CN-122000930-A - Fire-storage joint frequency modulation control method and device based on double closed-loop optimization

Abstract

The invention belongs to the field of electric power, and discloses a fire-storage joint frequency modulation control method and device based on double closed loop optimization, wherein the method comprises the steps of responding to an acquired AGC instruction signal, predicting a load instruction of a future period of a fire-storage joint system through a load prediction model, and constructing the load prediction model based on variation modal decomposition and two-way long-short-term memory network fusion; the method comprises the steps of inputting a predicted load instruction and an actual load instruction into a third-order frequency modulation demand calculation model to calculate the frequency modulation power demand of the fire-storage combined system, carrying out optimal distribution of output between a thermal power unit and an energy storage system on the basis of the frequency modulation power demand by taking comprehensive minimization of the adjustment cost, the frequency modulation deviation and the energy storage life loss of the fire-storage combined system as targets according to a chaotic quantum particle swarm optimization algorithm, evaluating the health of the energy storage system in the fire-storage combined system based on a capacity attenuation mechanism, and adjusting optimization target weight or constraint conditions in the chaotic quantum particle swarm optimization algorithm according to the health.

Inventors

• JI DEHENG
• ZHANG XINSHENG
• ZHANG JIANGFENG
• DING WEICONG
• WANG QI
• ZHENG KEKE
• WANG TIANYU
• DING NING
• ZHAO HONGYU

Assignees

• 国网浙江省电力有限公司电力科学研究院

Dates

Publication Date

20260508

Application Date

20260129

Claims (10)

1. 1. The fire and storage joint frequency modulation control method based on double closed loop optimization is characterized by comprising the following steps of: responding to the collected AGC command signal, predicting a load command of a fire-storage combined system in a future period through a load prediction model, wherein the load prediction model is constructed based on variation modal decomposition and two-way long-short-term memory network fusion; inputting a predicted load instruction and an actual load instruction obtained through prediction into a pre-constructed third-order frequency modulation demand calculation model containing proportional, integral and differential terms, and calculating the frequency modulation power demand of the fire-storage combined system; According to an improved chaotic quantum particle swarm optimization algorithm, with the aim of comprehensively minimizing the adjustment cost, the frequency modulation deviation and the energy storage life loss of the fire-storage combined system, carrying out optimal distribution of the output between the thermal power unit and the energy storage system on the frequency modulation power demand; and evaluating the health degree of an energy storage system in the fire-storage combined system based on a capacity attenuation mechanism, and adjusting an optimization target weight or constraint condition in the chaotic quantum particle swarm optimization algorithm according to the health degree so as to form life protection closed-loop control.
2. 2. The method according to claim 1, wherein the method further comprises: and taking the collected AGC input signal as an effective identifier, and removing abnormal values by adopting a moving average filtering and box diagram method to generate model input data for inputting the load prediction model.
3. 3. The method of claim 1, wherein predicting, by a load prediction model, load instructions for a future period of the combined fire and storage system comprises: decomposing model input data input into the load prediction model into a plurality of eigenmode function components according to a variation mode decomposition algorithm; And independently predicting each eigenmode function component through the two-way long-short term memory network, and reconstructing and superposing prediction results of the components to obtain the prediction load instruction.
4. 4. The method of claim 1, wherein the expression of the third order tone requirement calculation model is: ; Where P req (t) is the FM power demand, P AGCact (t) is the actual load command, P AGCpred （t+t p ) is the predicted load command, 、 、 For the frequency-modulated dynamic coefficient, Is the frequency modulation deviation at the t moment, For the AGC to command a gain factor, In order to predict the load instruction weight coefficient, The inertial time constant is commanded for the AGC, To predict the duration.
5. 5. The method of claim 1, wherein the expression of the optimization objective function of the chaotic quantum particle swarm optimization algorithm is: ; Wherein, the 、 、 To optimize the target coefficient, its initial value is determined by historical data training and adjusted based on the health feedback.
6. 6. The method of claim 5, wherein optimally distributing the output between the thermal power generation unit and the energy storage system for the modulated power demand comprises: Introducing information of the load prediction model to generate a fire-storage joint frequency modulation rolling time domain optimization strategy, wherein the expression is as follows: ; wherein, the thermal power generating unit outputs Output from energy storage system Meeting the constraint condition , In order to predict the duration of the time, Is the maximum climbing rate of the thermal power generating unit, 、 Respectively the maximum value and the minimum value of the stored energy charge and discharge, Is the maximum force of stored energy.
7. 7. The method of claim 1, wherein the estimating the health of the energy storage system in the combined fire and storage system based on the capacity fade mechanism comprises: an energy storage health evaluation model based on a capacity attenuation mechanism is constructed, the health is calculated based on the evaluation model, and the expression of the energy storage health dynamic evaluation model is as follows: ; Wherein, the In order to obtain the number of charge-discharge cycles, In order to account for the loss of capacity, Is the nominal capacity of the battery.
8. 8. Fire and storage joint frequency modulation control device based on double closed loop optimization, which is characterized by comprising: The prediction unit responds to the collected AGC command signal, and predicts a load command of a future period of the fire-storage combined system through a load prediction model, wherein the load prediction model is constructed based on variation modal decomposition and two-way long-short-period memory network fusion; The calculation unit is used for inputting the predicted load instruction and the actual load instruction obtained by prediction into a pre-constructed third-order frequency modulation demand calculation model comprising proportional, integral and differential terms, and calculating the frequency modulation power demand of the fire-storage combined system; The distribution unit is used for optimally distributing the output between the thermal power unit and the energy storage system according to the improved chaotic quantum particle swarm optimization algorithm and with the aim of comprehensively minimizing the adjustment cost, the frequency modulation deviation and the energy storage life loss of the fire-storage combined system; And the adjusting unit is used for evaluating the health degree of the energy storage system in the fire-storage combined system based on a capacity attenuation mechanism and adjusting the optimization target weight or constraint condition in the chaotic quantum particle swarm optimization algorithm according to the health degree so as to form life protection closed-loop control.
9. 9. An electronic device, comprising: A processor; A memory for storing processor-executable instructions; wherein the processor is configured to implement the steps of the method according to any of claims 1-7 by executing the executable instructions.
10. 10. A computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the method of any of claims 1-7.

Description

Fire-storage joint frequency modulation control method and device based on double closed-loop optimization Technical Field The invention belongs to the field of electric power, and particularly relates to a fire-storage combined frequency modulation control method and device based on double closed loop optimization. Background As a traditional frequency modulation main force, the thermal power generating unit has low response speed and limited climbing capacity, and is difficult to meet the rapidly-changing frequency modulation requirement. The energy storage system has millisecond response speed and flexible adjustment capability, but has high independent configuration cost and limited service life. Fire-storage combined frequency modulation becomes an effective means for improving the frequency modulation performance of the system. In the related art, the fire-storage combined frequency modulation control mostly adopts an open-loop or single-closed-loop control strategy, and the energy storage life loss and the prediction precision of the unit load in the frequency modulation process cannot be considered. The traditional frequency modulation method mainly has the problems of insufficient prediction precision, single optimization target, lack of closed-loop life protection, contradiction between response speed and precision and the like. Specifically, the AGC command signal has strong nonlinearity and randomness characteristics, a single prediction model is difficult to accurately capture the change rule, so that the deviation between a pre-allocation strategy and actual demands is larger, most methods only aim at minimum frequency modulation deviation or optimal economical efficiency, energy storage life loss is not comprehensively considered, energy storage capacity is quickly attenuated due to long-term operation, economical efficiency is reduced, the existing methods mainly adopt fixed weights or constraint conditions, a control strategy cannot be dynamically adjusted according to the actual health state of energy storage, the balance of life and economical efficiency is difficult to realize, the thermal power unit is slow in response but good in economical efficiency, energy storage response is quick but high in cost, and the existing allocation strategy is difficult to obtain optimal balance between quick response and long-term economical efficiency. Disclosure of Invention In view of the above, the invention discloses a fire-storage joint frequency modulation control method and device based on double closed loop optimization, which can solve the defects in the related art. In order to achieve the above purpose, the technical scheme disclosed by the invention is as follows: According to a first aspect of the present invention, a fire-storage joint frequency modulation control method based on double closed loop optimization is provided, including: responding to the collected AGC command signal, predicting a load command of a fire-storage combined system in a future period through a load prediction model, wherein the load prediction model is constructed based on variation modal decomposition and two-way long-short-term memory network fusion; inputting a predicted load instruction and an actual load instruction obtained through prediction into a pre-constructed third-order frequency modulation demand calculation model containing proportional, integral and differential terms, and calculating the frequency modulation power demand of the fire-storage combined system; According to an improved chaotic quantum particle swarm optimization algorithm, with the aim of comprehensively minimizing the adjustment cost, the frequency modulation deviation and the energy storage life loss of the fire-storage combined system, carrying out optimal distribution of the output between the thermal power unit and the energy storage system on the frequency modulation power demand; and evaluating the health degree of an energy storage system in the fire-storage combined system based on a capacity attenuation mechanism, and adjusting an optimization target weight or constraint condition in the chaotic quantum particle swarm optimization algorithm according to the health degree so as to form life protection closed-loop control. According to a second aspect of the present invention, there is provided a fire and storage joint frequency modulation control device based on double closed loop optimization, the device comprising: The prediction unit responds to the collected AGC command signal, and predicts a load command of a future period of the fire-storage combined system through a load prediction model, wherein the load prediction model is constructed based on variation modal decomposition and two-way long-short-period memory network fusion; The calculation unit is used for inputting the predicted load instruction and the actual load instruction obtained by prediction into a pre-constructed third-order frequency modulation demand ca