Frequency control method of microgrid based on wind-photovoltaic-fuel cell

Abstract

The present invention provides a method for frequency control of a microgrid based on wind-photovoltaic-fuel cell, the method comprising: using a linearized transfer function to model the power generation resources and energy storage system source of the microgrid respectively; integrating the power generation resources and the energy storage system to form a microgrid, and establishing a linearized model of a microgrid based on wind-photovoltaic-fuel cell; constructing a microgrid frequency controller for frequency control based on a linear auto-disturbance rejection control algorithm; and optimizing the performance of the microgrid frequency controller based on a bat optimization algorithm. The method of the present invention can effectively suppress the influence of wind speed and load disturbances on system frequency through a microgrid frequency controller designed based on a linear auto-disturbance rejection control algorithm.

Description

Technical Field

[0001] The present invention belongs to the technical field of microgrid frequency control, and in particular relates to a method for microgrid frequency control based on wind-photovoltaic-fuel cell. Background Art

[0002] Renewable energy sources such as solar photovoltaic, wind power, small hydropower, fuel cells, diesel engine generators, etc. have been introduced into the power grid system to meet the growing demand and minimize environmental pollution and transmission losses, thereby improving the reliability of power supply. However, these renewable energy sources are highly unpredictable and depend on climatic conditions. Therefore, they are integrated with fuel cells, backup diesel generators, and some energy storage systems (such as flywheel energy storage and battery energy storage) and supercapacitors to form a hybrid system for better operation and control. The uncertain characteristics of wind and photovoltaic power generation, or the mismatch between power generation and demand caused by changes in load, cause frequency deviations in the power system, which must be quickly controlled.

[0003] Currently, most of the works are conducted in conventional power systems with thermal / hydroelectric power stations. The number of frequency control studies is relatively less with the penetration of renewable energy and distributed generation. Frequency control of microgrids in the presence of flywheel energy storage and batteries in energy storage systems has been studied using fuzzy controllers based on PI and particle swarm optimization. Load-frequency control of wind-diesel and wind-diesel-micro-hydro hybrid systems has been studied to cope with step changes in wind and load in wind energy systems. A fuzzy logic-based controller has been designed for frequency regulation of a PV-diesel hybrid system in the presence of an energy storage system.

[0004] Although works on frequency control can be found in the literature based on heuristic optimization techniques, they are mainly studied in conventional thermal, hydroelectric systems where the only disturbance is in the form of load variation. However, in this study, the frequency control problem in microgrids in the presence of intermittent renewable energy sources such as wind and solar photovoltaics is studied. Here, in addition to the variations in load demand, the frequency controller is highly affected by variations in wind speed and solar intensity.

[0005] In view of the above problems, it is necessary to propose a method for frequency control of a microgrid based on wind-photovoltaic-fuel cell that is reasonably designed and effectively solves the above problems. Summary of the invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art and provide a method for frequency control of a microgrid based on wind-photovoltaic-fuel cell.

[0007] The present invention provides a method for frequency control of a microgrid based on wind-photovoltaic-fuel cell, the method comprising:

[0008] A linearized transfer function is used to model the power generation resource and the energy storage system source of the microgrid respectively;

[0009] Integrate the power generation resources and the energy storage system to form the microgrid, and establish a linearized model of the microgrid based on wind power, photovoltaic power and fuel cells;

[0010] A microgrid frequency controller for frequency control is constructed based on the linear active disturbance rejection control algorithm;

[0011] The performance of the microgrid frequency controller is optimized based on a bat optimization algorithm.

[0012] Optionally, the microgrid frequency controller for frequency control based on the linear active disturbance rejection control algorithm includes:

[0013] A second-order linear active disturbance rejection control algorithm is used to construct a microgrid frequency controller for frequency control, where:

[0014] Assume that the mathematical model of the microgrid frequency controller is:

[0015]

[0016] Among them, b 0 is the high-frequency gain of the controlled object, f(y,u,d) includes the internal uncertainty and external disturbance of the object;

[0017] Let z = [z 1 z 2 z 3 ] T , where z 1 =y, z 3 =f, then the mathematical model of the microgrid frequency controller is:

[0018]

[0019] Among them, E e =[0 0 1] T , C e =[1 0 0]

[0020] Optionally, the microgrid frequency controller for frequency control based on the linear active disturbance rejection control algorithm further includes:

[0021] The performance of the constructed microgrid frequency controller is optimized by evaluating the integral time absolute error of the system frequency deviation under different operating conditions; wherein,

[0022]

[0023]

[0024] Among them, st is the constraint condition, b 0min ,ω omi,n ω cm and b 0max ,ω omax ,ω cmax are the upper and lower limits of the parameters to be optimized, Δf(t) is the system frequency deviation, and ITAE is the integral time absolute error of the frequency deviation.

[0025] Optionally, optimizing the performance of the microgrid frequency controller based on a bat optimization algorithm includes:

[0026] Initialize the parameters of the bat optimization algorithm, where there are N populations, the individual loudness of the bats is a, the pulse emission rate is r, and the maximum number of iterations is t max ;

[0027] Initialize the population individuals of the bat optimization algorithm; wherein, multiple beacon points are evenly set in the search space to guide the bat optimization algorithm to find the correct search direction at the beginning of the iteration, and the multiple beacon points are set as follows:

[0028]

[0029]

[0030]

[0031] Wherein, L is the length of the search space, W is the width of the search space, and H is the height of the search space;

[0032] The fitness function of the bat optimization algorithm is set, and the fitness function is:

[0033]

[0034] Where Δf(t) is the system frequency deviation and ITAE is the time-integrated absolute error of the frequency deviation.

[0035] Optionally, the optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm further includes:

[0036] The performance parameters of the bat optimization algorithm are updated, wherein:

[0037] The position of a single bat is updated, and the update formula of the position of a single bat is:

[0038]

[0039] Among them, x i t is the position of the i-th bat in the t-th iteration, v i (t+1) is the speed of the ith bat in the (t+1)th iteration;

[0040] An adaptive weighting factor, a frequency parameter for controlling the local optimal flight direction, and a frequency parameter for controlling the global optimal flight direction are introduced to establish a speed frequency update formula, which is:

[0041]

[0042]

[0043] w v =w vmin +(w vmax +w vmin )S t-1 ,

[0044] Among them, Q i1 and Q i2 are the frequency parameters controlling the local and global optimal flight directions, W v is the adaptive weighting factor, v i (t+1) is the speed of the ith bat in the (t+1)th iteration, x pbest and x gbest are the local and global optimal positions respectively, rand1 and rand2 are random numbers between [0,1], and w v max and w v max are the maximum weight value and the minimum weight value respectively, S is the attenuation value, and k is a random number between [0,1].

[0045] Optionally, the optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm further includes:

[0046] The probability of generating a solution of the bat optimization algorithm is disturbed by chaotic mapping, where rand3 is assumed to be a random number generated between the set [0,1]. If rand3>r n is satisfied, then:

[0047] The position coordinates of each new individual are mapped by a chaotic sequence generated by a chaotic logic diagram {zy|y=1,2,3,...,N}, which is defined as:

[0048]

[0049] Among them, Z y is the yth element of the chaotic sequence, Z y+1 is the y+1th element of the chaotic sequence, Rand() is a random number between [0,1], and η is the control factor that controls the chaotic intensity factor;

[0050] Generates the interference value corresponding to each individual, expressed as:

[0051] D(1:y)=2(t max -t)Rand(1:y) / t max ,

[0052] Where D is the disturbance value generated, (1:y) represents y individuals from 1 to y, t max is the maximum value of chaos time;

[0053] If only the position coordinates of the new individual are disturbed and the other parameters remain unchanged, it can be expressed as:

[0054] pop new (1:y)=p best (1:y)-D(1:y)+2D(1:y)z(1:y),

[0055] Among them, pop new It is a new individual after the disturbance.

[0056] Optionally, the optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm further includes:

[0057] An individual selection strategy is performed to update the impulse response and loudness; wherein,

[0058] Assume that rand4 is a random number in the interval [0,1]. If rand4>a n and f(x new ) <f(x n ), then accept the new solution generated by the perturbation operation, then the impulse response and the loudness will be updated simultaneously, and the formulas of the impulse response and the loudness are respectively:

[0059]

[0060]

[0061] Among them, r i t is the impulse response, a i t r is loudness, i t+1is the impulse loudness value of the ith bat at the t+1th iteration, γ is the constant coefficient of the impulse rate, and a i t+1 is the response value of the i-th bat at the t+1th iteration, and ε is the attenuation coefficient of the loudness.

[0062] Optionally, the optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm further includes:

[0063] The bat optimization algorithm is iterated to obtain parameters of the microgrid frequency controller.

[0064] Optionally, the adopting of a linearized transfer function to model the power generation resources and energy storage system source of the microgrid respectively includes:

[0065] The random wind speed is modeled based on the Van der Hoeven model, where the simulated wind speed w(t) is:

[0066] in,

[0067]

[0068] Among them, A p is the amplitude, ω p is the discrete angular frequency, Z w (ω p ) is the power spectral density,

[0069] φ p is the phase;

[0070] The power generation of wind, photovoltaic, fuel cell and diesel generator is modeled using linear first-order transfer functions, where

[0071] Wind power generation G WPGK (s) and photovoltaic power generation G PV The linear first-order transfer functions of (s) are:

[0072]

[0073]

[0074] Among them, K WPG and K PV is the gain constant, T WPG and T PV are the time constants of wind and photovoltaic, respectively;

[0075] Water electrolyzer G AE (s), fuel cell G FC (s) and diesel generator G DEGThe linear first-order transfer functions of (s) are:

[0076]

[0077]

[0078]

[0079] Among them, K AE , K FC and K DGE are gain constants, T AE 、T FC and T DGE are the time constants of water electrolyzer, fuel cell and diesel generator respectively, and s is in s domain;

[0080] The flywheel, battery, and supercapacitor energy storage system is modeled using a linear first-order transfer function, where

[0081] Flywheel energy storage system G FES 、Battery energy storage system G BES and supercapacitor energy storage system G UC The linear transfer functions are:

[0082]

[0083]

[0084]

[0085] Among them, K FES , K BES and K UC are gain constants, T FES 、T BES and T UC are the time constants of the flywheel energy storage system, battery energy storage system and supercapacitor energy storage system, respectively, and s is in the s domain;

[0086] Optionally, the power generation resources and the energy storage system are integrated to form the microgrid, and a linearized model of the microgrid based on wind power, photovoltaic power and fuel cell is established, including:

[0087] The vertical power generation capacity P of the microgrid MG for:

[0088] P MG =P T +P DEG +P FCG +P PVG ±P FES ±(P BES or PUC ),

[0089] Among them, P T =P WPG +P PVPG -P AE ,

[0090] P T is the net power of wind turbines and photovoltaic power generation minus the power occupied by water electrolyzers, P DEG is the power generation of the diesel generator, P FCG is the power generation of the fuel cell, P PVG is the photovoltaic power generation, P FES is the power generation of the flywheel energy storage system, P BES is the power generation of the battery energy storage system, P UC Power generation for supercapacitor energy storage system;

[0091] Power balance between supply and demand ΔP e for:

[0092] ΔP e =P MG -P L , where P L is the total electricity demand;

[0093] The frequency change Δf is:

[0094] Among them, K SC is the characteristic constant of the microgrid;

[0095] The transfer function of the system is:

[0096]

[0097] Where M = K sys and D = K sys T sys are the equivalent inertia constant and damping constant of the system respectively, and s is the s domain.

[0098] The method for frequency control of a microgrid based on wind-photovoltaic-fuel cell provided by the present invention adopts a linearized transfer function to model the power generation resources and energy storage system sources of the microgrid respectively; integrates the power generation resources and the energy storage system to form a microgrid, and establishes a linearized model of a microgrid based on wind-photovoltaic-fuel cell; constructs a microgrid frequency controller for frequency control based on a linear self-disturbance rejection control algorithm; and optimizes the performance of the microgrid frequency controller based on a bat optimization algorithm. The microgrid frequency controller designed by the present invention based on the linear self-disturbance rejection control algorithm can effectively suppress the influence of wind speed and load disturbances on the system frequency. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 A schematic flow chart of a method for frequency control of a microgrid based on wind-photovoltaic-fuel cell according to an embodiment of the present invention;

[0100] Figure 2 It is a structural schematic diagram of a microgrid linearization model according to another embodiment of the present invention. DETAILED DESCRIPTION

[0101] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0102] like Figure 1 As shown, one aspect of the present invention provides a method S100 for frequency control of a microgrid based on wind-photovoltaic-fuel cell, the method S100 comprising:

[0103] S110, using linearized transfer functions to respectively model the power generation resources and energy storage system sources of the microgrid.

[0104] First, the random wind speed is modeled based on the van der Hoeven model.

[0105] Specifically, the wind speed varies according to the climatic conditions and can therefore be modeled as a nonlinear time-varying function. Therefore, the wind speed can be modeled based on the van der Hoven model. In this embodiment, the power spectrum of the horizontal wind speed is assumed to be in the range of 0.0007 to 900 times / hour. This frequency range contains the medium-term and long-term changes in the turbulent component and the spectrum range. The discrete angular frequency is recorded as ω p ,p=1to(K+1) and the corresponding power spectrum density is recorded as Z vv (ω p ). Based on the van der Hoeven model, the frequency ω p The harmonics at have amplitudes:

[0106]

[0107] Among them, A p is the amplitude, ω p is the discrete angular frequency, Z w (ω p ) is the power spectral density,

[0108] φ p is the phase, φ p is randomly generated with uniform distribution in the domain [π, -π].

[0109] The wind speed w(t) is simulated as:

[0110]

[0111] Among them, ω 0 =0,φ 0 =0 and is the average wind speed, calculated over a time span greater than the maximum period in the van der Hoeven characteristic (i.e., T = 2π / ω 1 In this embodiment, K=55 is used to perform the sampling operation of the Van der Hoven characteristic, and the following discrete frequencies are considered: f p = p·10 k Cycles / hour, p=1,2,.....,9, k=-3,-2,-1,0,1,2.

[0112] Secondly, a linear first-order transfer function is used to model the power generation of wind, photovoltaic, fuel cell, and diesel generators, where

[0113] Wind power generation G WPGK (s) and photovoltaic power generation G PV The linear first-order transfer functions of (s) are:

[0114]

[0115]

[0116] Among them, K WPG and K PV is the gain constant, T WPG and T PV are the time constants of wind and photovoltaic, respectively, and s is the s domain;

[0117] Water electrolyzer G AE (s), fuel cell G FC (s) and diesel generator G DEG The linear first-order transfer functions of (s) are:

[0118]

[0119]

[0120]

[0121] Among them, K AE , K FC and K DGE are gain constants, T AE 、T FC and T DGE are the time constants of water electrolyzer, fuel cell and diesel generator respectively;

[0122] The flywheel, battery, and supercapacitor energy storage system is modeled using a linear first-order transfer function, where

[0123] Flywheel energy storage system G FES 、Battery energy storage system G BES and supercapacitor energy storage system G UC The linear transfer functions are:

[0124]

[0125]

[0126]

[0127] Among them, K FES , K BES and K UC are gain constants, T FES 、T BES and T UC are the time constants of the flywheel energy storage system, battery energy storage system and supercapacitor energy storage system respectively, and s is the s domain.

[0128] S120, integrating the power generation resources and the energy storage system to form the microgrid, and establishing a microgrid linearization model based on wind power-photovoltaic-fuel cell.

[0129] like Figure 1 As shown, power generation resources and storage systems can be integrated into a microgrid, based on a linearized model of a wind-photovoltaic-fuel cell microgrid.

[0130] Power resources such as wind, photovoltaic, fuel cell and diesel generator are combined with water electrolyzers, which convert part of the electricity generated by wind or / and photovoltaic into hydrogen, which is supplied to the fuel cell as input. Two different energy storage combinations: flywheel energy storage system-battery energy storage system and flywheel energy storage system-supercapacitor are considered to be the best options for better frequency control and power management under different operating conditions. The combination of wind power-photovoltaic-fuel cell-diesel generator greatly reduces power management issues by coordinating their respective advantages and disadvantages. In addition, in order to improve the reliability of power supply, energy storage devices such as battery energy storage, flywheel energy storage and supercapacitors are integrated into microgrid systems, which greatly helps to contribute insufficient power during peak load demand.

[0131] Among them, the vertical power generation of the microgrid P MG for:

[0132] P MG =P T +P DEG +P FCG +P PVG ±P FES ±(P BES or PUC ), (7)

[0133] Among them, P T =P WPG +P PV -P AE , (8)

[0134] P T is the net power of wind turbines and photovoltaic power generation minus the power occupied by water electrolyzers, P DEG is the power generation of the diesel generator, P FCG is the power generation of the fuel cell, P PVG is the photovoltaic power generation, P FES is the power generation of the flywheel energy storage system, P BES is the power generation of the battery energy storage system, P UC Power generation for supercapacitor energy storage system;

[0135] Power balance between supply and demand ΔP e for:

[0136] ΔP e =P MG -P L , (9)

[0137] Among them, P L is the total electricity demand;

[0138] The frequency change Δf is:

[0139]

[0140] Among them, K SC is the characteristic constant of the microgrid;

[0141] The transfer function of the system is:

[0142]

[0143] Where M = K sys and D = K sys T sys are the equivalent inertia constant and damping constant of the system respectively, and s is the s domain.

[0144] In this embodiment, the parameters of the microgrid are shown in Table 1.

[0145] Table 1 Microgrid parameters

[0146]

[0147]

[0148] In this case, the frequency deviation of the microgrid using two energy storage combinations, flywheel energy storage system-battery energy storage system and flywheel energy storage system-supercapacitor, is analyzed under the condition of changing wind speed and load demand. The parameters of the microgrid are in pu and the wind speed is in meters per second. The step change of load and wind speed curves is set as follows:

[0149]

[0150]

[0151] S130. Construct a microgrid frequency controller for frequency control based on a linear active disturbance rejection control algorithm.

[0152] 1) Design process of microgrid frequency controller.

[0153] In the present invention, a second-order linear active disturbance rejection control (LADRC) algorithm is used to construct a microgrid frequency controller for frequency control, wherein:

[0154] Assume that the mathematical model of the microgrid frequency controller is:

[0155]

[0156] Among them, b 0 is the high-frequency gain of the controlled object, f(y,u,d) includes the internal uncertainty and external disturbance of the object;

[0157] Let z = [z 1 z 2 z 3 ] T , where z 1 =y, z 3 =f, then the mathematical model of the microgrid frequency controller is:

[0158]

[0159] in, Design a linear extended state observer for formula (14):

[0160]

[0161] in, is the estimated value of z, L 0 is the observer gain, which can be expressed as:

[0162] L o =[β 1 β 2 β 3 ] T(16)

[0163] When A e —L o When Ce is asymptotically stable, The total disturbance can be estimated. In order to eliminate the total disturbance through the feedback loop, select

[0164]

[0165] Among them, K 0 is the feedback control gain, which can be expressed as:

[0166] K o =[k p k d 1] / b 0 (18)

[0167] For the expanded reference input signal, it can be expressed as:

[0168]

[0169] The parameter that needs to be adjusted for the second-order LADRC is the feedback controller gain K 0 and the observer gain L of the ESO 0 In formulas (16) and (18), by adjusting the controller bandwidth and observer bandwidth ω c ,ω 0 Get the gain K 0 , L 0 Each element of , we get:

[0170] β 1 =3ω o ,

[0171]

[0172] 2) Optimization framework of microgrid frequency controller.

[0173] The performance of the constructed microgrid frequency controller is optimized by evaluating the integral time absolute error of the system frequency deviation under different operating conditions; wherein,

[0174]

[0175]

[0176] Among them, st is the constraint condition, b 0min ,ω omin ,ω cmin and b 0max ,ω omax ,ωcmax are the upper and lower limits of the parameters to be optimized, Δf(t) is the system frequency deviation, and ITAE is the integral time absolute error of the frequency deviation.

[0177] In this embodiment, under the microgrid configuration and load-wind speed step input, the performance optimization problem of the microgrid frequency controller can be expressed as follows:

[0178]

[0179]

[0180] S140. Optimizing the performance of the microgrid frequency controller based on a bat optimization algorithm.

[0181] The bat algorithm is a heuristic search algorithm proposed by Professor Yang in 2010 based on swarm intelligence. It is an effective method for searching the global optimal solution. It is a random search algorithm that simulates the use of sonar by bats in nature to detect prey and avoid obstacles, that is, it simulates the most basic detection and positioning capabilities of bats using ultrasound to detect obstacles or prey and links it with the optimization target function. The algorithm is an iterative optimization technique that is initialized as a set of random solutions, then searches for the optimal solution through iteration, and generates local new solutions around the optimal solution through random flight, which strengthens the local search. Compared with other algorithms, the bat algorithm is far superior to other algorithms in terms of accuracy and effectiveness, and there are not many parameters to adjust.

[0182] Considering the poor search capability of the traditional bat algorithm, the present invention integrates the adaptive weight coefficient, local frequency parameter and chaos algorithm into the bat algorithm. The improved hybrid bat algorithm includes the following steps:

[0183] 1) Initialize the bat optimization algorithm parameters.

[0184] The search space of the improved hybrid bat algorithm is defined in the space of L×W×H. L, W, and H represent the length, width, and height of the space, respectively. The position coordinates (x, y, z) of the bat are constrained in the three-dimensional space. In the initialization of the improved hybrid bat algorithm, there are N populations, the individual loudness of the bat is a, the pulse emission rate is r, and the maximum number of iterations is t. max .

[0185] 2) Initialize the population individuals of the bat optimization algorithm.

[0186] The complexity and stability of the algorithm are related to the number of initialized individuals. At present, in most literature, the first generation of population is often generated by random initialization. This initialization method can easily lead to sparse and uneven distribution of individuals. To address this problem, multiple beacon points are evenly set in the search space to guide the algorithm to quickly find the correct search direction at the beginning of the iteration. The setting of multiple beacon points is as follows:

[0187]

[0188]

[0189]

[0190] Wherein, L is the length of the search space, W is the width of the search space, and H is the height of the search space.

[0191] Compared with the random distribution of all points at initialization, in the first iteration, all individual bats are more likely to search in better directions.

[0192] 3) Setting the fitness function of the bat optimization algorithm.

[0193] The fitness function is:

[0194]

[0195] Where Δf(t) is the system frequency deviation and ITAE is the time-integrated absolute error of the frequency deviation.

[0196] 4) Update the performance parameters of the bat optimization algorithm.

[0197] The performance of the improved hybrid bat algorithm is related to three parameters, including the bat's location information, speed and frequency.

[0198] Update the position of a single bat. The update formula for a single bat is:

[0199]

[0200] Among them, x i t is the position of the i-th bat in the t-th iteration, v i (t+1) is the speed of the ith bat in the (t+1)th iteration.

[0201] In order to better control the flight direction of a single bat, local position information is introduced to update the speed. i1 and Q i2are the frequency parameters that control the local and global optimal flight directions respectively. In order to make the algorithm search quickly in the early stage and accurately in the later stage, an adaptive weighting factor W is introduced v , Q i1 and Q i2 To establish the speed frequency update formula, as shown below:

[0202] The speed frequency update formula is:

[0203]

[0204]

[0205] w v =w v min +(w v max +w v min )S t-1 , (26)

[0206] Among them, Q i1 and Q i2 are the frequency parameters controlling the local and global optimal flight directions, W v is the adaptive weighting factor, v i (t+1) is the speed of the ith bat in the (t+1)th iteration, x pbest and x gbest are the local and global optimal positions respectively, rand1 and rand2 are random numbers between [0,1], and w v max and w v max are the maximum weight value and the minimum weight value respectively, S is the attenuation value, and k is a random number between [0,1].

[0207] 5) The probability of generating a solution by the bat optimization algorithm is disturbed by chaotic mapping.

[0208] The chaotic map perturbs the probability of generating a solution. Assume that rand3 is a random number generated between the set [0,1]. If rand3>r n is satisfied, then:

[0209] The position coordinates of each new individual are mapped by a chaotic sequence generated by a chaotic logic diagram {zy|y=1,2,3,...,N}, which is defined as:

[0210]

[0211] Among them, Z y is the yth element of the chaotic sequence, Z y+1is the y+1th element of the chaotic sequence, Rand() is a random number between [0,1], and η is the control factor that controls the chaotic intensity factor;

[0212] Generates the interference value corresponding to each individual, expressed as:

[0213] D(1:y)=2(t max -t)Rand(1:y) / t max , (28)

[0214] Where D is the disturbance value generated, (1:y) represents y individuals from 1 to y, t max is the maximum value of chaos time;

[0215] If only the position coordinates of the new individual are disturbed and the other parameters remain unchanged, it can be expressed as:

[0216] pop new (1:y)=p best (1:y)-D(1:y)+2D(1:y)z(1:y), (29)

[0217] Among them, pop new It is a new individual after the disturbance.

[0218] 6) Perform individual selection strategy to update the impulse response and loudness.

[0219] Assume that rand4 is a random number in the interval [0,1]. If rand4>a n and f(x new ) <f(x n ), then accept the new solution generated by the perturbation operation, then the impulse response and the loudness will be updated simultaneously, and the formulas of the impulse response and the loudness are respectively:

[0220]

[0221]

[0222] Among them, r i t is the impulse response, a i t r is loudness, i t+1 is the impulse loudness value of the ith bat at the t+1th iteration, γ is the constant coefficient of the impulse rate, and a i t+1 is the response value of the i-th bat at the t+1th iteration, and ε is the attenuation coefficient of the loudness.

[0223] 7) Iterating the bat optimization algorithm.

[0224] In this embodiment, the maximum number of iterations is regarded as the termination condition of the algorithm. If the number of iterations does not reach the maximum number of iterations, steps 3)-6) are repeated until the number of iterations reaches the maximum number of iterations (the maximum number of iterations is 200). After the optimization is completed, the optimal parameters of the microgrid frequency controller are obtained.

[0225] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for frequency control of a microgrid based on wind-photovoltaic-fuel cell, characterized in that: The method comprises: A linearized transfer function is used to model the power generation resource and the energy storage system source of the microgrid respectively; Integrate the power generation resources and the energy storage system to form the microgrid, and establish a linearized model of the microgrid based on wind power, photovoltaic power and fuel cells; A microgrid frequency controller for frequency control is constructed based on the linear active disturbance rejection control algorithm; The performance of the microgrid frequency controller is optimized based on the bat optimization algorithm; wherein, The adopting of linearized transfer functions to model the power generation resources and energy storage system sources of the microgrid respectively includes: The random wind speed is modeled based on the Van der Hoeven model, where the simulated wind speed w(t) is: in, Among them, A p is the amplitude, ω p is the discrete angular frequency, Z w (ω p ) is the power spectral density, φ p is the phase; The power generation of wind, photovoltaic, fuel cell and diesel generator is modeled using linear first-order transfer functions, where Wind power generation G WPGK (s) and photovoltaic power generation G PV The linear first-order transfer functions of (s) are: Among them, K WPG and K PV is the gain constant, T WPG and T PV are the time constants of wind and photovoltaic, respectively, and s is the s domain; Water electrolyzer G AE (s), fuel cell G FC (s) and diesel generator G DEG The linear first-order transfer functions of (s) are: Among them, K AE , K FC and K DGE are gain constants, T AE , T FC and T DGE are the time constants of water electrolyzer, fuel cell and diesel generator respectively; The flywheel, battery, and supercapacitor energy storage system is modeled using a linear first-order transfer function, where Flywheel energy storage system G FES 、Battery energy storage system G BES and supercapacitor energy storage system G UC The linear transfer functions are: Among them, K FES , K BES and K UC are gain constants, T FES , T BES and T UC are the time constants of the flywheel energy storage system, battery energy storage system and supercapacitor energy storage system, respectively, and s is in the s domain; The step of integrating the power generation resources and the energy storage system to form the microgrid and establishing a linearized model of the microgrid based on wind power, photovoltaic power and fuel cells includes: The vertical power generation capacity P of the microgrid MG for: P MG =P T +P DEG +P FCG +P PVG ±P FES ±(P BES or P UC ), Among them, P T =P WPG +P PV -P AE , P T is the net power of wind turbines and photovoltaic power generation minus the power occupied by water electrolyzers, P DEG is the power generation of the diesel generator, P FCG is the power generation of the fuel cell, P PVG is the photovoltaic power generation, P FES is the power generation of the flywheel energy storage system, P BES is the power generation of the battery energy storage system, P UC Power generation for supercapacitor energy storage system; Power balance between supply and demand ΔP e for: ΔP e =P MG -P L , where P L is the total electricity demand; The frequency change Δf is: Among them, K SC is the characteristic constant of the microgrid; The transfer function of the system is: Where M = K sys and D = K sys T sys are the equivalent inertia constant and damping constant of the system respectively, and s is the s domain.

2. The method according to claim 1, characterized in that The microgrid frequency controller for frequency control based on the linear active disturbance rejection control algorithm comprises: A second-order linear active disturbance rejection control algorithm is used to construct a microgrid frequency controller for frequency control, where: Assume that the mathematical model of the microgrid frequency controller is: Where b0 is the high-frequency gain of the controlled object, and f(y,u,d) contains the internal uncertainty and external disturbance of the object; Let z=[z1 z2 z3] T , where z1=y, z3=f, then the mathematical model of the microgrid frequency controller is: Among them, E e =[001] T , C e =[1 0 0].

3. The method according to claim 1, characterized in that The microgrid frequency controller for frequency control based on the linear active disturbance rejection control algorithm further includes: The performance of the constructed microgrid frequency controller is optimized by evaluating the integral time absolute error of the system frequency deviation under different operating conditions; wherein, Among them, st is the constraint condition, b 0min ,ω omin ,ω cmin and b 0max ,ω omax ,ω cmax are the upper and lower limits of the parameters to be optimized, Δf(t) is the system frequency deviation, and ITAE is the integral time absolute error of the frequency deviation.

4. The method according to claim 3, characterized in that The performance of the microgrid frequency controller is optimized based on a bat optimization algorithm, including: Initialize the parameters of the bat optimization algorithm, where there are N populations, the individual loudness of the bats is a, the pulse emission rate is r, and the maximum number of iterations is t max ; Initialize the population individuals of the bat optimization algorithm; wherein, multiple beacon points are evenly set in the search space to guide the bat optimization algorithm to find the correct search direction at the beginning of the iteration, and the multiple beacon points are set as follows: Wherein, L is the length of the search space, W is the width of the search space, and H is the height of the search space; The fitness function of the bat optimization algorithm is set, and the fitness function is: Where Δf(t) is the system frequency deviation and ITAE is the time-integrated absolute error of the frequency deviation.

5. The method according to claim 4, characterized in that The optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm also includes: The performance parameters of the bat optimization algorithm are updated, wherein: The position of a single bat is updated, and the update formula of the position of a single bat is: Among them, x i t is the position of the i-th bat in the t-th iteration, v i (t+1) is the speed of the ith bat in the (t+1)th iteration; An adaptive weighting factor, a frequency parameter for controlling the local optimal flight direction, and a frequency parameter for controlling the global optimal flight direction are introduced to establish a speed frequency update formula, which is: w v =w vmin +(w vmax +w vmin )S t-1 , Among them, Q i1 and Q i2 are the frequency parameters controlling the local and global optimal flight directions, W v is the adaptive weighting factor, v i (t+1) is the speed of the ith bat in the (t+1)th iteration, x pbest and x gbest are the local and global optimal positions respectively, rand1 and rand2 are random numbers between [0,1], and w v max and w v max are the maximum weight value and the minimum weight value respectively, S is the attenuation value, and k is a random number between [0,1].

6. The method according to claim 5, characterized in that The optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm also includes: The probability of generating a solution of the bat optimization algorithm is disturbed by chaotic mapping, where rand3 is assumed to be a random number generated between the set [0,1]. If rand3>r n is satisfied, then: The position coordinates of each new individual are mapped by a chaotic sequence generated by a chaotic logic diagram {zy|y=1,2,3,...,N}, which is defined as: Among them, Z y is the yth element of the chaotic sequence, Z y+1 is the y+1th element of the chaotic sequence, Rand() is a random number between [0,1], and η is the control factor that controls the chaotic intensity factor; Generates the interference value corresponding to each individual, expressed as: D(1:y)=2(t max -t)Rand(1:y) / t max , Where D is the disturbance value generated, (1:y) represents y individuals from 1 to y, t max is the maximum value of chaos time; If only the position coordinates of the new individual are disturbed and the other parameters remain unchanged, it can be expressed as: pop new (1:y)=p best (1:y)-D(1:y)+2D(1:y)z(1:y), Among them, pop new It is a new individual after the disturbance.

7. The method according to claim 6, characterized in that The optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm also includes: An individual selection strategy is performed to update the impulse response and loudness; wherein, Assume that rand4 is a random number in the interval [0,1]. If rand4>a n and f(x new ) <f(x n ), then accept the new solution generated by the perturbation operation, then the impulse response and the loudness will be updated simultaneously, and the formulas of the impulse response and the loudness are respectively: r i t+1 ＝r i 0 (1-e -γt ) Among them, r i t is the impulse response, a i t r is loudness, i t+1 is the impulse loudness value of the ith bat at the t+1th iteration, γ is the constant coefficient of the impulse rate, and a i t+1 is the response value of the i-th bat at the t+1th iteration, and ε is the attenuation coefficient of the loudness.

8. The method according to claim 7, characterized in that The optimizing the performance of the microgrid frequency controller based on the bat optimization algorithm also includes: The bat optimization algorithm is iterated to obtain parameters of the microgrid frequency controller.

Images