CN116260164A

CN116260164A - Power distribution method based on hybrid energy storage system and application thereof

Info

Publication number: CN116260164A
Application number: CN202211473300.2A
Authority: CN
Inventors: 祇会强; 张敏; 张世锋; 李冉; 宋金鸽; 肖莹; 李�瑞; 常潇
Original assignee: State Grid Electric Power Research Institute Of Sepc
Current assignee: State Grid Electric Power Research Institute Of Sepc
Priority date: 2022-11-21
Filing date: 2022-11-21
Publication date: 2023-06-13

Abstract

A power distribution method based on a hybrid energy storage system, comprising the steps of: CEEMD decomposition of wind power generation data; the energy entropy of the intrinsic mode function is calculated; power allocation based on fuzzy control optimization. The invention provides a method for decomposing wind-solar power generation power fluctuation based on a CEEMD method, and optimizing energy storage configuration combination of power distribution after finishing primary distribution of energy entropy by using a fuzzy control method. The invention is helpful to ensure the precision of power distribution and further improve the efficiency of stabilizing and decomposing the fluctuation power. The protection point is to make the fluctuation range of the super capacitor and the storage battery SOC lower than 10% and the electric quantity always maintain about 50% of the battery SOC, so that the two energy storage devices can exert the maximum effect, and the overall efficiency and the service life of the hybrid energy storage system are improved.

Description

Power distribution method based on hybrid energy storage system and application thereof

Technical Field

The present invention relates to a power distribution method and an application thereof, and more particularly, to a power distribution method based on a hybrid energy storage system and an application thereof.

Background

In the current large environment that electricity utilization is increasingly intense and electricity limiting news is not fresh, in order to improve the service life and the operation efficiency of various energy storage devices, an efficient and accurate power distribution method is required to be provided, power fluctuation generated by wind and light power generation is stabilized, and stable and continuous operation of a power grid is maintained.

Complementary ensemble empirical mode decomposition (Complementary Ensemb le Empi r ica l Mode Decomposition, CEEMD) is a method for processing adaptive signals that is derived from a gradual derivation of optimization from the ensemble empirical mode decomposition (Ensemble Empi r ical Mode Decompos it ion, EEMD) method. CEEMD is based on EEMD, the step of adding white noise in the EEMD method is converted into two signals of adding white noise to the original signal and subtracting white noise from the original signal, and the two signals are subjected to EMD averaging at the same time, so that white noise residues in the reconstructed signal after decomposition by the original EEMD method are reduced, the rate of decomposition iteration is accelerated, the total average frequency is reduced, and the integral average frequency is reduced from hundreds of EEMD to tens of CEEMD. In general, the larger the lumped average number, the smaller the signal noise after reconstruction. Therefore, the self-adaptive decomposition of the fluctuating power based on the CEEMD method is more efficient and reliable in the decomposition process.

In wind-solar power generation systems, high-frequency band fluctuation and low-frequency band fluctuation are the main types of power fluctuation caused by unstable weather conditions. Two types of energy storage devices are associated to absorb or release energy. The power density of the power type energy storage device is high, but the energy density is low, such as a super capacitor, and the charging and discharging frequency is high; in contrast to supercapacitors, energy storage devices have low power density but high energy density, and batteries are typically energy-type devices with lower charge and discharge frequencies than supercapacitors. Therefore, the combination of the two methods is planned in an overall way, the advantages and the characteristics of the two methods are utilized to the greatest extent, and the method has an important pushing effect on the absorption stabilization of power fluctuation.

In the prior art, wind power fluctuation is stabilized by using an EEMD decomposition method, such as Guo Lingjuan, wei and Han Xiaoqing, information entropy and energy entropy data are analyzed, and fluctuation power is distributed and corrected; li Dayong, which are bright and based on EEMD method, the primary distribution order k is obtained by adopting standardized modulus cumulative average method for the inherent mode function (I ntr i ns ic Mode Funct ion, IMF) ₁ Then adaptively find the secondary distribution power command k ₂ The method comprises the steps of carrying out a first treatment on the surface of the According to the parameter requirements on power quota and time response speed in relevant national standard documents of China, the EEMD method is utilized to decompose the original wind power data, so that the energy accumulator executes relevant instruction actions under a specific scene; li Yang proposes an adaptive droop factor improvement method based on inductor current, which stabilizes the bus voltage while adjusting the current voltage, thereby improving the power distribution accuracy. However, in the above-mentioned prior art, the process of using the EEMD method mostly uses the EEMD to decompose the high-frequency and low-frequency power signal, to complete the primary distribution of the power, and then uses the fuzzy control to optimize and restrict the SOC condition of the energy storage element, so as to finally obtain the secondary distribution result of the power.

Disclosure of Invention

The invention aims to solve the problems of decomposing wind-solar power generation power fluctuation by using a CEEMD method and improving power distribution configuration on the basis of one-time distribution, and the final aim is to enable an energy storage system to continuously and efficiently work and maintain stable operation; the invention aims to decompose wind-solar power generation power fluctuation by using a CEEMD method, calculate an intrinsic mode function (I MF) and a margin thereof, perform primary distribution of power according to energy entropy, set limiting conditions for SOC of an energy storage device based on a fuzzy control method, and perform secondary distribution and optimization, thereby achieving the target of optimization, and the technical scheme is as follows:

a power distribution method based on a hybrid energy storage system is characterized by comprising the following steps: the method comprises the following steps:

step one: CEEMD decomposition of wind power generation data;

step two: the energy entropy of the intrinsic mode function is calculated;

step three: power allocation based on fuzzy control optimization.

Advantageous effects

After wind-solar power generation power fluctuation is decomposed by using a CEEMD method and related secondary power distribution of the energy storage device is completed based on a fuzzy control method, the SOC state of the energy storage device can be maintained within a reasonable interval, and the power fluctuation is reduced, so that positive influence is generated on fluctuation stabilization of the whole wind power generation system, and the normal working output requirement of the wind power system is met.

Drawings

FIG. 1 is a CEEMD algorithm step diagram;

FIG. 2 is an input supercapacitor SOC;

FIG. 3 shows the input super capacitor ΔSOC _SC A schematic diagram;

FIG. 4 is a diagram of a fuzzy control input and output membership function of a supercapacitor;

FIG. 5 is a schematic diagram of the energy entropy of each IMF component;

FIG. 6 is a schematic diagram of the energy entropy difference of each IMF component;

FIG. 7 is a graph of SOC before and after supercapacitor and battery adjustment;

fig. 8 is a schematic diagram of supercapacitor and battery power distribution.

Detailed Description

step one: CEEMD decomposition of wind power generation data (see fig. 1);

firstly, the method is improved on the EEMD method, and because the EEMD method which is most widely applied at present refers to an integrated empirical mode decomposition method, the decomposition principle is that when additional white noise is uniformly distributed in the whole time-frequency space, the time-frequency space is composed of different scale components divided by a filter bank, and the time-frequency space can be gradually decomposed; the step is improved on the above known EEMD method, and based on multiple empirical mode decomposition of superimposed Gaussian white noise, the single-direction white noise signal is converted into a pair of adaptive white noise signals with positive and negative characteristics, so we can assume that the original signal is x ₀ (t), then adding bidirectional self-adaptive white noise signal as + -f (t), obtaining two new signals X ₁ (t) and X ₂ (t):

The above procedure is to add positive and negative noise in pairs once, but in practical application, the average number NE is set in the program first in the general procedure, and can be understood as the number of additions, assuming ne=n, i.e. n positive and negative white noise±f (t) are added. After adding n white noise times, carrying out EMD decomposition on the n pairs of new signal sources, and obtaining corresponding calculation results by using the following calculation formula:

IMF in _i Is X ₁ The ith IMF component after decomposition, IMF _j Is X ₂ The ith IMF component after the decomposition; delta (t) and lambda (t) are decomposition residues, namely EMD decomposition results of one pair in the algorithm, and n pairs of similar decomposition results are assumed to be shared in the whole process; next, if a final decomposition result is required, first:

the final decomposition results were as follows:

wherein:

the q-th IMF variable is decomposed after averaging;

the remainder after the total average value is calculated; after n times of decomposition according to the calculation, the final decomposition allowance a _i (t) and the remainder σ (t) are as follows:

step two: the energy entropy of the intrinsic mode function is calculated;

entropy is a measure of the degree of confusion of a substance within a system and is a measure of the state of a substance system. The intrinsic mode function IMF also contains a certain amount of energy, so that the total amount of energy of each IMF function can be considered to be measured by using the energy entropy. And combining the energy entropy value, and taking the maximum energy entropy difference value as a demarcation point, namely one-time distribution of power. Assuming that the energy entropy of each IMF function is E _i The total energy is E, and the remainder energy is generally not taken into account. The respective calculation formulas are as follows

Wherein p is _i Representing each ofEnergy entropy E of modal function _i Accounting for the specific gravity of the total energy E. According to the calculated energy entropy of each IMF component, calculating the energy entropy difference value between every two adjacent components, taking the interval with the largest difference value as a boundary for distinguishing a high frequency band from a low frequency band, and setting the boundary point k for one-time distribution of power after checking.

After the k value of the demarcation point is determined, an IMF component before k is used as a high-frequency component, and energy can be absorbed or released by using an energy storage device super capacitor with higher power because of higher frequency; otherwise, all IMF components after k are absorbed or released as low frequency components by the battery.

P in the formula _sc Is the power that the super capacitor needs to absorb or release; p (P) _bat Is the power that the battery needs to absorb or release; epsilon _e Is the margin of IMF; n is the total number of IMF components.

Step three: power allocation based on fuzzy control optimization.

After the above steps are completed, the safety range of the SOC value of the whole energy storage system is mainly considered, and in short, the SOC of the storage battery and the supercapacitor must be ensured to be within the safety range at the same time, which is taken as a constraint condition of the energy storage system. By the following formula, the super-capacitance state of charge change ΔSOC will be calculated first _sc The calculation formula is as follows

P in formula (10) regardless of the self-discharge rate of the two energy storage batteries _sc (t) a high-frequency power instruction of the super capacitor at the moment t; η (eta) _c1 And eta _d1 The charging efficiency and the discharging efficiency of the super capacitor are respectively; e (E) _sc And rated working capacity value of the super capacitor. Can be read by the above calculationThe charge state of the super capacitor in the operation process is calculated according to P _sc The positive and negative of (t) and the SOC state of the super capacitor, and the charge and discharge instruction parameters are properly adjusted according to specific conditions.

The input parameters of the fuzzy controller comprise the current state of charge (SOC) value of the super capacitor and delta SOC calculated according to the formula (10) _sc Is a value of (2). The output of the fuzzy control is the regulating coefficient K of the high-frequency power corresponding to the super capacitor _p ，(0<K _p <1). Then (1-K) _p )P _sc And (t) is the power that the storage battery needs to bear additionally.

Then, we introduce the calculation formula of the state of charge variation of the storage battery

Similarly, P in _bat (t) a high-frequency power command of the battery at time t; η (eta) _c2 And eta _d2 The charging efficiency and the discharging efficiency of the storage battery are respectively; e (E) _bat The working capacity of the super capacitor. The super capacitor reads the charge state in the running process and according to P _bat The positive and negative of (t) and the state of charge of the battery, and the charge-discharge process of the battery are appropriately adjusted.

A fuzzy controller is created, the input parameters of which are the current state of charge (SOC) value of the storage battery and delta SOC calculated according to formula (11) _bat Is a value of (2). The output is the regulating coefficient K of the accumulator corresponding to the low-frequency power _b (0<K _b <1)。

Finally, P is _bat (t)K _b And P _bat (t)K _b +(1-K _p )P _sc (t) comparing the two, if the two differ by not more than + -2%, then the two are considered to be the same, and the adjustment coefficient is K _p Or K _b Can be used; if the difference between the two is within the range of (+/-) (2% -5%), the average value of the two can be taken as the corresponding low-frequency power of the storage battery; if the two are more than + -5% different, then the energy signal needs to be decomposed again to set new parameters, as shown in Table 1.

TABLE 1 super capacitor Power adjustment scheme

And respectively drawing input and output membership function diagrams of the fuzzy controller corresponding to the supercapacitor according to the calculated data, as shown in figure 2. The common fuzzy variable language values for fuzzy control are as follows: NB (Negative Big):negative big, NM (Negative Medium):negative medium, NS (Negative Small):negative small, ZO (AImost Zero):few zero, PS (Positive Small):positive small, PM (Positive Medium):positive, PB (Positive Big):positive big.

And respectively drawing input and output membership functions of the fuzzy controller corresponding to the supercapacitor, as shown in figures 2, 3 and 4. The fuzzy rule control table is shown in table 2, and the fuzzy control method can reduce the charge and discharge cycle times of the storage battery as much as possible, thereby prolonging the service life of the storage battery, simultaneously playing the role of the super capacitor to the greatest extent and sharing the pressure of the storage battery for adjusting the charge and discharge power.

Table 2 fuzzy rule control table

Will P _bat (t)K _b And P _bat (t)K _b +(1-K _p )P _sc (t) comparing the two, if the two differ by not more than + -2%, then the two are considered to be the same, and the adjustment coefficient is K _p Or K _b Can be used.

If the difference between the two is within the range of (+/-) (2% -5%), the average value of the two can be taken as the corresponding low-frequency power of the storage battery:

in summary, the whole process is as follows:

firstly, the system carries out CEEMD decomposition on input wind power generation data to obtain a group of IMF components and remainder, then calculates the energy entropy value of each IMF component and the difference between the energy entropy values by utilizing the energy entropy principle, finds out two IMF components with the largest difference, takes the IMF components as primary power distribution points, and divides the wind power generation into two groups of high-frequency energy and low-frequency energy. Then, the two sets of high-frequency energy and low-frequency energy are respectively calculated as delta SOC _sc And delta SOC _bat Respectively inputting the two fuzzy controllers into a fuzzy control optimization method to perform reasoning operation to obtain two adjustment parameters K _p (0<K _p <1) And K _b (0<K _b <1), the comparison method is used preferentially to obtain final distribution adjustment parameters, and the equation (12) and the equation (13) are used for carrying out secondary power distribution of the super capacitor and the storage battery.

Examples

Step one: acquisition and CEEMD decomposition of wind power generation data

The invention adopts the original data of a certain wind power plant, the installed capacity of the wind power plant is 66MW, the sampling interval is 15 minutes, 96 points are sampled every day, and 100 days of data are called by using a data mining method. According to the formula, after n times of decomposition, the final decomposition allowance and residual are calculated, and an IMF component curve and residual curve diagram of each order are drawn according to a CEEMD algorithm.

Step two: intrinsic mode function energy entropy calculation

After the drawn IMF component curves of all orders are drawn, the energy entropy value p of the IMF functions of all orders is solved according to the formula, and the energy entropy difference value between the IMF functions of all orders can be obtained through subtraction calculation, wherein the energy entropy of each IMF component and the energy entropy difference value of each IMF component are respectively shown in fig. 5 and 6.

From the above graph, it can be seen that, without considering the remainder, the difference between IMF4 and IMF5 is the largest among IMF components of each order, and can be used as a demarcation point to distinguish between high and low frequency power, where the power demarcation point takes k=5. Namely

Step three: power allocation based on fuzzy control optimization.

Next, simulation modeling is performed on the hybrid energy storage system, whether the fluctuation variation range of the super capacitor and the storage battery SOC meets the limiting condition or not is further verified, and the specific initial parameter configuration of the whole hybrid energy storage system is shown in table 3:

table 3 hybrid energy storage device parameters

Parameters (parameters)	Super capacitor	Storage battery
			Rated capacity/(MW. H)	5.9	6.6
SOC range/%	25-85	25-80
			Initial SOC/%	50	50
Rated charging power/kW	4450	1620
			Rated discharge power/kW	4555	2875
Charge-discharge efficiency/%	98	98

According to the fuzzy control rule and the parameter configuration specification of table 3, SOC adjustment curves of the two energy storage devices before and after charging and discharging can be obtained, as shown in fig. 7. And drawing an original power data graph and the height.

Meanwhile, an original power data diagram and a schematic diagram of the distribution of high-frequency power and low-frequency power between the super capacitor and the storage battery are drawn, as shown in fig. 8.

According to the SOC curves of fig. 7, after the power distribution adjustment of the system is completed, the SOC state curves of the two energy storage devices are optimized to a certain extent, and the SOC state peaks and valleys of the two energy storage devices are reduced to be more stable. Specifically, the change of the SOC of the storage battery is slow and stable near 0.58, and the whole is stable, so that the constraint condition can be met; compared with the prior art, the super capacitor has obvious change, and is stable near 0.40 in the tail period, and the integral curve is near 0.50, so that the super capacitor is mainly responsible for coping with sudden power change in the later period, can provide high-frequency power for the outside, and enables the state of the energy storage system to fluctuate slightly near an expected value.

In fig. 8, it can be found that the super capacitor has significantly higher charge and discharge times than the storage battery, and mainly bears the high frequency component in the power fluctuation, and the storage battery mainly bears the low frequency component in the power fluctuation. The CEEMD method for power distribution fully plays the characteristics and advantages of two energy storage devices, can finish peak clipping and valley filling, stabilize fluctuation, maintain stable operation of the wind power system and prolong the service life of the whole energy storage system.

Finally, the invention provides a CEEMD method-based method for decomposing wind-solar power generation power fluctuation to obtain an Intrinsic Mode Function (IMF) and a margin, and an energy storage configuration combination of power distribution is optimized after primary distribution of energy entropy is completed by using a fuzzy control method, so that the optimal planning of a system is achieved in a proper SOC range. The invention is helpful to ensure the precision of power distribution and further improve the efficiency of stabilizing and decomposing the fluctuation power. The protection point is to control the SOC of the energy storage device in a certain interval range, the scheme finally enables the fluctuation range of the SOC of the super capacitor and the storage battery to be lower than 10%, and the electric quantity is always maintained at about 50% of the SOC of the battery, so that the two energy storage devices can play the maximum role, the expected target of the prior reservation is achieved, and the overall efficiency and the service life of the hybrid energy storage system are improved.

The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made therein without departing from the spirit and scope of the invention, which is defined by the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims

1. A power distribution method based on a hybrid energy storage system is characterized by comprising the following steps: the method comprises the following steps:

step one: CEEMD decomposition of wind power generation data;

step two: the energy entropy of the intrinsic mode function is calculated;

step three: power allocation based on fuzzy control optimization.

2. The hybrid energy storage system based power distribution method of claim 1, wherein: the step 1 further comprises the following steps:

firstly, the EEMD method which is most widely used at present refers to an ensemble Empirical Mode Decomposition (EMD) method, and the decomposition principle is that when additional white noise is uniformly distributed in the whole time-frequency space, the time-frequency space is composed of different scale components divided by a filter bank, so that the EEMD method can be gradually decomposed; the EEMD method is improved, the only one-way white noise signal is converted into a pair of adaptive white noise signals with positive and negative characteristics on the basis of multiple empirical mode decomposition of superimposed Gaussian white noise, and the original signal is assumed to be x ₀ (t), then adding bidirectional self-adaptive white noise signal as + -f (t), obtaining two new signals X ₁ (t) and X ₂ (t):

Setting an average number NE, and assuming that ne=n, namely adding n times of positive and negative white noise ± f (t), after adding n times of white noise, performing EMD decomposition on the n pairs of new signal sources, and obtaining a corresponding calculation result by using the following calculation formula:

the final decomposition results were as follows:

wherein:

the q-th IMF variable is decomposed after averaging; />

。

3. the hybrid energy storage system based power distribution method of claim 1, wherein: the step 1 further comprises the following steps:

the energy entropy is utilized to measure the total energy of each IMF function, the maximum energy entropy difference value is taken as a demarcation point, namely one-time distribution of power, by combining the energy entropy value, and the energy entropy of each IMF function is assumed to be E respectively _i The total energy is E, and the rest energy is not counted into a formula; the respective calculation formulas are as follows

Wherein p is _i The energy entropy E representing each modal function _i The specific gravity of the total energy E; according to the calculated energy entropy of each IMF component, calculating every two adjacent componentsThe energy entropy difference between the quantities takes the interval with the largest difference as a demarcation for distinguishing the high frequency band from the low frequency band, and after the error is checked, the interval can be set as a demarcation point k for once distribution of power;

after the k value of the demarcation point is determined, the IMF component before k is used as a high-frequency component, and the super capacitor can be used for absorbing or releasing energy because of higher frequency; conversely, all IMF components after k are absorbed or released as low frequency components by the battery:

4. The hybrid energy storage system based power distribution method of claim 1, wherein: the step 1 further comprises the following steps:

calculating the variation delta SOCsc of the charge state of the super capacitor, wherein the calculation formula is as follows

P in formula (10) regardless of the self-discharge rate of the two energy storage batteries _sc (t) a high-frequency power instruction of the super capacitor at the moment t; η (eta) _c1 And eta _d1 The charging efficiency and the discharging efficiency of the super capacitor are respectively; e (E) _sc Rated working capacity value for super capacitor; the charge state of the super capacitor in the running process can be read through the calculation, and the super capacitor is controlled according to P _sc The positive and negative of (t) and the SOC state of the super capacitor, and the charge and discharge instruction parameters are properly adjusted according to specific conditions.

5. The hybrid energy storage system based power distribution method of claim 4, wherein: the step 1 further comprises the following steps:

the input parameters of the fuzzy controller comprise the current state of charge (SOC) value of the super capacitor and delta SOC calculated according to the formula (10) _sc Is a value of (2); the output of the fuzzy control is the regulating coefficient K of the high-frequency power corresponding to the super capacitor _p ，(0<K _p <1). Then (1-K) _p )P _sc (t) is the power that the battery needs to bear additionally;

then, a calculation formula of the state of charge change of the storage battery is introduced

P in the formula _bat (t) a high-frequency power command of the battery at time t; η (eta) _c2 And eta _d2 The charging efficiency and the discharging efficiency of the storage battery are respectively; e (E) _bat The working capacity of the super capacitor; the super capacitor reads the charge state in the running process and according to P _bat The positive and negative of (t) and the SOC state of the storage battery, and properly adjusting the charge and discharge processes of the storage battery; a fuzzy controller is created, the input parameters of which are the current state of charge (SOC) value of the storage battery and delta SOC calculated according to formula (11) _bat Is a value of (2); the output is the regulating coefficient K of the accumulator corresponding to the low-frequency power _b (0<K _b <1) The method comprises the steps of carrying out a first treatment on the surface of the Finally, P is _bat (t)K _b And P _bat (t)K _b +(1-K _p )P _sc (t) comparing the two, if the two differ by not more than + -2%, then the two are considered to be the same, and the adjustment coefficient is K _p Or K _b Can be used;

if the difference between the two is within the range of (+/-) (2% -5%), the average value of the two can be taken as the corresponding low-frequency power of the storage battery;

if the two are more than + -5%, then the energy signal needs to be decomposed again to set new parameters.

6. A non-volatile storage medium, characterized in that the non-volatile storage medium comprises a stored program, wherein the program, when run, controls a device in which the non-volatile storage medium is located to perform the method of any one of claims 1 to 5.

7. An electronic device comprising a processor and a memory; the memory has stored therein computer readable instructions for executing the processor, wherein the computer readable instructions when executed perform the method of any of claims 1 to 5.