Energy storage cooperative control method for multifunctional energy storage converter

Through the combination of virtual oscillator control based on PSOA algorithm and grid-type energy storage converter, the problem of energy storage system sharing energy storage participation in frequency regulation in microgrid island mode is solved, and better dynamic response and frequency support are achieved , improves the power supply capacity of distributed power supplies.

CN119010128APending Publication Date: 2024-11-22NANJING INST OF TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410914714.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-09
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the microgrid island mode, the shared energy storage system's energy storage system lacks mature strategies in frequency regulation, resulting in difficulty in cooperating with grid-type and network-type equipment, especially due to its heterogeneous frequency response characteristics, The power distribution and system frequency response mechanisms are complex.

Method used

Using a virtual oscillator control method based on PSOA algorithm, combined with the control strategy of network-type energy storage converter, a robust optimization scheduling model for distributed new energy uncertainty is established, and the solution is simplified through KKT conditions, and optimized inertial weight is adopted. and improve particle swarm algorithm to set control parameters and build a multi-time scale shared energy storage energy management system.

Benefits of technology

Improve the power supply capacity of distributed power supply in the grid-connected operation state, achieve better dynamic response and frequency support, avoid active and reactive power measurement, have better dynamic response than traditional sag control, and pass the energy management system Complete automatic power distribution in seconds and frequency support on long-term scales.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119010128A_ABST
    Figure CN119010128A_ABST
Patent Text Reader

Abstract

The invention discloses an energy storage cooperative control method for a multifunctional energy storage converter. The method comprises the following steps: (1) establishing a power distribution network robust optimization scheduling model considering the uncertainty of distributed new energy; (2) carrying out simplified solution by adopting a KKT (Karush-Kuhn-Tucker condition); (3) establishing a virtual oscillator control method based on a PSO algorithm; setting a control parameter by adopting an improved particle swarm algorithm for optimizing an inertia weight and a learning factor; (4) constructing the autonomous frequency support characteristic of the network type energy storage converter and a multi-time scale cooperative control strategy of a shared energy storage energy management system controlled and adjusted by the network type energy storage converter; according to the invention, the power supply capability of the distributed power supply in the grid-connected operation state can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage coordinated control, and in particular to a method for coordinated energy storage control of a multifunctional energy storage converter. Background Art

[0002] When virtual oscillator control is applied to the microgrid island mode, it can achieve faster synchronization speed and better power equal distribution than droop control and virtual synchronous machine control, and the control system is more concise. Virtual oscillator control is more suitable for weak grids because of its simulated droop characteristics and faster response speed.

[0003] Since such energy storage is usually distributed and has control limitations, the concept of shared energy storage has emerged on the power supply side. Although shared energy storage has attracted widespread attention due to its advantages such as improving energy storage utilization and reducing total costs, there is still little research on the participation of shared energy storage in system frequency regulation, and the immature frequency regulation strategy has also hindered the further development of shared energy storage. At present, the application of grid-building technology to energy storage is still in its early stages. Domestic and foreign scholars have proposed a variety of grid-building control strategies. Droop control based on the simulation of the operation mechanism of synchronous generators and virtual synchronous generator control are widely used; in addition, nonlinear control methods such as matching control and virtual oscillator control have also received widespread attention. Grid-building energy storage converters have autonomous frequency support characteristics, while grid-following energy storage converters are regulated by the upper EMS, resulting in certain coordination problems between energy storage EMS and grid-building and grid-following converters. Since grid-following and grid-building devices have heterogeneous frequency response characteristics, when grid-following and grid-building devices participate in system frequency regulation together, the power distribution and system frequency response mechanism will become more complicated. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a method for coordinated control of energy storage of a multifunctional energy storage inverter, which combines the virtual oscillator control based on the PSOA algorithm and the control strategy of the grid-connected energy storage inverter, and can effectively improve the power supply capacity of the distributed power source in the grid-connected operation state.

[0005] Technical solution: A multifunctional energy storage converter energy storage coordinated control method according to the present invention comprises the following steps:

[0006] (1) A robust optimization dispatch model for distribution networks considering the uncertainty of distributed renewable energy was established;

[0007] (2) Use KKT, i.e. Karush-Kuhn-Tucker conditions, to simplify the solution;

[0008] (3) Establish a virtual oscillator control method based on the PSO algorithm; use an improved particle swarm algorithm that optimizes inertia weight and learning factor to adjust the control parameters;

[0009] (4) Construct a multi-time scale collaborative control strategy for a shared energy storage management system with autonomous frequency support characteristics of grid-type energy storage converters and controlled regulation of grid-type energy storage converters.

[0010] Furthermore, step (1) is as follows: First, an uncertainty set is used to describe the uncertainty of wind and solar output:

[0011]

[0012] Among them, P WP To provide new energy, namely wind power and photovoltaic power; is the expected output of new energy, obtained through historical data analysis; λ is a variable that determines the direction of deviation, which takes +1 when the output of new energy is higher than the expected value and takes -1 when it is lower than the expected value; △P WP is the deviation between the expected output of renewable energy and the dispatched output, and its value is determined by formula (2), where σ is the mean square error between the actual value and the expected value of renewable energy output, and α is a conservative estimation parameter.

[0013] Secondly, the goal of the lower-level model problem is to minimize the running cost and the formula is as follows:

[0014]

[0015] Where T is the number of daily dispatch periods; Ngen is the number of traditional small power generators connected to the distribution network; P gen,i,t Dispatching output for traditional units; P buy,t Purchase electricity from the upper grid for the distribution network; is the cost function of the traditional generator set; C buy,t The price of electricity purchased from the upper power grid;

[0016] The goal of establishing the upper model problem is to maximize the penalty cost. The formula is as follows:

[0017]

[0018] Among them, T is the number of daily scheduling periods, N WP is the number of distributed wind farms and photovoltaic power stations connected to the distribution network, △P WP,i,t is the deviation between the expected output of renewable energy and the dispatched output, C i,t This is a penalty electricity price for random deviations in renewable energy output while taking into account the robustness of the algorithm.

[0019] Finally, the objective function of the two-layer scheduling model is as follows:

[0020] f(p)=min[max(f up )+(f down )].

[0021] Furthermore, step (2) is specifically as follows: establishing constraint conditions including: new energy coefficient variable constraint, node power balance constraint, new energy uncertainty coefficient constraint, unit output constraint, traditional unit ramp rate constraint and network power flow constraint; wherein,

[0022] New energy coefficient variable constraints:

[0023]

[0024] Where λ1 and λ2 are Boolean values ​​to satisfy the definition of λ.

[0025] Node power balance constraints:

[0026] P WP,i,t +P gen,i,t +P in,i,t =P LD,i,t +P out,i,t (7)

[0027] Where: P WP,i,t is the renewable energy power generation output of node i at time t; P gen,i,t is the active power output of the traditional unit at node i at time t; P LD,i,t is the active load of node i at time t; P in,i,t With P out,i,t are the active power flowing into node i and the active power flowing out of node i from other lines at time t respectively.

[0028] New energy uncertainty coefficient constraints:

[0029]

[0030] In the formula, t is the conservative coefficient of new energy uncertainty, which is determined according to actual demand. The larger the value, the more robust the scheduling result.

[0031] Unit output constraints:

[0032]

[0033] Where: P max,i With P min,i is the upper and lower limits of the active output of the i-th coal-fired / gas-fired unit; P wp,max,i is the upper limit of the output of the ith typhoon / photovoltaic power station.

[0034] Traditional unit ramp rate constraints:

[0035] P down,i ≤P gen,i,t -P gen,i,t-1 ≤P up,i (10)

[0036] Where: P down,i With P up,i is the ramp rate limit of the i-th traditional unit.

[0037] Network flow constraints:

[0038] -P ij,max ≤P ij,t ≤P ij,max (11)

[0039] Where: P ij,t P is the power flow from node i to node j. It is defined that the flow from i to j is positive, and vice versa is negative. ij,max is the maximum capacity of line ij.

[0040] Further, step (3) is specifically as follows: Assume that two energy storage converters are connected to each other and operate in an island mode, V dc The DC side capacitor C is responsible for smoothing the DC bus voltage for the input DC power. The filter inductance, capacitance and resistance are respectively L f , C f and r on The on-resistance of IGBT is represented by r on Indicates, including: V ta 、V tb and V tc is the terminal voltage of the converter, and V sa 、V sb and V sc is the voltage after filtering, I fa ,I fb ,I fc is the capacitor current, I La ,I Lb ,I Lc is the load current.

[0041] Furthermore, the specific process of step (3) is as follows:

[0042] (S1) Initialize N populations;

[0043] (S2) According to the position of the particle Obtain the objective function value of the particle, i.e., fitness;

[0044] (S3) Iteratively update the individual extreme value Pbest and the global extreme value Gbest to achieve The update in the kth iteration, the iteration formula is as follows:

[0045]

[0046] Where: r1, r2 are random numbers between 0 and 1; is the optimal value of the i-th particle after the k-th iteration; g k is the optimal value of all particles in the kth iteration; w is its inertia coefficient; x and v are the position and velocity vectors of the particles, respectively. In order to improve the particle swarm algorithm, the inertia weight ω update formula in equation (12) is first modified:

[0047]

[0048] Where: max is the initial value of inertia weight; ω min is the inertia weight of the last iteration; G max is the number of iterations, and then the learning factor is adjusted asynchronously and dynamically. The calculation formulas are:

[0049]

[0050]

[0051] in,

[0052]

[0053] Where: C max , C min They are the upper and lower limits of C1 and C2 respectively;

[0054] (S4) Using the PSOA algorithm to design the parameters of the VOC; wherein the fitness function and constraint expression formula are as follows:

[0055]

[0056] Among them, Z net (jω) is the filter impedance.

[0057] Furthermore, the specific process of step (4) is as follows: the energy storage converter port adopts virtual synchronous generator control, and the electromagnetic power P output by the synchronous generator e And the input mechanical power P m The following relationship exists:

[0058]

[0059] Where, ω is the mechanical angular velocity; J is the moment of inertia; T m and T e are the mechanical torque and electromagnetic torque of the synchronous generator respectively;

[0060] Considering the damping characteristics of the generator, we have:

[0061]

[0062] D is the damping coefficient; ω N is the rated angular velocity.

[0063] Formula (23) is the rotor motion equation. By simulating formula (23), the active power-frequency control equation of the virtual synchronous generator can be obtained:

[0064]

[0065] P ref , P e are the simulated values ​​of mechanical power and electromagnetic power respectively; θ is the electrical angle; in order to simulate the primary frequency modulation characteristics of the synchronous generator, the active power-frequency droop control equation is added:

[0066] P ref =P set +K ω (ω N -ω) (25)

[0067] Among them, P set is the set value of active power;

[0068] The SOC and rated capacity of different energy storage units are used as constraints, and the rated capacity has a higher priority than the SOC. When the system is disturbed, if the frequency change exceeds the threshold, first determine whether the absolute value of the current active output of the energy storage |P| exceeds the limit. If |P| is within the required limit, select the constraint coefficient K1 according to the size of |P|. After determining K1, determine whether the energy storage is charging or discharging by detecting the positive or negative frequency change, and then input the battery SOC to constrain the SOC; divide the energy storage battery SOC into 5 intervals, each of which represents a different working state of the battery, namely: deep over-discharge, over-discharge, normal operation, over-charge, and deep overcharge.

[0069] The charging and discharging power weights of batteries in different SOC intervals are different. The weights K2 of charging and discharging of each battery are calculated according to the hierarchical analysis method. After determining K1 and K2, the upper-level energy storage EMS control sends the charging and discharging active power instructions to each battery according to formula (26):

[0070] P set =P0+K1|△f|+K2SOC (26)

[0071] Where: P0 is the initial charging and discharging power of the energy storage, |△f| is the absolute value of the frequency change.

[0072] A multifunctional energy storage converter energy storage coordinated control system according to the present invention comprises:

[0073] Optimization and dispatching module: used to establish a robust optimization and dispatching model for distribution networks that takes into account the uncertainty of distributed renewable energy;

[0074] Simplified solution module: used to simplify the solution using KKT, i.e. Karush-Kuhn-Tucker conditions;

[0075] Control module: used to establish a virtual oscillator control method based on the PSO algorithm; the control parameters are adjusted using an improved particle swarm algorithm that optimizes inertia weights and learning factors;

[0076] Multi-time scale collaborative module: used to construct the multi-time scale collaborative control strategy of the shared energy storage energy management system with autonomous frequency support characteristics of grid-type energy storage converters and controlled regulation of grid-type energy storage converters.

[0077] An electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, any one of the multifunctional energy storage converter energy storage collaborative control methods is implemented.

[0078] A storage medium described in the present invention stores a computer program, and when the computer program is executed by a processor, it implements any one of the multifunctional energy storage converter energy storage coordinated control methods described.

[0079] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: without changing the operating state of the PCS, a virtual oscillator control based on the PSOA algorithm is proposed, and the improved particle swarm algorithm that optimizes the inertia weight and the learning factor is used to adjust the controller parameters, thereby avoiding the measurement of active and reactive power, and having better dynamic response than the traditional droop control; a control strategy for a grid-type energy storage inverter is proposed, which fully considers that the energy management system and the active support of the inverter have different time scales, completes automatic power distribution in seconds through the energy management system, and realizes frequency support on a long time scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 It is a structural diagram of the parallel energy storage converter of the energy storage system of the present invention;

[0081] Figure 2 It is the flow chart of the SOA algorithm of the present invention;

[0082] Figure 3 It is the EMS control flow chart of the present invention;

[0083] Figure 4 It is a block diagram of sub-process 1 of the present invention;

[0084] Figure 5 It is a block diagram of sub-process 2 of the present invention;

[0085] Figure 6 It is a block diagram of sub-process 3 of the present invention; DETAILED DESCRIPTION

[0086] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0087] like Figure 1 As shown, an embodiment of the present invention provides a multifunctional energy storage converter energy storage coordinated control method, comprising the following steps:

[0088] (1) A robust optimization dispatching model for distribution networks considering the uncertainty of distributed renewable energy is established. The details are as follows: First, an uncertainty set is used to describe the uncertainty of wind and solar power output:

[0089]

[0090] Among them, P WP To provide new energy, namely wind power and photovoltaic power; is the expected output of new energy, obtained through historical data analysis; λ is a variable that determines the direction of deviation, which takes +1 when the output of new energy is higher than the expected value and takes -1 when it is lower than the expected value; △P WP is the deviation between the expected output of renewable energy and the dispatched output, and its value is determined by formula (2), where σ is the mean square error between the actual value and the expected value of renewable energy output, and α is a conservative estimation parameter.

[0091] Secondly, the goal of the lower-level model problem is to minimize the running cost and the formula is as follows:

[0092]

[0093] Where T is the number of daily dispatch periods; Ngen is the number of traditional small power generators connected to the distribution network; P gen,i,t Dispatching output for traditional units; P buy,t Purchase electricity from the upper grid for the distribution network; is the cost function of the traditional generator set; C buy,t The price of electricity purchased from the upper power grid;

[0094] The goal of establishing the upper model problem is to maximize the penalty cost. The formula is as follows:

[0095]

[0096] Among them, T is the number of daily scheduling periods, N WP is the number of distributed wind farms and photovoltaic power stations connected to the distribution network, △P WP,i,t is the deviation between the expected output of renewable energy and the dispatched output, C i,t This is a penalty electricity price for random deviations in renewable energy output while taking into account the robustness of the algorithm.

[0097] Finally, the objective function of the two-layer scheduling model is as follows:

[0098] f(p)=min[max(f up )+(f down )].

[0099] (2) Using KKT, i.e. Karush-Kuhn-Tucker conditions, to simplify the solution; specifically, the constraints established include: new energy coefficient variable constraints, node power balance constraints, new energy uncertainty coefficient constraints, unit output constraints, traditional unit ramp rate constraints, and network power flow constraints; among them,

[0100] New energy coefficient variable constraints:

[0101]

[0102] Where λ1 and λ2 are Boolean values ​​to satisfy the definition of λ.

[0103] Node power balance constraints:

[0104] P WP,i,t +P gen,i,t +P in,i,t =P LD,i,t +P out,i,t (7)

[0105] Where: P WP,i,t is the renewable energy power generation output of node i at time t; P gen,i,t is the active power output of the traditional unit at node i at time t; P LD,i,t is the active load of node i at time t; P in,i,t With P out,i,t are the active power flowing into node i and the active power flowing out of node i from other lines at time t respectively.

[0106] New energy uncertainty coefficient constraints:

[0107]

[0108] In the formula, t is the conservative coefficient of new energy uncertainty, which is determined according to actual demand. The larger the value, the more robust the scheduling result.

[0109] Unit output constraints:

[0110]

[0111] Where: P max,i With P min,i is the upper and lower limits of the active output of the i-th coal-fired / gas-fired unit; P wp,max,i is the upper limit of the output of the ith typhoon / photovoltaic power station.

[0112] Traditional unit ramp rate constraints:

[0113] P down,i ≤P gen,i,t -P gen,i,t-1 ≤P up,i (10)

[0114] Where: P down,i With P up,i is the ramp rate limit of the i-th traditional unit.

[0115] Network flow constraints:

[0116] -P ij,max ≤P ij,t ≤P ij,max (11)

[0117] Where: P ij,t P is the power flow from node i to node j. It is defined that the flow from i to j is positive, and vice versa is negative. ij,max is the maximum capacity of line ij.

[0118] (3) Figure 2 As shown in the figure, a virtual oscillator control method based on PSO algorithm is established; the control parameters are adjusted by using the improved particle swarm algorithm that optimizes the inertia weight and learning factor; the details are as follows: suppose two energy storage converters are connected to each other and operate in island mode, V dc The DC side capacitor C is responsible for smoothing the DC bus voltage for the input DC power. The filter inductance, capacitance and resistance are respectively L f , C f and r on The on-resistance of IGBT is represented by r on Indicates, including: V ta 、V tb and V tc is the terminal voltage of the converter, and V sa 、V sb and V sc is the voltage after filtering, I fa ,I fb ,I fc is the capacitor current, I La ,I Lb ,I Lc is the load current. The specific process is as follows:

[0119] (S1) Initialize N populations;

[0120] (S2) According to the position of the particle Obtain the objective function value of the particle, i.e., fitness;

[0121] (S3) Iteratively update the individual extreme value Pbest and the global extreme value Gbest to achieve The update in the kth iteration, the iteration formula is as follows:

[0122]

[0123] Where: r1, r2 are random numbers between 0 and 1; is the optimal value of the i-th particle after the k-th iteration; g k is the optimal value of all particles in the kth iteration; w is its inertia coefficient; x and v are the position and velocity vectors of the particles, respectively. In order to improve the particle swarm algorithm, the inertia weight ω update formula in equation (12) is first modified:

[0124]

[0125] Where: max is the initial value of inertia weight; ω min is the inertia weight of the last iteration; G max is the number of iterations, and then the learning factor is adjusted asynchronously and dynamically. The calculation formulas are:

[0126]

[0127] in,

[0128]

[0129] Where: C max , C min They are the upper and lower limits of C1 and C2 respectively;

[0130] (S4) Using the PSOA algorithm to design the parameters of the VOC; wherein the fitness function and constraint expression formula are as follows:

[0131]

[0132] Among them, Z net (jω) is the filter impedance.

[0133] (4) Construct a multi-time scale collaborative control strategy for a shared energy storage energy management system that has the autonomous frequency support characteristics of a grid-type energy storage converter and the controlled regulation of a grid-type energy storage converter. The specific process is as follows: The energy storage converter port is controlled by a virtual synchronous generator, and the electromagnetic power P output by the synchronous generator e And the input mechanical power P m The following relationship exists:

[0134]

[0135] Where, ω is the mechanical angular velocity; J is the moment of inertia; T m and T e are the mechanical torque and electromagnetic torque of the synchronous generator respectively;

[0136] Considering the damping characteristics of the generator, we have:

[0137]

[0138] D is the damping coefficient; ω N is the rated angular velocity.

[0139] Formula (23) is the rotor motion equation. By simulating formula (23), the active power-frequency control equation of the virtual synchronous generator can be obtained:

[0140]

[0141] P ref , P e are the simulated values ​​of mechanical power and electromagnetic power respectively; θ is the electrical angle; in order to simulate the primary frequency modulation characteristics of the synchronous generator, the active power-frequency droop control equation is added:

[0142] P ref =P set +K ω (ω N -ω) (25)

[0143] Among them, P set is the set value of active power;

[0144] The SOC and rated capacity of different energy storage units are used as constraints, and the rated capacity has a higher priority than the SOC. When the system is disturbed, if the frequency change exceeds the threshold, first determine whether the absolute value of the current active output of the energy storage |P| exceeds the limit. If |P| is within the required limit, select the constraint coefficient K1 according to the size of |P|. After determining K1, determine whether the energy storage is charging or discharging by detecting the positive or negative frequency change, and then input the battery SOC to constrain the SOC; divide the energy storage battery SOC into 5 intervals, each of which represents a different working state of the battery, namely: deep over-discharge, over-discharge, normal operation, over-charge, and deep overcharge.

[0145] The charging and discharging power weights of batteries in different SOC intervals are different. The weights K2 of charging and discharging of each battery are calculated according to the hierarchical analysis method. After determining K1 and K2, the upper-level energy storage EMS control sends the charging and discharging active power instructions to each battery according to formula (26):

[0146] P set =P0+K1|△f|+K2SOC (26)

[0147] Where: P0 is the initial charging and discharging power of the energy storage, |△f| is the absolute value of the frequency change. The above overall control flow chart is as follows Figure 3 As shown, the sub-processes are as follows Figures 4 to 6 As shown. Figure 3 In the control process shown, the frequency change |△f| is first input to determine whether |△f| is within the threshold of 0.05. If so, the energy storage operates according to the initial charge and discharge power P0; if it exceeds the threshold, it enters the rated capacity judgment link. The rated capacity of the energy storage unit used in this article is 1MW, so it is necessary to determine whether the current output of the battery exceeds 1MW. If it exceeds 1MW, the constraint coefficients K1 and K2 are set to zero, and the energy storage operates with the initial active power; otherwise, different K1s are taken according to the absolute value of the current active output of the energy storage |P|.

[0148] An embodiment of the present invention provides a multifunctional energy storage converter energy storage coordinated control system, comprising:

[0149] Optimization and dispatching module: used to establish a robust optimization and dispatching model for distribution networks that takes into account the uncertainty of distributed renewable energy;

[0150] Simplified solution module: used to simplify the solution using KKT, i.e. Karush-Kuhn-Tucker conditions;

[0151] Control module: used to establish a virtual oscillator control method based on the PSO algorithm; the control parameters are adjusted using an improved particle swarm algorithm that optimizes inertia weights and learning factors;

[0152] Multi-time scale collaborative module: used to construct the multi-time scale collaborative control strategy of the shared energy storage energy management system with autonomous frequency support characteristics of grid-type energy storage converters and controlled regulation of grid-type energy storage converters.

[0153] An electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, any one of the multifunctional energy storage converter energy storage collaborative control methods is implemented.

[0154] An embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is loaded into the processor, any one of the multifunctional energy storage converter energy storage collaborative control methods is implemented.

[0155] An embodiment of the present invention provides a storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements any one of the multifunctional energy storage converter energy storage coordinated control methods.

Claims

1. A method for coordinated control of energy storage of a multifunctional energy storage converter, characterized in that: The following steps are involved: (1) A robust optimization dispatch model for distribution networks considering the uncertainty of distributed renewable energy was established; (2) Use KKT, i.e. Karush-Kuhn-Tucker conditions, to simplify the solution; (3) Establish a virtual oscillator control method based on the PSO algorithm; use an improved particle swarm algorithm that optimizes inertia weight and learning factor to adjust the control parameters; (4) Construct a multi-time scale collaborative control strategy for a shared energy storage management system with autonomous frequency support characteristics of grid-type energy storage converters and controlled regulation of grid-type energy storage converters.

2. A method for coordinated control of energy storage of a multifunctional energy storage converter according to claim 1, characterized in that: Step (1) is as follows: First, an uncertainty set is used to describe the uncertainty of wind and solar output: Among them, P WP To provide new energy, namely wind power and photovoltaic power; is the expected output of new energy, obtained through historical data analysis; λ is a variable that determines the direction of deviation, which takes +1 when the output of new energy is higher than the expected value and takes -1 when it is lower than the expected value; △P WP is the deviation between the expected output of renewable energy and the dispatched output, and its value is determined by formula (2), where σ is the mean square error between the actual value and the expected value of renewable energy output, and α is a conservative estimation parameter. Secondly, the goal of the lower-level model problem is to minimize the running cost and the formula is as follows: Where T is the number of daily dispatch periods; Ngen is the number of traditional small power generators connected to the distribution network; P gen,i,t Dispatching output for traditional units; P buy,t Purchase electricity from the upper grid for the distribution network; is the cost function of the traditional generator set; C buy,t The price of electricity purchased from the upper power grid; The goal of establishing the upper model problem is to maximize the penalty cost. The formula is as follows: Among them, T is the number of daily scheduling periods, N WP is the number of distributed wind farms and photovoltaic power stations connected to the distribution network, △P WP,i,t is the deviation between the expected output of renewable energy and the dispatched output, C i,t This is a penalty electricity price for random deviations in renewable energy output while taking into account the robustness of the algorithm. Finally, the objective function of the two-layer scheduling model is as follows: f(p)=min[max(f up )+(f down )]。 3. A method for coordinated control of energy storage of a multifunctional energy storage converter according to claim 1, characterized in that: Step (2) is as follows: Establish constraint conditions including: new energy coefficient variable constraint, node power balance constraint, new energy uncertainty coefficient constraint, unit output constraint, traditional unit ramp rate constraint and network power flow constraint; where: New energy coefficient variable constraints: Where λ1 and λ2 are Boolean values ​​to satisfy the definition of λ. Node power balance constraints: P WP,i,t +P gen,i,t +P in,i,t =P LD,i,t +P out,i,t (7) Where: P WP,i,t is the renewable energy power generation output of node i at time t; P gen,i,t is the active power output of the traditional unit at node i at time t; P LD,i,t is the active load of node i at time t; P in,i,t With P out,i,t are the active power flowing into node i and the active power flowing out of node i from other lines at time t respectively. New energy uncertainty coefficient constraints: In the formula, t is the conservative coefficient of new energy uncertainty, which is determined according to actual demand. The larger the value, the more robust the scheduling result. Unit output constraints: Where: P max,i With P min,i is the upper and lower limits of the active output of the i-th coal-fired / gas-fired unit; P wp,max,i is the upper limit of the output of the ith typhoon / photovoltaic power station. Traditional unit ramp rate constraints: P down,i ≤P gen,i,t -P gen,i,t-1 ≤P up,i (10) Where: P down,i With P up,i is the ramp rate limit of the i-th traditional unit. Network flow constraints: -P ij,max ≤P ij,t ≤P ij,max (11) Where: P ij,t P is the power flow from node i to node j. It is defined that the flow from i to j is positive, and vice versa is negative. ij,max is the maximum capacity of line ij.

4. A method for coordinated control of energy storage of a multifunctional energy storage converter according to claim 1, characterized in that: Step (3) is as follows: Assume that two energy storage converters are connected to each other and operate in island mode, V dc The DC side capacitor C is responsible for smoothing the DC bus voltage for the input DC power. The filter inductance, capacitance and resistance are respectively L f , C f and r on The on-resistance of IGBT is represented by r on Indicates, including: V ta 、V tb and V tc is the terminal voltage of the converter, and V sa 、V sb and V sc is the voltage after filtering, I fa ,I fb ,I fc is the capacitor current, I La ,I Lb ,I Lc is the load current.

5. A method for coordinated control of energy storage of a multifunctional energy storage converter according to claim 3, characterized in that: The specific process of step (3) is as follows: (S1) Initialize N populations; (S2) According to the position of the particle Obtain the objective function value of the particle, i.e., fitness; (S3) Iteratively update the individual extreme value Pbest and the global extreme value Gbest to achieve The update in the kth iteration, the iteration formula is as follows: Where: r1, r2 are random numbers between 0 and 1; is the optimal value of the i-th particle after the k-th iteration; g k is the optimal value of all particles in the kth iteration; w is its inertia coefficient; x and v are the position and velocity vectors of the particles, respectively. In order to improve the particle swarm algorithm, the inertia weight ω update formula in equation (12) is first modified: Where: max is the initial value of inertia weight; ω min is the inertia weight of the last iteration; G max is the number of iterations, and then the learning factor is adjusted asynchronously and dynamically. The calculation formulas are: in, Where: C max , C min They are the upper and lower limits of C1 and C2 respectively; (S4) Using the PSOA algorithm to design the parameters of the VOC; wherein the fitness function and constraint expression formula are as follows: Among them, Z net (jω) is the filter impedance.

6. A method for coordinated control of energy storage of a multifunctional energy storage converter according to claim 3, characterized in that: The specific process of step (4) is as follows: the energy storage converter port adopts virtual synchronous generator control, and the electromagnetic power P output by the synchronous generator is e And the input mechanical power P m The following relationship exists: Where, ω is the mechanical angular velocity; J is the moment of inertia; T m and T e are the mechanical torque and electromagnetic torque of the synchronous generator respectively; Considering the damping characteristics of the generator, we have: D is the damping coefficient; ω N is the rated angular velocity. Formula (23) is the rotor motion equation. By simulating formula (23), the active power-frequency control equation of the virtual synchronous generator can be obtained: P ref , P e are the simulated values ​​of mechanical power and electromagnetic power respectively; θ is the electrical angle; in order to simulate the primary frequency modulation characteristics of the synchronous generator, the active power-frequency droop control equation is added: P ref =P set +K ω (oh N -ω) (25) Among them, P set is the set value of active power; The SOC and rated capacity of different energy storage units are used as constraints, and the rated capacity has a higher priority than the SOC. When the system is disturbed, if the frequency change exceeds the threshold, first determine whether the absolute value of the current active output of the energy storage |P| exceeds the limit. If |P| is within the required limit, select the constraint coefficient K1 according to the size of |P|. After determining K1, determine whether the energy storage is charging or discharging by detecting the positive or negative frequency change, and then input the battery SOC to constrain the SOC; divide the energy storage battery SOC into 5 intervals, each of which represents a different working state of the battery, namely: deep over-discharge, over-discharge, normal operation, over-charge, and deep overcharge. The charging and discharging power weights of batteries in different SOC intervals are different. The weights K2 of charging and discharging of each battery are calculated according to the hierarchical analysis method. After determining K1 and K2, the upper-level energy storage EMS control sends the charging and discharging active power instructions to each battery according to formula (26): P set =P0+K1|△f|+K2SOC (26) Where: P0 is the initial charging and discharging power of the energy storage, |△f| is the absolute value of the frequency change.

7. A multifunctional energy storage converter energy storage coordinated control system, characterized in that: include: Optimization and dispatching module: used to establish a robust optimization and dispatching model for distribution networks that takes into account the uncertainty of distributed renewable energy; Simplified solution module: used to simplify the solution using KKT, i.e. Karush-Kuhn-Tucker conditions; Control module: used to establish a virtual oscillator control method based on the PSO algorithm; the control parameters are adjusted using an improved particle swarm algorithm that optimizes inertia weights and learning factors; Multi-time scale collaborative module: used to construct the multi-time scale collaborative control strategy of the shared energy storage energy management system with autonomous frequency support characteristics of grid-type energy storage converters and controlled regulation of grid-type energy storage converters.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into the processor, a multifunctional energy storage converter energy storage coordinated control method according to any one of claims 1 to 6 is implemented.

9. A storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, a multifunctional energy storage converter energy storage coordinated control method according to any one of claims 1 to 6 is implemented.

Citation Information

Cited By

  • Improved energy storage converter network following control algorithm

    CN121238621A