regulation support. In the second layer of control, the AGC control cycle is used as the execution cycle. The AGC scheduling command is received and the dynamic power allocation described in step three is executed. The flywheel array undertakes the rapidly changing part of the power deviation, and the thermal power unit gradually takes over and undertakes the trend-changing part of the power. In the third layer of control, when the average state of charge of the flywheel array deviates from the preset ideal range or the state of charge of a single flywheel unit exceeds the limit, the output of the thermal power unit is adjusted to provide reverse power support, so that the state of charge of the flywheel array gradually returns to the ideal range.

8. The method according to claim 1, characterized in that, The automatic increase of flywheel array output power for dynamic compensation in step six specifically includes: calculating the difference between the actual output change rate of the thermal power unit and the command change rate, multiplying the difference by a preset compensation coefficient and adding it to the target power command of the flywheel array, so that the flywheel array can quickly make up for the power gap caused by the response lag of the thermal power unit.

9. The method according to claim 1, characterized in that, The comprehensive frequency modulation performance evaluation index K in step seven p The calculation method is as follows: K p =w1×K delay +w2×K accuracy +w3×K correlation ; Among them, K delay To respond to latency metrics, K accuracy To adjust the accuracy index, K correlation For relevance indicators, w1, w2, and w3 are weighting coefficients set according to the power grid dispatch assessment rules.

10. The method according to claim 9, characterized in that, The adaptive optimization adjustment in step seven specifically includes: when the comprehensive frequency modulation performance evaluation index K... p When the value is lower than the preset target value, according to K delay K accuracy K correlation Each deviation direction is optimized separately, if K delay If K is too low, the power distribution ratio of the flywheel array will increase in the initial stage of the response. accuracy If K is too low, adjust the power distribution filter parameters or the state-of-charge balancing factor. correlation If the value is too low, the moving average window width in multi-timescale collaborative control will be optimized.

Images