A method for establishing a joint control strategy for load and frequency of power grid considering hybrid energy storage

Abstract

The present invention relates to a method for establishing a joint control strategy for load frequency of a hybrid energy storage-assisted power grid, which belongs to the technical field of power systems. In order to solve the frequency control problem faced by power systems under the background of "double highs", energy storage equipment is added in a timely manner to perform efficient load frequency control to meet the frequency regulation requirements of the power grid. A method for establishing a joint control strategy for load frequency of a hybrid energy storage-assisted power grid is provided, comprising the following steps: S1) analyzing a load frequency control model, S2) building a hybrid energy storage-assisted LFC model, S3) designing a FOPID controller and using chaotic particle swarm optimization to adjust its control parameters, and S4) simulation analysis and verification. The control strategy proposed in the present invention exhibits good dynamic performance under both step and continuous load disturbance conditions, reduces frequency deviations while shortening recovery time, and promptly restores frequency stability of the interconnected power system.

Description

Technical Field

[0001] The present invention relates to a method for establishing a hybrid energy storage-assisted power grid load frequency joint control strategy, and belongs to the technical field of power systems. Background Art

[0002] With the worsening of environmental problems and the depletion of primary energy, traditional power systems are unable to quickly adapt to the requirements of the "dual carbon" goals. Therefore, a high proportion of renewable energy and a high proportion of power electronic equipment are gradually becoming the trend and characteristics of future power system development. The grid connection of new energy represented by wind power and photovoltaics has, on the one hand, reduced environmental pollution and dependence on fossil energy, and on the other hand, changed the structure of traditional power grids, which has also brought new challenges to the frequency stability of regional power grids.

[0003] As an important indicator for measuring power quality, grid frequency change is of great significance to the stable and safe operation of the power system. As one of the basic purposes of automatic power generation control, load frequency control is an important means to achieve grid frequency stability. In order to make up for the problem of unsatisfactory LFC of traditional units, fast response frequency regulation resources represented by pumped storage and energy storage batteries have been widely used in auxiliary grid frequency regulation in recent years [4-6]. Reference [4] established a two-region steam turbine unit LFC model taking into account nonlinear links and governor dead zone, and proposed a fractional-order PID control strategy for regional load frequency of pumped storage power stations on this basis, but did not adjust the control parameters. It is also proposed to combine energy storage batteries with pumped storage power stations to participate in secondary grid frequency regulation, and the advantages of various frequency regulation resources complement each other to give full play to the frequency regulation potential.

[0004] A hybrid energy storage system refers to the combination of two or more energy storage systems with complementary performance. It is an effective way to solve the current economic inefficiency of a single energy storage technology. We usually divide energy storage equipment into energy storage represented by various types of energy storage batteries and power storage represented by supercapacitors and flywheels. The former has a large capacity but a short energy storage life, while the latter has a large power density and a long energy storage life but a low energy density. By utilizing the complementarity of the above two types of energy storage resources in terms of technical characteristics, it can provide flexible and high-quality frequency regulation auxiliary services for the power grid while smoothing the volatility of wind power. At present, there have been preliminary results in the frequency regulation of HESS-assisted power grids. A control strategy for a hybrid energy storage system consisting of supercapacitors and batteries to participate in AGC has been proposed. Subsequently, in order to smooth the fluctuation of wind power output, a wind power hybrid energy storage system has also been proposed. Some studies have proposed to establish an opportunity-constrained programming model with the goal of minimizing average cost, and to reasonably configure the power and capacity of supercapacitors and batteries. In view of the problem of sensor failure in the hybrid energy storage system during the transmission and distribution of electric energy, an integral sliding mode fault-tolerant controller is designed to perform active fault-tolerant control on the sensor.

[0005] In summary, pumped storage and single energy storage battery systems, as good frequency regulation resources, are limited by geographical location and high costs and cannot be applied on a large scale. Introducing HESS into the grid-assisted frequency regulation process makes up for the above problems while also having good frequency regulation performance. However, the current research on HESS mainly considers the power frequency deviation of the interconnection line of the LFC system under step disturbances, and does not analyze the output and charge state changes of the energy storage battery when the HESS assists the grid frequency regulation under actual wind farm access. Therefore, in order to effectively smooth out the frequency deviation and interconnection line power changes under wind power disturbances, solve the problem of life attenuation caused by frequent charging and discharging, and optimize the output and SOC of energy storage batteries, the control method of HESS-assisted grid frequency regulation needs further in-depth research. Summary of the invention

[0006] In order to solve the frequency control problem faced by the power system under the background of "double highs", the present invention timely adds energy storage equipment to perform efficient load frequency control to meet the grid frequency regulation requirements, and provides a method for establishing a joint control strategy for load frequency of a hybrid energy storage-assisted grid.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for establishing a joint control strategy of load frequency of a hybrid energy storage-assisted power grid, comprising the following steps:

[0008] S1) Analyze the load frequency control model;

[0009] S2) Building a hybrid energy storage combined with LFC model;

[0010] S3) Design FOPID controller and use chaotic particle swarm optimization to tune its control parameters;

[0011] S4) Simulation analysis and verification.

[0012] Compared with the prior art, the present invention has the following beneficial effects: the proposed control strategy exhibits good dynamic performance under both step and continuous load disturbance conditions, reduces frequency deviation while shortening recovery time, and promptly restores frequency stability of the interconnected power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described in detail below in conjunction with the accompanying drawings;

[0014] Figure 1 This is the model diagram of the reheat steam turbine unit considering GRC;

[0015] Figure 2 It is the transfer function model diagram of energy storage power supply;

[0016] Figure 3 It is the transfer function model diagram of supercapacitor;

[0017] Figure 4 This is a diagram of the joint frequency regulation model with the participation of hybrid energy storage;

[0018] Figure 5 A diagram of the method for HESS to participate in secondary frequency modulation;

[0019] Figure 6 This is the FOPID control system model diagram;

[0020] Figure 7 This is the controller performance comparison curve after parameter optimization;

[0021] Figure 8 It is the frequency deviation curve under step disturbance;

[0022] Fig. 9 It is the transmission power curve of the tie line under step disturbance;

[0023] Fig.10 is the random power fluctuation curve of the wind farm;

[0024] Fig.11 To take into account the frequency deviation curve of wind power disturbance;

[0025] Fig.12 Transmission power curve of tie line considering wind power disturbance;

[0026] Fig.13 It is the output curve of energy storage battery under different systems;

[0027] Fig.14 The SOC change curve of the energy storage battery under different systems. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0029] The present invention relates to a method for establishing a hybrid energy storage-assisted power grid load frequency joint control strategy, comprising the following steps:

[0030] S1) Analyze the load frequency control model;

[0031] S2) Building a hybrid energy storage combined with LFC model;

[0032] S3) Design FOPID controller and use chaotic particle swarm optimization to tune its control parameters;

[0033] S4) Simulation analysis and verification.

[0034] The step S1) analyzing the load frequency control model comprises:

[0035] S11) System frequency response model

[0036] According to the strategy of calculating the area control error (ACE), AGC is divided into three modes: FFC (fixed frequency control), FTC (constant exchange power control) and TBC (interconnection line frequency deviation control). The TBC mode, that is, the interconnection line frequency deviation control mode, is a combination of the FFC and FTC modes. It can effectively stabilize the frequency deviation of the interconnected power grid, as shown in equations (1) and (2):

[0037] ACE i =ΔP tiei,j +β i Δf i (1)

[0038]

[0039] In the formula, ΔP tiei,j is the power exchange deviation of the tie line between area i and area j; Δf i is the system frequency deviation; β i is the response coefficient; R i and D i are the adjustment coefficient and load damping coefficient respectively; ACE i represents the error signal of region i;

[0040] In summary, the TBC mode is adopted to i As the output feedback item of the LFC controller to achieve the purpose of maintaining the frequency stability of the interconnected power system;

[0041] S12) LFC component model

[0042] Build an LFC (load frequency control) system, model the prime mover, governor, generator-load and tie line components in the system, and select the reheat steam turbine as the prime mover. The model is shown in formula (3):

[0043]

[0044] In the formula, K r is the proportion of steam power generated in the high-pressure cylinder section; T r is the reheater time coefficient; T t is the steam chamber time coefficient; s is the Laplace transform operator;

[0045] Generation Rate Constraints (GRC) are used to limit the unit output by a given maximum change value. The reheat steam turbine unit model considering GRC is constructed as follows: Figure 1 As shown, the limiter value of the limiting link is set to ±0.0017puMW / s;

[0046] Formula (4) is the transfer function model of the turbine speed governor. Setting the speed governor dead zone for the unit can reduce the frequent operation of the speed governor caused by the frequency fluctuation of the power grid, thereby protecting the speed governor and extending its service life. At present, the speed governor dead zone in the power system is generally less than 1r / min (i.e. ±0.017Hz). By linearizing it, the speed governor model with dead zone can be obtained, as shown in Formula (5):

[0047]

[0048]

[0049] Where, T g is the time constant of the speed regulator, N1 and N2 are linearization coefficients;

[0050] The generator-load model describes the relationship between power change and frequency change in the interconnected power system, as shown in equation (6):

[0051]

[0052] In the formula, ΔP m is the output power of the generator; ΔP i is the applied load disturbance; M i is the inertia constant of the generator; D i is the load damping coefficient;

[0053] S13) Tie line model

[0054] The tie line model has been described in the S11) system frequency response model and is often used in conjunction with TBC for secondary frequency modulation, such as equations (1) and (2).

[0055] The step S2) of building a hybrid energy storage combined with LFC model includes:

[0056] S21) Energy storage battery frequency modulation model

[0057] The simplified energy storage power transfer function model is as follows: Figure 2 As shown, the description of its mathematical model is shown in formula (7) to formula (12):

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064] In the formula, ACE i For the input control signal, select ACE i , that is, the error signal of region i is used as the control signal, and the current is selected as the response variable inside the battery; ΔU oc , ΔU serious , ΔU c , ΔU t and ΔU b They are the open circuit voltage, internal resistance voltage increment, connection impedance voltage increment, transient voltage increment and the battery terminal voltage value is obtained by superimposing the above increments; ΔI b is the current flowing through the battery; T b is the time constant of the first-order inertia link of the battery; C0 and C p are the initial capacity and rated capacity of the battery respectively; R series is the internal resistance of the battery cell; n is the number of batteries in series; m is the number of subsystems in parallel; k is the number of energy storage units in parallel; R c is the connection impedance between the energy storage unit and the PCS; R t is the overvoltage resistance; C t is the overvoltage capacitance; ΔP b is the actual output value of the active power of the energy storage battery;

[0065] Open circuit voltage U oc It is a function of SOC (state of charge), usually expressed as a percentage. The battery SOC expression is as follows:

[0066]

[0067] In the formula, S(t) and S(t0) are the SOC of the energy storage battery at time t and the initial time t0 respectively; C is the rated capacity of the battery; η bat is the battery charge / discharge efficiency, representing the internal power dissipated during the charge / discharge cycle, SOC∈[0.2,0.8];

[0068] S22) Supercapacitor frequency modulation model

[0069] The most commonly used equivalent form of supercapacitors is a parallel circuit of resistors and capacitors. It is difficult to ensure that the capacitor voltage remains stable when considering the initial voltage of the capacitor. In order to solve this problem, a voltage feedback loop can be introduced to keep it stable. Its transfer function model is as follows: Figure 3 As shown, the description of the mathematical model is shown in equations (14) to (16):

[0070]

[0071]

[0072] ΔP sc =(U do +ΔU d )ΔI d (16)

[0073] In the formula, ΔI d , ΔU d and ΔP sc are the actual output values ​​of current, voltage and active power flowing through the supercapacitor respectively; T C is the time constant; K CA ACE is the control signal i Gain factor; K vd is the voltage feedback loop gain coefficient; R is the equivalent resistance; C is the equivalent capacitance; U d0 is the initial voltage of the supercapacitor;

[0074] S23) Hybrid energy storage combined frequency regulation model

[0075] Combining the auxiliary frequency regulation model proposed in step S21) and S22), the joint frequency regulation model with the participation of hybrid energy storage can be obtained as shown in equations (17) and (18):

[0076] ΔP m -ΔP i +ΔP hess =(M i s+D i )Δf i (17)

[0077] ΔP hess =ΔP b +ΔP sc (18)

[0078] Where ΔP hess is the actual output value of HESS active power; HESS stands for hybrid energy storage system;

[0079] The energy storage battery frequency regulation model and the supercapacitor frequency regulation model are respectively connected to the IEEE standard two-region LFC model to obtain the joint frequency regulation model with the participation of hybrid energy storage, as shown in Figure 4 shown.

[0080] The methods of HESS participating in secondary frequency regulation of power grid are as follows: Figure 5 Assume that the load frequency characteristic curve in the initial state is P L1 When the load suddenly increases, the load frequency characteristic curve will move to P L2 At this time, the traditional power source starts the primary frequency regulation function, the grid operation point will move from the stable operation point a to the unstable point b, and a frequency deviation Δf1 (which is a negative value) will be generated at the same time, and the primary frequency regulation ends. The traditional power source starts the secondary frequency regulation function. Assuming that its spare capacity is insufficient, the power frequency characteristic curve will change from P G1 Move to P G2 , the secondary frequency modulation output is ΔP G , the grid operation point will move from point b to point c, that is, the frequency deviation rises from Δf1 to Δf2, and the frequency deviation is still not completely eliminated. In this scenario, the power command controls the HESS to discharge and output active power P hess , the frequency deviation will gradually return to 0. That is, the traditional power source and HESS participate in the secondary frequency modulation, and through the reasonable allocation of the regional control error (ACE) signal, the output of the traditional power source is ΔP G , the output of HESS is P hess , and ultimately achieve error-free regulation of the grid frequency.

[0081] When the HESS composed of energy storage batteries and supercapacitors is connected to the grid and participates in joint frequency regulation, Figure 6 It can be seen that it assists the load frequency control unit of the traditional generator set to jointly complete the primary and secondary frequency regulation tasks. At this time, in order to make each control unit work together to achieve the best control effect, it is necessary to consider the load characteristics, unit characteristics, etc., and it is necessary to select a suitable controller and adjust the control parameters.

[0082] In the past, LFC models mostly used PI and PID controllers. In recent years, PID has been expanded from the integer order field to the fractional order field and achieved good control effects. Based on this, the FOPID controller is selected and its control parameters are tuned and optimized using intelligent algorithms to achieve better frequency modulation effects.

[0083] The step S3) designs the FOPID controller and uses the chaotic particle swarm optimization to adjust its control parameters, specifically:

[0084] S31) FOPID control

[0085] The continuous FOPID operator is usually defined as shown in formula (19):

[0086]

[0087] In the formula, α is the order of calculus, t and α are the upper and lower limits of calculus respectively;

[0088] The FOPID controller is a deepening of the integer-order PID controller in the fractional-order field. It adds two adjustable parameters, λ and μ, to the PID control and makes it have a better adjustment range. Its transfer function expression is:

[0089]

[0090] In the formula, K p , K i , K d are the proportional, integral and differential coefficients respectively, λ and μ are the integral order and differential order respectively;

[0091] Compared with the traditional PID controller, only K needs to be adjusted p , K i , K d 3 parameters, K needs to be tuned in FOPID controller p , K i , K d , λ and μ, a total of 5 parameters, so it is necessary to select a suitable high-order optimization algorithm to obtain the optimal control parameters;

[0092] Combining the above formula with the unit model, the FOPID control system model can be obtained as follows: Figure 6 As shown, r i and i are the input and output signals of the control system, u i is the output signal of the controller, ACE i =r i -y i is the regional control error signal, which is used as the output feedback term to maintain the stability of the interconnected power system. The research frequency band of this invention is [10 -3 , 10 3 ], the improved Oustaloup filtering algorithm is used to approximate the FOPID operator. A filter is added to the original Oustaloup approximation algorithm, so that it can achieve better approximation both within the frequency band and at the endpoints;

[0093] S32) Optimization model

[0094] (1) Optimizing the objective function

[0095] As a control system performance evaluation index with good engineering practicality and selectivity, the performance index of time multiplied by the integral of the absolute value of the error (ITAE) has been widely used. By setting ITAE as the objective function, the ACE of area i is taken as the instantaneous error e(t) to establish the optimization model of the system, as shown in Equation (21).

[0096]

[0097] Since there are only two variables, ACE and t, in formula (21), it is difficult to determine whether the error in the system is a positive error or a negative error. At the same time, the overshoot will be sacrificed to reduce the rise time of the system. Therefore, the objective function in formula (21) is modified and a penalty link is added to avoid the occurrence of excessive overshoot. The modified objective function is shown in formula (22):

[0098]

[0099] In the formula, t r is the rise time; ξ1, ξ2 are weight coefficients;

[0100] The penalty function is used to achieve dynamic adjustment. When ξ1>>ξ2 in equation (22), the rise time of the system is sacrificed to reduce or eliminate the overshoot of the system. When ξ1<<ξ2 in equation (22), the overshoot of the system increases, but the rise time of the system decreases.

[0101] (2) Optimize the process

[0102] The particle swarm optimization algorithm searches for the motion state of particles from the individual and global perspectives, and continuously iterates and updates the particle speed and position by combining formulas (23)-(24), and finally obtains the local optimal solution and the global optimal solution:

[0103]

[0104]

[0105] Where ω is the inertia weight, Respectively represent the particle's speed, position, individual optimal solution and global optimal solution; c1 and c2 are acceleration factors; r1 and r2 are random numbers between [0,1];

[0106] However, the traditional particle swarm optimization algorithm is prone to fall into the local optimal solution, and the particle stagnation leads to premature algorithmization. The chaotic particle swarm optimization algorithm uses the ergodicity of chaotic variables to effectively prevent the above problems, and has the advantages of good applicability and fast convergence speed for the high-dimensional parameter optimization problem in FOPID. Mapped to the domain of the Logistic equation through equation (25):

[0107]

[0108]

[0109]

[0110] The chaotic particle swarm optimization algorithm is used to optimize the FOPID control parameters online.

[0111] The online optimization process of FOPID control parameters using chaotic particle swarm optimization algorithm is as follows:

[0112] 1) Initialize the particle swarm and set relevant parameters such as particle swarm size and maximum number of iterations;

[0113] 2) According to the particle position, assign a value to K p , K i and K d or K p , K i , K d , λ and μ;

[0114] 3) Run the two-region LFC model and return the target value J gbest ;

[0115] 4) Global optimal value x of particle swarm gbest Perform chaos optimization:

[0116] ① Through formula (25) Mapped to the domain of the Logistic equation [0,1];

[0117] ② Using formula (26) Perform M iterations to obtain the chaotic sequence

[0118] ③ According to formula (27), the chaotic sequence is inversely mapped back to the original solution space to obtain the feasible solution sequence x ki (1),x ki (2)……x ki (M);

[0119] ④ Run the two-region LFC model to obtain the optimal solution after chaos optimization

[0120] ⑤ Randomly replace any particle group in the current particle group with the optimal particle group obtained in the chaotic sequence;

[0121] ⑥ Update the objective function value, such as J' gbest <J gbest , then J gbest =J' gbest,

[0122] 5) Update the particle velocity and position according to the individual optimal solution and the global optimal solution. Let k = k + 1. If k < K, then return to step 2; otherwise, the obtained is the global optimal solution;

[0123] 6) Output the corresponding parameters as the optimal control parameters, and the solution is completed.

[0124] The simulation analysis and verification content in step S4) includes: controller performance verification, step load disturbance verification, and continuous disturbance verification considering the uncertainty of wind power.

[0125] 1) Controller performance verification

[0126] In the present invention, a reheating steam turbine unit LFC model considering GRC in two areas is built in MATLAB / Simulink, and a hybrid energy storage assisted frequency modulation model with energy storage batteries and supercapacitors is built on the basis of the model. The capacity of the model system is 1000MW. First, the controller performance is verified under the traditional two-area LFC system without energy storage batteries or supercapacitors considering nonlinearity. Table 1 shows the basic parameters of the two-area LFC model. The unit parameter values in the two areas are set the same. The present invention simulates the dynamic response of the system during load fluctuations by setting a step signal. A step signal of 0.01 p.u. is applied to area i at 2s as the connected load, and the load is removed at 45s. The values of each optimization parameter are K p ∈[0, 200], K i ∈[0, 100], K d ∈[0, 100], λ ∈ [0, 2], μ ∈ [0, 2]. The basic parameters of the optimized controller are shown in Table 2.

[0127] Table 1 Basic parameters of two-region LFC model

[0128] Table1 Basic parameters of two-region LFC model

[0129]

[0130] Table 2 The basic parameters of the controller has been optimized

[0131] Table2 The basic parameters of the controller has been optimized

[0132]

[0133] By Figure 7 It can be seen that after the sudden increase in load, if the PID controller is optimized by the PSO algorithm, the maximum frequency deviation is 0.08026Hz and the recovery time is 28.5s; if the FOPID controller is optimized by the PSO algorithm or the CPSO algorithm, the maximum frequency deviation is 0.06132Hz and 0.05575Hz, which is 23.47% and 30.54% less than the former, and the recovery time is 16.67s and 14.12s, which is about 41.51% and 50.46% shorter, respectively. Overall, compared with the traditional PID controller, the maximum frequency deviation and recovery time of the FOPID controller are obviously superior, reflecting good control performance; compared with the PSO-FOPID controller, the CPSO-FOPID controller does not show obvious advantages in recovery time, but it shows better regulation performance in terms of maximum frequency deviation and overshoot. Similarly, after the load is removed, the LFC system using the CPSO-FOPID controller also shows good frequency regulation performance.

[0134] 2) Step load disturbance verification

[0135] First, the following assumptions are made before simulation verification: 1) Both the energy storage battery and the supercapacitor have the ability to be connected to the grid to assist the grid frequency regulation; 2) The energy storage battery and the supercapacitor have sufficient available capacity and backup capacity; 3) The influencing factors such as the life of the energy storage battery and the supercapacitor are not considered in the optimization model. The energy storage battery is selected as a 50Ah lithium iron phosphate battery with a configuration capacity of 1MW·h, an initial SOC of 0.5, and SOC∈[0.2,0.8]; the supercapacitor is configured with a capacity of 0.5MW·h. The other power model parameters in the HESS are shown in Table 3.

[0136] Table 3 Parameters of each HESS power model

[0137] Table3 All power model parameters of hybrid energy storage system

[0138]

[0139] The present invention sets three operating modes, namely, the traditional generator set frequency modulation mode without the participation of the energy storage system, the BESS single energy storage auxiliary frequency modulation mode with the participation of lithium batteries, and the hybrid energy storage auxiliary frequency modulation mode with the participation of lithium batteries and supercapacitors proposed in the present invention. The controller adopts the FOPID controller designed by the present invention and optimized by the CPSO algorithm, and the control parameters remain unchanged.

[0140] Figure 8is the frequency deviation of different operating modes after applying a step disturbance to area i. The maximum frequency deviations of the three modes of traditional frequency modulation, BESS auxiliary frequency modulation and auxiliary frequency modulation proposed in this invention are 0.06072 Hz, 0.02834 Hz ​​and 0.01972 Hz, respectively, and the adjustment times are 22.19 s, 13.54 s and 11.95 s, respectively. Compared with the traditional frequency modulation mode, the maximum frequency differences of the latter two modes are reduced by about 53.33% and 67.52%, respectively, and the adjustment time is shortened by 38.98% and 46.15%, respectively. Through comparative analysis, it can be seen that compared with the traditional frequency modulation mode, the latter two frequency modulation modes, especially the hybrid energy storage frequency modulation mode proposed in this paper, greatly reduce the disturbance range, reduce the maximum frequency deviation value by 2 / 3, greatly shorten the recovery time, and have better frequency modulation performance.

[0141] Fig. 9 is the transmission power variation curve of the tie line under step disturbance. It can be seen from the figure that after the load suddenly increases, if the system has no energy storage system to assist frequency regulation, the peak transmission power of the tie line can reach 8.762×10 -3 puMW, while after adopting the auxiliary frequency modulation method proposed in the present invention, the peak transmission power of the tie line is only 4.221×10 -3 puMW, which is a reduction of nearly 51.8%. At the same time, it also makes up for the problem of long recovery time and frequency oscillation leading to system instability when a single energy storage system participates in frequency modulation.

[0142] 3) Continuous disturbance verification considering wind power uncertainty

[0143] In order to further verify the control performance of the control method proposed in this paper in actual scenarios, a random wind power sequence is generated, and the wind farm output power fluctuation per unit value (reference power is 10MVA) curve is as follows: Fig.10 As shown in Figure 2, the overall trend of the wind farm is to adjust the frequency upward, that is, the energy storage system is more needed to absorb the excess power generated by the wind farm. The frequency control effect of different control modes under the random disturbance of wind power is compared. The frequency deviation curve of region i is shown in Figure 2. Fig.11 As shown in the figure, the traditional control mode is seriously affected by the random disturbance of wind power, the regional frequency deviation fluctuates deeply, and a large deviation value of 0.02872Hz appears. The system continues to oscillate when approaching the stable point and is difficult to maintain stability. Compared with the other two control modes, the frequency deviation of the system is significantly improved when the auxiliary frequency modulation mode proposed in the present invention is adopted. The maximum frequency difference is 0.0089Hz, and the curve is smoother when approaching the stable point. The system can effectively track the random power disturbance of wind power and has stronger anti-disturbance ability and dynamic performance.

[0144] After the random fluctuation of wind power is connected, the change of transmission power of tie lines between interconnected systems is given by Fig.12As shown in the figure, after HESS participates in frequency modulation, the power change curve is relatively gentle, which plays a greater role in the overall stability of the system.

[0145] The output and SOC changes of energy storage batteries in single energy storage system and hybrid energy storage system modes are as follows: Figure 13-14 As shown in the figure, compared with the single energy storage auxiliary grid frequency regulation mode, the present invention utilizes the different power generation characteristics of the energy storage battery and the supercapacitor to complement each other's advantages to form the HESS, which effectively makes up for the economic disadvantages of the single energy storage system, reduces the shortening of the life of the energy storage battery due to frequent charging and discharging, and better meets the multi-level requirements of the modern power grid for energy and power.

[0146] In view of the load frequency control problem faced by the current power system under the background of "double high", this paper first built a two-region LFC model containing energy storage batteries and supercapacitors for hybrid energy storage auxiliary frequency regulation, and on this basis designed a FOPID controller based on chaotic particle swarm optimization, and deeply studied the joint optimization model of hybrid energy storage auxiliary frequency regulation and traditional unit frequency regulation. The simulation results show that:

[0147] 1) Compared with the traditional PID controller, the FOPID controller has better dynamic performance. The frequency regulation potential of the traditional unit can be fully utilized by optimizing its control parameters by using the improved ITAE as the objective function and the excellent performance of the chaotic particle swarm algorithm;

[0148] 2) By simulating the switching of loads in the power grid under step load disturbance conditions, compared with the frequency regulation of energy storage batteries alone, the frequency regulation of hybrid energy storage shows good frequency regulation potential and superiority, so as to better meet the increasing reliability requirements of the power grid;

[0149] 3) Taking into account the uncertainty of wind power conditions, hybrid energy storage control can still reduce battery loss while stabilizing frequency deviation in the case of random disturbances in wind power, and has strong applicability.

[0150] The feasibility and robustness of the control method proposed in this invention are demonstrated through simulation analysis. The hybrid energy storage stage control can effectively reduce the frequency deviation under different load disturbance conditions and improve the stability of the power system. This lays a certain foundation for the subsequent capacity configuration and economic evaluation of hybrid energy storage participating in grid frequency regulation.

[0151] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for establishing a joint control strategy for load frequency of a hybrid energy storage-assisted power grid, characterized in that The following steps are involved: S1) Analyze the load frequency control model; The step S1) analyzing the load frequency control model comprises: S11) System frequency response model The automatic generation control AGC adopts the tie line frequency deviation control TBC mode to efficiently stabilize the frequency deviation of the interconnected power grid, as shown in equations (1) and (2): ACE i =ΔP tiei,j +b i Δf i (1) Where ΔP tiei,j is the power exchange deviation of the tie line between area i and area j; Δf i is the system frequency deviation; β i is the response coefficient; R i and D i are the adjustment coefficient and load damping coefficient respectively; ACE i represents the error signal of region i; Using TBC mode, ACE i As the output feedback item of the LFC controller to achieve the purpose of maintaining the frequency stability of the interconnected power system; S12) LFC component model Build the LFC system, model the prime mover, speed governor, generator-load and tie line in the system, and select the reheat steam turbine as the prime mover. Its model is shown in formula (3): In the formula, K r is the proportion of steam power generated in the high-pressure cylinder section; T r is the reheater time coefficient; T t is the steam chamber time coefficient; s is the Laplace transform operator; The generator rate constraint is to limit the unit output by a given maximum change value. A reheat steam turbine unit model considering GRC is built, and the limiter value of the limiting link is set to ±0.0017puMW / s; Formula (4) is the transfer function model of the steam turbine governor. The governor dead zone is set for the unit, where the governor dead zone is less than 1r / min. After linearization, the governor model with dead zone can be obtained, as shown in Formula (5): Where, T g is the time constant of the speed regulator, N1 and N2 are linearization coefficients; The generator-load model describes the relationship between power change and frequency change in the interconnected power system, as shown in equation (6): Where ΔP m is the output power of the generator; ΔP i is the applied load disturbance; M i is the inertia constant of the generator; D i is the load damping coefficient; S13) Tie line model The tie line model is often used to cooperate with TBC for secondary frequency modulation, such as equation (1)-equation (2); S2) Build a hybrid energy storage combined with load frequency control LFC model: The step S2) of building a hybrid energy storage combined with LFC model includes: S21) Energy storage battery frequency modulation model The simplified energy storage power transfer function model is described in the mathematical model as shown in equations (7) to (12): In the formula, ACE i For the input control signal, select ACE i , that is, the error signal of region i is used as the control signal, and the current is selected as the response variable inside the battery; ΔU oc , ΔU serious , ΔU c , ΔU t and U b They are the battery’s open circuit voltage increment, internal resistance voltage increment, connection impedance voltage increment, transient voltage increment, and the battery terminal voltage value obtained by superimposing the above increments; ΔI b is the current increment flowing through the battery; T b is the time constant of the first-order inertia link of the battery; C0 and C p are the initial capacity and rated capacity of the battery respectively; R series is the internal resistance of the battery cell; n is the number of batteries in series; m is the number of subsystems in parallel; k is the number of energy storage units in parallel; R c is the connection impedance between the energy storage unit and the PCS; R t is the overvoltage resistance; C t is the overvoltage capacitance; ΔP b is the actual output value of the active power of the energy storage battery; Open circuit voltage U oc It is a function of SOC, usually expressed as a percentage. The battery SOC expression is as follows: In the formula, S(t) and S(t0) are the SOC of the energy storage battery at time t and the initial time t0 respectively; C is the rated capacity of the battery; η bat is the battery charge / discharge efficiency, representing the internal power dissipated during the charge / discharge cycle, SOC∈[0.2,0.8]; S22) Supercapacitor frequency modulation model The equivalent form of a supercapacitor is a parallel circuit of a resistor and a capacitor. A voltage feedback loop is introduced to keep it stable. The mathematical model of its transfer function model is described as shown in equations (14) to (16): ΔP sc =(U do +ΔU d )ΔI d (16) In the formula, ΔI d , ΔU d and ΔP sc are the increment of current, voltage and actual output value of active power flowing through the supercapacitor respectively; T C is the time constant; K CA ACE is the control signal i Gain factor; K vd is the voltage feedback loop gain coefficient; R is the equivalent resistance; C is the equivalent capacitance; U d0 is the initial voltage of the supercapacitor; S23) Hybrid energy storage combined frequency regulation model Combining the auxiliary frequency regulation model proposed in step S21) and S22), the joint frequency regulation model with the participation of hybrid energy storage can be obtained as shown in equations (17) and (18): ΔP m -ΔP i +ΔP hess =(M i s+D i )Δf i (17) ΔP hess =ΔP b +ΔP sc (18) In the formula, ΔP hess The actual output value increment of HESS active power; The energy storage battery frequency regulation model and the supercapacitor frequency regulation model are respectively connected to the IEEE standard two-region LFC model to obtain a joint frequency regulation model with the participation of hybrid energy storage. S3) Design a fractional-order proportional-integral-differential FOPID controller and use chaotic particle swarm optimization to tune its control parameters: The step S3) designs a fractional-order proportional-integral-differential FOPID controller and uses chaotic particle swarm optimization to adjust its control parameters, specifically: S31) FOPID control The continuous FOPID operator is defined as shown in formula (19): In the formula, α is the order of calculus, t and α are the upper and lower limits of calculus respectively; The transfer function expression of FOPID controller is: In the formula, K p , K i , K d are the proportional, integral and differential coefficients respectively, λ and μ are the integral order and differential order respectively; The FOPID control system model can be obtained by combining the above formula with the unit model. i and i are the input and output signals of the control system, u i is the output signal of the controller, ACE i =r i -y i is the regional control error signal, which is used as the output feedback term to maintain the stability of the interconnected power system. The research frequency band is set to [10 -3 , 10 3 ], the improved Oustaloup filtering algorithm is used to approximate the FOPID operator. A filter is added to the original Oustaloup approximation algorithm, so that it can achieve better approximation both within the frequency band and at the endpoints; S32) Optimization model (1) Optimizing the objective function By multiplying time by the absolute value integral of the error ITAE as the objective function, the ACE of area i is taken as the instantaneous error e(t) to establish the optimization model of the system, as shown in formula (21): Since there are only two variables, ACE and t, in formula (21), it is difficult to determine whether the error in the system is a positive error or a negative error. At the same time, the overshoot will be sacrificed to reduce the rise time of the system. Therefore, the objective function in formula (21) is modified and a penalty link is added to avoid the occurrence of excessive overshoot. The modified objective function is shown in formula (22): Where, t r is the rise time; ξ1, ξ2 are weight coefficients; The penalty function is used to achieve dynamic adjustment. When ξ1>>ξ2 in equation (22), the rise time of the system is sacrificed to reduce or eliminate the overshoot of the system. When ξ1<<ξ2 in equation (22), the overshoot of the system increases, but the rise time of the system decreases. (2) Optimize the process The particle swarm optimization algorithm searches for the motion state of particles from the individual and global perspectives, and continuously iterates and updates the particle speed and position by combining formulas (23)-(24), and finally obtains the local optimal solution and the global optimal solution: Where ω is the inertia weight, Respectively represent the particle's speed, position, individual optimal solution and global optimal solution; c1 and c2 are acceleration factors; r1 and r2 are random numbers between [0,1]; Chaotic particle swarm optimization is used to optimize Mapped to the domain of the Logistic equation through equation (25): The chaotic particle swarm optimization algorithm is used to optimize the FOPID control parameters online; The process of online optimization of FOPID control parameters using the chaotic particle swarm optimization algorithm in the optimization model of step S32) is as follows: 1) Initialize the particle swarm and set the particle swarm size and maximum number of iterations. 2) According to the particle position, assign a value to K p , K i and K d or K p , K i , K d , λ and μ; 3) Run the two-region LFC model and return the target value J gbest ; 4) Global optimal value x of particle swarm gbest Perform chaos optimization: ① Through formula (25) Mapped to the domain of the Logistic equation [0,1]; ② Using formula (26) Perform M iterations to obtain the chaotic sequence ③ According to formula (27), the chaotic sequence is inversely mapped back to the original solution space to obtain the feasible solution sequence x ki (1),x ki (2)……x ki (M); ④ Run the two-region LFC model to obtain the optimal solution J after chaos optimization g ' best =minJ(x k i (j)); ⑤ Randomly replace any particle group in the current particle group with the optimal particle group obtained in the chaotic sequence; ⑥ Update the objective function value, such as J g ' best <J gbest , then J gbest =J g ' best , 5) Update the particle velocity and position according to the individual optimal solution and the global optimal solution. Let k = k + 1. If k < K, return to step 2; otherwise, the obtained is the global optimal solution; 6) The corresponding parameters output are the optimal control parameters, and the solution is completed; S4) Simulation analysis verification: The simulation analysis verification content of step S4) includes: controller performance verification, step load disturbance verification and continuous disturbance verification taking into account wind power uncertainty.

Images