A multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system

Abstract

The application discloses a primary frequency modulation active power distribution multi-objective optimization method of an energy storage system. The method comprises the following steps: calculating the total active power output required for the energy storage system to participate in frequency modulation; obtaining the active power efficiency of the power supply part and the PCS part according to the circuit of the sub-energy storage system, so as to form a first objective function for minimizing the active power loss; forming a second objective function based on the deviation value of the SOC of the sub-energy storage system; the first objective function and the second objective function form the multi-objective function of the present optimization; establishing an optimization constraint, solving the multi-objective function of the present optimization to obtain the Pareto frontier of the active power output distribution; normalizing the objective function value of the obtained Pareto frontier through a fuzzy membership function, and selecting the optimal compromise solution from the Pareto frontier through weighting. The application is helpful to improve the economy of the energy storage power station participating in frequency modulation, reduce the SOC deviation, and prolong the service life of the energy storage power station.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power system load distribution, and particularly relates to a multi-objective optimization method for active power distribution of primary frequency modulation of an energy storage system. BACKGROUND

[0002] Under the background of global fossil energy gradually drying up and the greenhouse effect becoming increasingly serious, the wave of clean energy is rising worldwide. The energy structure is changing from mainly relying on fossil energy to mainly relying on renewable energy such as wind energy and solar energy. As an important part of the energy structure, the promotion of clean electricity will effectively promote the transformation of the overall energy structure. In recent years, the installed capacity of new energy power generation such as wind power and photovoltaic power has increased rapidly.

[0003] However, due to the volatility and randomness of wind energy and solar energy, there is uncertainty in the frequency modulation performance of wind farms and photovoltaic farms, which may affect the frequency modulation effect. After large-scale access of wind power and photovoltaic power generation, the frequency modulation capability and anti-interference capability of the power system may decrease, and the power system with a high proportion of renewable energy penetration will face serious frequency modulation challenges, and new control means are urgently needed. Battery energy storage has strong load change tracking capability, fast response speed, accurate output control, and bidirectional regulation capability, and shows great potential in the field of frequency modulation.

[0004] In view of this challenge, scholars have proposed different control strategies. For example, KIM W W et al. proposed a control strategy for distributing adjustment instructions among multiple energy storage systems, which considers the current state and future available capacity of the energy storage system to respond to the adjustment instructions as much as possible while ensuring the utilization rate of the energy storage system. However, it only considers the overall energy storage power station and does not consider the SOC and efficiency of the sub-energy storage systems within the energy storage power station. Yan G G et al. proposed a power distribution strategy for battery sub-energy storage system groups participating in the secondary frequency modulation of the power system, studied the relationship between the charging and discharging power of the sub-energy storage system and the charging and discharging efficiency of the battery and the PCS efficiency, optimized the number and power of the sub-energy storage systems participating in the frequency modulation response, and reduced power loss and improved operating efficiency. However, this method only considers the efficiency optimization of the sub-energy storage system and does not consider reducing the SOC deviation of each sub-energy storage system, which may reduce the service life of the energy storage system to some extent. SUMMARY

[0005] In order to overcome the problems in the prior art, the present application provides a multi-objective optimization method for active power distribution of primary frequency modulation of an energy storage system. The present application constructs an optimization objective function with the minimum frequency modulation active power loss and the minimum SOC deviation of the sub-energy storage systems of the energy storage power station, and then selects the optimal compromise solution by weight analysis of the obtained Pareto frontier, which can improve the economy of the energy storage power station participating in the primary frequency modulation of the power grid and fully consider the influence of SOC on the service life of the energy storage power station to improve the service life.

[0006] The application is realized by the technical scheme as follows:

[0007] A multi-objective optimization method for active power distribution of primary frequency modulation of an energy storage system, comprising:

[0008] Step one: monitor the frequency of the power grid to which the current energy storage system is connected, and obtain the total active power output required for the energy storage system to participate in frequency modulation based on the current frequency through a primary frequency modulation formula;

[0009] Step two: the energy storage system comprises a plurality of sub-energy storage systems, and each sub-energy storage system comprises a power supply part and a PCS part; the active power efficiency of the power supply part and the PCS part is obtained according to the circuit of the power supply part and the PCS part, and a first objective function for minimizing the active power loss is formed based on the power supply efficiency and the PCS efficiency;

[0010] Step three: a second objective function is formed based on the deviation value of the SOC of the sub-energy storage system; the first objective function and the second objective function constitute the multi-objective function of this optimization;

[0011] Step four: the SOC, frequency modulation capacity and total active power output required for the energy storage system to participate in frequency modulation are used as optimization constraints to solve the multi-objective function of this optimization, and the Pareto frontier of the optimal active power output distribution of each sub-energy storage system in this frequency modulation is obtained;

[0012] Step five: the first objective function value and the second objective function value of the obtained Pareto frontier are normalized through a fuzzy membership function, the normalized results of the first objective function value and the second objective function value are weighted and summed to form a satisfaction function, and the optimal compromise solution is selected from the Pareto frontier by solving the satisfaction function;

[0013] Step six: the optimal compromise solution is used for primary frequency modulation of the energy storage system.

[0014] The application quantifies the frequency modulation active power loss of the energy storage power station, considers the influence of the SOC of the energy storage power station on the service life, proposes a multi-objective optimization method for active power distribution of primary frequency modulation of an energy storage system, constructs an optimization objective function for minimizing the frequency modulation active power loss and minimizing the SOC deviation of the sub-energy storage system of the energy storage power station, and then selects the optimal compromise solution through weight analysis of the obtained Pareto frontier, which can improve the economy of the energy storage power station participating in the primary frequency modulation of the power grid and fully consider the influence of the SOC on the service life of the energy storage power station to improve the service life. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 It is a flowchart of the multi-objective optimization method for active power distribution of primary frequency modulation of an energy storage system;

[0016] Figure 2 It is a circuit structure diagram of a sub-energy storage system;

[0017] Figure 3 This is a primary frequency modulation active power output diagram of each subsystem of the energy storage system in an embodiment of the present invention;

[0018] Figure 4 This is a graph showing the percentage of primary frequency modulation active power loss to total output of the energy storage system in an embodiment of the present invention;

[0019] Figure 5 1 is a diagram showing the SOC changes of each subsystem of the energy storage system in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be further described and illustrated below in conjunction with specific embodiments. The embodiments are merely illustrative of the present disclosure and do not limit its scope. The technical features of the various embodiments of the present invention may be combined accordingly, provided that there is no conflict between them.

[0021] A multi-objective optimization method for primary frequency regulation active power distribution of energy storage system, such as Figure 1 The specific implementation process is as follows:

[0022] Step 1: The PMU (Phasor Measurement Unit) system monitors the grid frequency in real time and applies the resulting low-frequency signal to the primary frequency modulation formula to derive the total active power required for the energy storage power station to participate in frequency modulation.

[0023] Step 2: According to the circuits of the power supply part and the PCS part of the sub-energy storage system, the power supply efficiency and the PCS efficiency are obtained to form an objective function that minimizes the active power loss.

[0024] Step 3: Considering the impact of the SOC of the energy storage system on the life of the energy storage station, the SOC deviation value is used to form the objective function, and combined with the objective function of step 2 to form the multi-objective function of this optimization. Among them, the SOC, frequency regulation capability, and total active output of the energy storage system constitute the optimization constraints of this optimization.

[0025] Step 4: Use the MOGA algorithm to solve the multi-objective optimization problem and obtain the Pareto frontier of the optimal active output of the sub-energy storage system for this frequency regulation.

[0026] Step 5: Normalize the objective function values ​​obtained on the Pareto front using a fuzzy membership function. Then, consider the frequency modulation demand and the current deviation of the SOC to form the weight value of the objective function. This weighting allows the optimal compromise solution to be selected from the Pareto front.

[0027] Step 6: Use the optimal compromise solution to distribute the active power of the energy storage system's primary frequency regulation.

[0028] Since then, the multi-objective optimization method for primary frequency regulation active power distribution within the energy storage system has been completed.

[0029] The preferred embodiments of the present application are described in detail below.

[0030] The energy storage system described in the present application is connected to the power grid as a new energy unit to supply power to the power grid. The energy storage system comprises a plurality of sub-energy storage systems and a phasor measurement unit (PMU), and the circuit model of the sub-energy storage system is shown in Figure 2 , which comprises a power supply part and a power conversion system (PCS) part. The sub-energy storage system can be a single battery cluster and an upper converter, or a plurality of battery clusters connected in parallel to form a battery stack and an upper converter, or other energy storage facilities.

[0031] In step one, the phasor measurement unit monitors the grid frequency in real time, and the total active power P required for the energy storage system to participate in frequency regulation is obtained by the primary frequency modulation formula sum :

[0032]

[0033] where ΔP is the total active power output required for the energy storage system to participate in frequency regulation, K d is the droop coefficient, f is the current grid frequency, f0 is the standard frequency, which is taken as 50 Hz here, and K i is the inertia coefficient.

[0034] In step two, the active power efficiency of the power supply part and the active power efficiency of the PCS part are obtained according to the circuit of the power supply part and the PCS part of the sub-energy storage system, so as to form an objective function for minimizing the active power loss, which is as follows:

[0035] The measured maximum active power P inmax of the PCS part of the sub-energy storage system is fitted into a function curve P inmax =f(SOC) with the SOC data of the sub-energy storage system, and P inmax (SOC i ) represents the maximum active power of the PCS part corresponding to the SOC i of the i-th sub-energy storage system on the fitted function curve of the sub-energy storage system, and the active power efficiency p pcs (ac power / DC power) of the PCS part of the sub-energy storage system is fitted into a function curve p pcs =f(Pout) with the AC active power Pout of the PCS of the sub-energy storage system, and p pcs,i (Pout i ) represents the active power efficiency value of the PCS part corresponding to Pout i of the i-th sub-energy storage system on the fitted curve of the sub-energy storage system. The active power efficiency p batt,iSOC i , Pout i ):

[0036]

[0037] (U OCV SOC i )-I batt *R)*I batt,i =Pout i / ρ pcs,i (Pout i ) (2)

[0038] where I batt,i is the output current of the power supply end of the i-th sub-energy storage system, R is the equivalent internal resistance of the power supply part, U OCV (SOC i ) is the open circuit voltage of the power supply part of the i-th sub-energy storage system when the remaining power is SOC i .

[0039] The battery efficiency ρ batt of the sub-energy storage module and the converter efficiency ρ pcs are obtained, and a target function that minimizes the active power loss P LOSS can be formed.

[0040]

[0041] f LOSS (SOC i ,Pout i )=Pout i / ρ pcs,i (Pout i ) / ρ batt,i (SOC i ,Pout i )-Pout i (4)

[0042] In step three, the influence of the SOC of the sub-energy storage system of the energy storage power station on the service life of the energy storage power station is considered, and the deviation value SOC bias of the SOC is formed into a target function (see formula 5).

[0043]

[0044] f bias (SOC i )=(SOC i (t)-SOC mean (t-1)) 2 (6)

[0045] Among them, t represents the end time of this frequency modulation control, t-1 represents the start time of this frequency modulation control, SOC mean (t-1) is the average SOC of the sub-energy storage system at time t-1, SOC i (t) is the SOC of the i-th sub-energy storage system at time t, and n is the number of sub-energy storage systems.

[0046] In step 4, the SOC, frequency regulation capability, and total active output of the sub-energy storage system constitute the optimization constraints:

[0047] SOC i (t) = SOC i (t-1)-U OCV (SOC i )*I batt,i *t / E k,i (7)

[0048] SOC min ≤SOC i (t)≤SOC max (8)

[0049] Pout i / ρ pcs,i (Pout i )≤P inmax (SOC i ) (9)

[0050] Pout i ≤P rate,i (10)

[0051]

[0052] Among them, SOC i (t-1) is the initial SOC of the ith sub-storage system in this frequency modulation control, E k,i is the capacity of the i-th sub-storage system, SOC i (t) is the SOC of the i-th sub-storage system at the end of this control cycle, which must be within the SOC limit. min and SOC max is the minimum SOC and maximum SOC of the sub-energy storage system. inmax (SOC i ) is the maximum active power of the PCS DC terminal of the aforementioned sub-energy storage system. , P rate,i is the rated active power of the i-th sub-energy storage system.

[0053] The MOGA algorithm is used to solve the multi-objective optimization problem, and the optimal active power output of the sub energy storage system of this frequency modulation is obtained. The solution in the Pareto frontier is the active power output of each sub energy storage system, and the value of the i-th objective function is obtained by substituting the solution of the Pareto frontier into the i-th objective function i , f i is normalized and weighted (see formula 12), and the satisfaction degree y of each optimization solution is obtained. At this time, the optimization solution with the highest satisfaction degree is the optimal compromise solution of the Pareto frontier.

[0054] y = w1*f'1 + w2*f'2 (12)

[0055] Where w1 and w2 are the inertia weights of the importance of frequency modulation active loss and SOC deviation, both of which range from 0.1 to 1 in this embodiment, and f'1 and f'2 are the normalized results of the optimal values of the two objective functions. The normalization method uses fuzzy membership function (see formula 13), f i max , f i min are the maximum and minimum values of the i-th objective function, respectively.

[0056]

[0057] The value of w1 is related to the demand of this frequency modulation, and the greater the total active power required by the energy storage, the greater w1. The value of w2 is related to the SOC deviation value of each sub energy storage system during this frequency modulation, and the greater the deviation value, the greater w2. The optimal value of the weight is calculated according to the following formula:

[0058]

[0059] Where P rate is the rated active power of the energy storage power station as a whole; a is the target weight definition value, when the total active power required by the energy storage P sum is less than a, it is considered that the active power of this frequency modulation is small, and the active loss is small, so the objective function does not need to be emphasized, otherwise it is considered that the active power of this frequency modulation is large, and the active loss is high, and the objective function is increasingly emphasized with the increase of P sum ; SOC BIAS (t) is the sum of the difference between the SOC of each sub energy storage system in the energy storage power station at the last time and the average SOC of the sub energy storage system at time t, SOC i (t-Δt) is the SOC of the i-th sub energy storage system at t-Δt, which is the last time of time t, and SOC mean (t-Δt) is the average SOC of all sub energy storage systems at t-Δt, and n is the number of sub energy storage systems in the energy storage power station; SOC BIASMAX is the SOCBIAS The theoretical maximum value of (t).

[0060] In this experimental example, the frequency regulation control cycle is 12 seconds. The energy storage station consists of three sub-storage systems with initial SOCs of 0.8, 0.65, and 0.5, respectively. Each sub-storage system has a rated power of 30 MW and a capacity of 3 MW·h.

[0061] The results of this experiment can be seen in the attached figure. Figure 3 This is a primary frequency modulation active power output diagram of each subsystem of the energy storage system in an embodiment of the present invention. Figure 4 This is a graph showing the percentage of primary frequency modulation active power loss to total output of the energy storage system in an embodiment of the present invention. Figure 5 This is a diagram showing the SOC changes of each subsystem of the energy storage system according to the embodiment of the present invention. Figures 3-5 It can be seen that when active power demand is low, this method tends to reduce the SOC deviation of each sub-storage system, and the sub-system with a higher SOC discharges more heavily, so the three sub-systems alternately output large amounts of power. However, when active power demand is high, this method tends to reduce the active power loss of each sub-storage system. It can be seen that the percentage of active power loss to total power output is much smaller than when active power demand is low.

[0062] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system, characterized in that: include: Step 1: Monitor the frequency of the grid to which the energy storage system is currently connected, and use the primary frequency regulation formula based on the current frequency to determine the total active power output required for the energy storage system to participate in frequency regulation. Step 2: The energy storage system includes multiple sub-energy storage systems, each of which includes a power supply part and a PCS part; the active efficiency of the power supply part and the PCS part is obtained based on the circuits of the power supply part and the PCS part of the sub-energy storage system, and a first objective function is formed based on the power supply efficiency and the PCS efficiency to minimize the active power loss; Step 3: A second objective function is formed based on the deviation value of the SOC of the sub-energy storage system; the first objective function and the second objective function constitute a multi-objective function of this optimization; Step 4: Using the SOC, frequency regulation capability, and total active power output required for the energy storage system to participate in frequency regulation as optimization constraints, solve the multi-objective function for this optimization and obtain the Pareto frontier of the optimal active power output allocation for each sub-energy storage system in this frequency regulation. Step 5: Normalize the first and second objective function values ​​obtained on the Pareto front using the fuzzy membership function, and take the weighted sum of the normalized results of the first and second objective function values ​​to form a satisfaction function. Select the optimal compromise solution from the Pareto front by solving the satisfaction function. Step 6: Use the optimal compromise solution to distribute the active power of the primary frequency regulation of the energy storage system.

2. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 1, characterized in that: In step 1, the obtained low-frequency signal is applied to the primary frequency modulation formula to obtain the total active output required by the energy storage system to participate in frequency modulation. The specific calculation formula is: Where ΔP is the total active power output required by the energy storage system to participate in frequency regulation, K d is the droop coefficient, f is the current grid frequency, f0 is the standard frequency, K i is the coefficient of inertia.

3. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 1, characterized in that: In step 2, the active efficiency of the power supply part and the PCS part is obtained based on the circuits of the power supply part and the PCS part of the energy storage sub-system, specifically: The circuits of the power supply and PCS parts of the sub-energy storage system are tested, and the active efficiency ρ of the PCS part is obtained based on the measured active power at the AC end and the active power at the DC end of the PCS part. pcs , the active efficiency P of the PCS part pcs The active power Pout of the AC end of the PCS is fitted into a function curve, and the active efficiency ρ of the power supply part of the i-th sub-energy storage system is calculated by the following formula: batt，i (SOC i , Pout i ): (U OCV (SOC i )-I batt *R)*I batt，i =Pout i / ρ pcs,i (Pout i ) Among them, I batt，i is the output current of the power supply end of the i-th sub-energy storage system, R is the equivalent internal resistance of the power supply part, U OCV (SOC i ) is the open circuit voltage of the power supply part of the i-th sub-energy storage system, Pout i is the AC active power of the PCS part of the i-th sub-energy storage system, ρ pcs,i (Pout i ) is the Pout on the fitting function curve of the i-th sub-energy storage system i The corresponding active efficiency value of the PCS part, U OCV (SOC i ) is the power supply of the i-th sub-energy storage system when the remaining power is SOC i The open circuit voltage at .

4. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 3, characterized in that: In step 2, the first objective function that minimizes active power loss is formed based on the power supply efficiency and PCS efficiency, and its expression is: f LOSS (SOC i ,Pout i )=Pout i / ρ pcs,i (Pout i ) / ρ batt，i (SOC i ,Pout i )-Pout i Among them, P LOSS Active power loss.

5. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 1, characterized in that: In step 3, the deviation value based on the SOC of the sub-energy storage system constitutes a second objective function, which is expressed as follows: f bias (SOC i )=(SOC i (t)-SOC mean (t-1)) 2 Among them, SOC bias is the deviation value of the SOC of the sub-energy storage system, SOC mean (t-1) is the average SOC value of each sub-storage system at time t-1, SOC i (t) is the SOC of the i-th sub-energy storage system at time t, and n is the number of sub-energy storage systems.

6. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 4, characterized in that: In step 4, the SOC of the sub-storage system, the frequency regulation capability, and the total active power output required for the energy storage system to participate in frequency regulation are used as optimization constraints. The expression is: SOC i (t)=SOC i (t-1)-U OCV (SOC i )*I batt，i *t / E k，i SOC min ≤SOC i (t)≤SOC max Pout i / ρ pcs,i (Pout i )≤P inmax (SOC i ) Pout i ≤P rate，i Among them, SOC i (t-1) is the initial SOC of the ith sub-storage system in this frequency modulation, E k，i is the capacity of the i-th sub-storage system, SOC i (t) is the SOC of the i-th sub-storage system at the end of this frequency modulation cycle, SOC min and SOC max is the minimum SOC and maximum SOC of the sub-storage system, P inmax (SOC i ) is the remaining capacity of the i-th sub-storage system SOC i The maximum active power of the PCS DC terminal at the time of rate，i is the rated active power of the i-th sub-energy storage system, P sum The total active power output required for the energy storage system to participate in frequency regulation.

7. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 1, characterized in that: In step 4, the MOGA algorithm is used to solve the multi-objective function of this optimization.

8. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 1, characterized in that: In step 5, the first objective function value and the second objective function value of the obtained Pareto front are normalized by the fuzzy membership function, and the calculation formula is: Among them, f i ′ is the normalized result of the i-th objective function value, f i is the i-th objective function value on the Pareto front, f i max and f i min is the maximum and minimum value of the i-th objective function.

9. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 8, characterized in that: In step 5, the normalized results of the first objective function value and the second objective function value are weighted and summed to form a satisfaction function, which is expressed as follows: y=w1*f′1+w2*f′2 Among them, w1 and w2 are weights, and y is satisfaction; The method of selecting the optimal compromise solution from the Pareto front by solving the satisfaction function specifically involves solving the satisfaction function so that the solution with the highest satisfaction is the optimal compromise solution in the Pareto front.

10. The multi-objective optimization method for primary frequency modulation active power distribution of an energy storage system according to claim 9, characterized in that: The values ​​of weights w1 and w2 are calculated according to the following formula: Among them, P rate is the total rated active output of the energy storage power station, a is the target weight limit value, SOC i (t-Δt) is the SOC of the i-th sub-storage system at time t-Δt, SOC mean (t-Δt) is the mean SOC of all sub-storage systems at time t-Δt, n is the number of sub-storage systems in the energy storage power station; SOC BIASMAx For SOC VIAS The theoretical maximum value of (t).

Images