MPSO-based capacity optimization configuration method for wind-solar-hydrogen storage microgrid system

Abstract

The present application relates to the technical field of energy storage system, and particularly relates to a wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO, which comprises setting a parameter acquisition period, acquiring power generation parameters in a wind power generation device and a photovoltaic power generation device based on the acquisition period, and synchronously acquiring tie-line power parameters based on the acquisition period; filtering the acquired power generation parameters and tie-line power parameters; obtaining the filtered power generation parameters and tie-line power parameters, and combining the two to build a learning model through an MPSO algorithm; the present application can more reasonably configure the capacity of a wind-solar-hydrogen storage micro-grid system, reliably measure the power generation power of the device, hydrogen storage system parameters, load demand and other conditions according to specific conditions, establish an improved multi-objective optimization particle swarm algorithm model, and take the overall benefits of all parties, the economy and reliability of micro-grid system operation as optimization targets to reliably configure the capacity of the micro-grid system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage systems, and particularly relates to a wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO. BACKGROUND

[0002] The wind-solar-hydrogen storage micro-grid system is a micro-grid that integrates wind power generation, solar photovoltaic power generation, hydrogen power generation, and energy storage systems. This system uses wind energy and solar energy, two clean energy sources, to generate electricity, and uses the excess electricity generated to electrolyze water to produce hydrogen, storing energy in the form of hydrogen, which is more convenient. When the demand for electricity increases or the amount of renewable energy generated is insufficient, the stored hydrogen can be converted into electricity through a fuel cell to meet the demand for electricity. This system can improve energy utilization efficiency, enhance the stability and reliability of the power grid. In addition, the wind-solar-hydrogen storage micro-grid system can be self-sufficient, and can also transmit excess electricity back to the power grid, or obtain electricity from the power grid when the power is insufficient, to maintain a balance between supply and demand.

[0003] In recent years, with increasing attention to new energy generation, people are increasingly focusing on how to stabilize the instability of wind and solar power generation to reduce energy waste and power outages. In order to address these challenges, the wind-solar-hydrogen storage micro-grid system has emerged, which effectively improves the utilization efficiency and economic benefits of renewable energy by integrating wind, solar and hydrogen energy, while promoting clean production and storage of energy. Although this system has great potential, it still faces challenges in terms of technical maturity, and further research and innovation are needed to enhance its performance and stability.

[0004] When operating the system, a core problem is how to determine its optimal capacity configuration, which requires careful design based on actual energy demand. Currently, the research on capacity optimization configuration of the system is not perfect, and many studies mainly focus on the overall system benefits.

[0005] Therefore, a wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO is proposed. SUMMARY

[0006] In view of the above-mentioned shortcomings of the prior art, the present application provides a wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO, which solves the technical problems raised in the background art.

[0007] To achieve the above purpose, the present application realizes the following technical scheme:

[0008] The wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO comprises:

[0009] The parameter acquisition period is set, the power generation parameters are acquired based on the acquisition period in the wind power generation device and the photovoltaic power generation device, and the tie line power parameters are acquired synchronously based on the acquisition period; the acquired power generation parameters and tie line power parameters are filtered; the filtered power generation parameters and tie line power parameters are acquired, and a learning model is built by combining the two through the MPSO algorithm; a wind-solar-hydrogen storage micro-grid system capacity configuration target is set, and the parameters of the configuration target are imported into the learning model; the learning model is started, and the iterative search process of the particle swarm is repeated to continuously find an optimal capacity configuration scheme meeting the target.

[0010] Further, the power generation parameters include wind power generation parameters and photovoltaic power generation parameters, the wind power generation parameters include wind speed, wind direction, wind wheel rotating speed, power output, air density and turbulence intensity, and the photovoltaic power generation parameters include light intensity, temperature, open-circuit voltage of a photovoltaic cell, short-circuit current, maximum power point power and filling factor, and the tie line power parameters include active power, power direction, power fluctuation rate and power factor.

[0011] In the formula, the parameter acquisition period is set based on historical power generation parameters.

[0012] Further, the setting logic of the parameter acquisition period is represented as:

[0013]

[0014] In the formula, T is the parameter acquisition period; ω, collect is a weight; F is a power generation equipment fluctuation factor; R is a power generation equipment response factor; and S is a factor based on storage limitation. γ is a weight; F is a power generation equipment fluctuation factor; R is a power generation equipment response factor; and S is a factor based on storage limitation.

[0015] In the formula, ω, γ are positive numbers, and the sum of the weights ω, γ is 1, the weight values are defined by a user terminal, and the initial settings of the weights are 0.5, 0.3 and 0.2.

[0016] Further, the calculation formulae of F, R and S are as follows:

[0017]

[0018] In the formula, α and β are weights; σ w and σ s are wind power fluctuation standard deviation and photovoltaic power fluctuation standard deviation; are average power of wind power generation in a historical time period and average power of photovoltaic power generation in the historical time period; T w and T s are cumulative response time of a wind power generation equipment and cumulative response time of a photovoltaic power generation equipment; and M maxis the maximum data storage amount per unit time; m is the average storage byte number of a single data point; T is the total time length of the historical time period of the wind power generation and the historical time period of the photovoltaic power generation tolal is the corresponding time length of the union of the historical time period of the wind power generation and the historical time period of the photovoltaic power generation;

[0019] wherein the values of the weights a and b are defined by the user end user, the weights a and b are both in the range of [0, 1], the historical time period of the wind power generation and the historical time period of the photovoltaic power generation are defined by the user end, M max , and m are derived from the storage device used after the power generation parameters are collected.

[0020] Further, before the power generation parameters and the tie-line power parameters are subjected to the filtering processing, the parameters are subjected to initial processing through any one of a state transition matrix, an observation matrix, a process noise covariance matrix, and a measurement noise covariance matrix, and after the parameters are subjected to the initial processing, the parameters subjected to the initial processing are further subjected to the filtering processing based on a Kalman filtering algorithm.

[0021] Further, when the power generation parameters and the tie-line power parameters subjected to the filtering processing are processed by the MPSO algorithm, the parameters are optimized by introducing a dynamic inertia weight or adding a Gaussian disturbance, and after the MPSO algorithm completes the optimization, the corresponding learning model is built;

[0022] wherein the optimization of the MPSO algorithm is preferentially performed by introducing the dynamic inertia weight, and the optimization steps include:

[0023] a dynamic inertia weight function is set:

[0024] ω′=ω max -(ω max -ω min )*(iter / iter max );

[0025] In the formula: ω′ is the current inertia weight; ω max is the set maximum value of the inertia weight; ω min is the set minimum value of the inertia weight; iter is the current iteration number; iter max is the total maximum iteration number;

[0026] Further, the dynamic inertia weight obtained based on the dynamic inertia weight function is embedded into the particle velocity update part of the MPSO algorithm, that is, introduced into the particle velocity update formula:

[0027]

[0028] In the formula, v[i][j] is the speed of the ith particle in the jth dimensional space; c1 and c2 are learning factors; rand() is a random number generating function; pbest[i][j] and gbest[i][j] are the jth dimensional component of the individual optimal position of the ith particle and the jth dimensional component of the group optimal position; and x[i][j] is the jth dimensional component of the current position of the ith particle.

[0029] wherein rand(), ω max , ω max are defined by the user terminal.

[0030] Further, the step of building the learning model comprises:

[0031] In the learning model, the position vector of the particle is set as the capacity size of the component of the wind-solar-hydrogen storage micro-grid system, the speed vector of the corresponding particle is used to control the moving state of the particle in the search space, the fitness function is constructed according to the target expected to be achieved by the system capacity configuration, and the fitness function is used to determine the good or bad degree of the capacity configuration scheme represented by different particles.

[0032] The value range of the penalty factor and the learning factor is set.

[0033] The preprocessed data is divided into a training set and a validation set according to a suitable proportion, the validation set is applied to verify the model, the fitting degree between the capacity configuration result output by the model and the actual expected result is viewed, the penalty factor and the learning factor of the model are coordinated according to the verification result, and finally the learning model is output.

[0034] The component of the wind-solar-hydrogen storage micro-grid system includes a wind turbine, a photovoltaic panel and a hydrogen storage device, and the target expected to be achieved by the system capacity configuration includes cost control and power supply reliability guarantee. The cost control in the target expected to be achieved by the system capacity configuration includes power value and cost, and the power supply reliability guarantee in the target expected to be achieved by the system capacity configuration includes the fault frequency of the wind-solar-hydrogen storage micro-grid system and the average energy storage of the wind-solar-hydrogen storage micro-grid system.

[0035] Further, the fitness function is:

[0036]

[0037] In the formula, F(x) is the fitness value; C(x) is the cost objective function; T l is the power shortage duration; T is the total duration of the system in a given time period; E out is the effective electric energy output by the system in a certain time period; E w is the total wind energy input; and E s is the total solar energy input.

[0038] Wherein, the cost target function C(x) = C inv +C om +C f , C inv , C om , C f is the equipment investment cost, operation and maintenance cost, fuel cost.

[0039] Further, the wind-solar-hydrogen storage micro-grid system capacity configuration target includes utilization rate, cost, power load, energy loss rate;

[0040] Wherein, the user end self-defined period or master iterates the wind-solar-hydrogen storage micro-grid system capacity configuration target, and when the wind-solar-hydrogen storage micro-grid system capacity configuration target is changed, the refreshing step is executed, and the learning model is synchronously iterated and reconstructed.

[0041] Compared with the known prior art, the technical scheme provided by the application has the following beneficial effects:

[0042] The application provides a wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO, which can perform more reasonable capacity configuration on the wind-solar-hydrogen storage micro-grid system during execution, reliably measure the power generation of the device, the hydrogen storage system parameters, the load demand and the like according to specific conditions, establish an improved multi-objective optimization particle swarm algorithm model, take the overall benefits of all parties, the economy and reliability of the micro-grid system operation as optimization targets, and reliably perform capacity optimization configuration on the micro-grid system. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical scheme in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0044] Figure 1 It is a flowchart of the wind-solar-hydrogen storage micro-grid system capacity optimization configuration method based on MPSO;

[0045] Figure 2 It is a structure diagram of the wind-solar-hydrogen storage micro-grid system in the application;

[0046] Figure 3 It is a schematic diagram of the overall operation principle of the wind-solar-hydrogen storage micro-grid system in the application;

[0047] Figure 4 It is a schematic diagram of the overall flow of the multi-objective optimization particle swarm algorithm in the application. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0049] The present invention will be further described below with reference to the embodiments.

[0050] Example:

[0051] The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method of this embodiment is as follows: Figure 1 Shown, including:

[0052] Setting a parameter collection cycle, collecting power generation parameters in the wind power generation device and the photovoltaic power generation device based on the collection cycle, and synchronously collecting the tie line power parameters based on the collection cycle;

[0053] The setting logic of the parameter acquisition cycle is expressed as:

[0054]

[0055] Where: T collect is the parameter acquisition period; ω, γ is the weight; F is the power generation equipment fluctuation factor; R is the power generation equipment response factor; S is the factor based on storage limitation;

[0056] Among them, the weight ω, γ are all positive numbers, and the weights ω, The sum of γ is 1, and the weight value is customized by the user. The weights are initially set to 0.5, 0.3, and 0.2;

[0057] The calculation formulas for F, R, and S are:

[0058]

[0059] Where: α, β are weights; σ w , σ s is the standard deviation of wind power generation fluctuation and photovoltaic power generation fluctuation; is the average power of wind power generation in the historical period, and the average power of photovoltaic power generation in the historical period; T w 、T s is the cumulative response time of wind power generation equipment and photovoltaic power generation equipment; M maxis the maximum data storage amount per unit time acceptable; m is the average storage byte number of a single data point; T tolal is the corresponding time length of the union of the historical time periods of wind power generation and photovoltaic power generation;

[0060] wherein the values of the weights α and β are defined by the user of the user terminal, the weights α and β are both in the range of [0, 1], the historical time periods of wind power generation and photovoltaic power generation are defined by the user of the user terminal, M max , and m are derived from the storage device used after the collection of the power generation parameters;

[0061] The collected power generation parameters and tie-line power parameters are subjected to filtering processing;

[0062] The power generation parameters and tie-line power parameters subjected to filtering processing are obtained, and a learning model is built by combining the two through the MPSO algorithm;

[0063] A wind-solar-hydrogen storage micro-grid system capacity configuration target is set, and the parameters of the configuration target are imported into the learning model;

[0064] The learning model is started, and the iterative search process of the particle swarm is repeated to continuously find an optimal capacity configuration scheme that meets the target;

[0065] When the power generation parameters and tie-line power parameters subjected to filtering processing are processed by the MPSO algorithm, optimization is performed through the introduction of a dynamic inertia weight or the addition of Gaussian disturbance, and after the optimization of the MPSO algorithm is completed, the building of the corresponding learning model is performed;

[0066] wherein the optimization of the MPSO algorithm is preferentially performed through the introduction of a dynamic inertia weight, and the optimization steps include:

[0067] A dynamic inertia weight function is set:

[0068] ω′=ω max -(ω max -ω min )*(iter / iter max );

[0069] In the formula: ω′ is the current inertia weight; ω max is the maximum value of the set inertia weight; ω min is the minimum value of the set inertia weight; iter is the current iteration number; iter max is the total maximum iteration number;

[0070] The dynamic inertia weight obtained based on the dynamic inertia weight function is further embedded in the part of particle speed update of the MPSO algorithm, that is, introduced into the particle speed update formula:

[0071]

[0072] wherein: v[i][j] is the velocity of the ith particle in the jth dimensional space; c1, c2 are learning factors; rand() is a random number generating function; pbest[i][j], gbest[i][j] are the jth dimensional component of the individual optimal position of the ith particle and the jth dimensional component of the group optimal position; x[i][j] is the jth dimensional component of the current position of the ith particle;

[0073] wherein: rand(), ω max , ω max are defined by the user terminal;

[0074] The steps for building the learning model include:

[0075] In the learning model, the position vector of the particle is set as the capacity size of the components of the wind-solar-hydrogen storage micro-grid system, and the velocity vector of the corresponding particle is used to control the moving state of the particle in the search space. According to the target to be achieved by the system capacity configuration, a fitness function is constructed to determine the good or bad degree of the capacity configuration scheme represented by different particles.

[0076] The value range of the penalty factor and the learning factor is set;

[0077] The preprocessed data is divided into a training set and a validation set according to a suitable proportion. The validation set is applied to verify the model, and the fitting degree between the capacity configuration result output by the model and the actual expected result is viewed. According to the verification result, the penalty factor and the learning factor of the model are coordinated, and finally the learning model is output.

[0078] The components of the wind-solar-hydrogen storage micro-grid system include wind turbines, photovoltaic panels, and hydrogen storage equipment. The target to be achieved by the system capacity configuration includes cost control and power supply reliability guarantee. The cost control in the target to be achieved by the system capacity configuration includes power value and cost. The power supply reliability guarantee in the target to be achieved by the system capacity configuration includes the fault frequency of the wind-solar-hydrogen storage micro-grid system and the average energy storage of the wind-solar-hydrogen storage micro-grid system.

[0079] The fitness function is:

[0080]

[0081] wherein: F(x) is the fitness value; C(x) is the cost objective function; T l is the power shortage duration; T is the total duration of the system within a given time period; E out is the effective electric energy output by the system within a certain time period; E w is the total wind energy input; E s is the total solar energy input;

[0082] Wherein, the cost objective function C(x) = C inv +C om +C f , C inv , C om , C f is the equipment investment cost, operation and maintenance cost, fuel cost;

[0083] The capacity configuration target of the wind-solar-hydrogen storage micro-grid system includes utilization rate, cost, power load, and energy loss rate.

[0084] Wherein, the user end customizes a period or the master controls the iteration of the capacity configuration target of the wind-solar-hydrogen storage micro-grid system, and when the capacity configuration target of the wind-solar-hydrogen storage micro-grid system changes, the refreshing step is executed, and the learning model is iteratively reconstructed.

[0085] In this embodiment, by establishing an improved multi-objective optimization particle swarm algorithm model, the overall benefits of all parties, the economy and reliability of the micro-grid system are taken as optimization targets, and the reliable capacity optimization configuration of the micro-grid system is performed.

[0086] Referring to Figure 3 , the micro-grid is provided with power by new energy power generation devices such as wind power generation and photovoltaic power generation. When the wind power generation capacity and the photovoltaic power generation capacity are greater than the load demand, the excess power is stored in the form of hydrogen through an electrolytic tank and the like, or the excess power can be sold to the power grid to improve the economic benefit of the micro-grid system. When the wind power generation capacity and the photovoltaic power generation capacity are less than the load demand, the hydrogen energy stored in the foregoing is used to generate power by a fuel cell to make up for the load difference, maintain the reliability of the micro-grid system operation, and avoid the generation of load interruption. If the load demand still cannot be met, power can be purchased from the power grid through the tie line.

[0087] As shown in Figure 1 , the power generation parameters include wind power generation parameters and photovoltaic power generation parameters. The wind power generation parameters include wind speed, wind direction, wind wheel speed, power output, air density, and turbulence intensity. The photovoltaic power generation parameters include light intensity, temperature, open-circuit voltage of a photovoltaic cell, short-circuit current, maximum power point power, and filling factor. The tie line power parameter includes active power, power direction, power fluctuation rate, and power factor.

[0088] Wherein, the parameter collection period is set based on historical power generation parameters.

[0089] Through the above setting, the contents of the power generation parameters and the tie line power parameters are further limited.

[0090] On the other hand, the mathematical modeling of the wind-solar-hydrogen storage micro-grid system is as follows:

[0091] The economic benefits should consider the situation of three investment parties of wind power, photovoltaic and hydrogen storage. The three investment parties form a non-cooperative game relationship in the whole capacity configuration process, which has both competition and cooperation. The game party set can be expressed as formula (3-1) :

[0092] N = {WT, PV, H} (3-1) ;

[0093] The investment party profit function is shown in formula (3-2) :

[0094] G all = G sell -G buy -G punish -G run (3-2)

[0095] In the above formula, G all represents the total income of the investment party, G sell represents the electricity sales income of the investment party, G buy represents the electricity purchase cost of the investment party, G punish represents the penalty factor introduced by the investment party due to wind or light abandonment or load interruption, and G run represents the operation and maintenance cost of the investment party.

[0096] The electricity sales income is shown in formula (3-3) :

[0097]

[0098] Where A represents the electricity sales price at that time, P sell represents the sold power, and T represents the system operation cycle.

[0099] The electricity purchase cost is shown in formula (3-4), P buy represents the power purchased from the grid, and B represents the electricity purchase price at that time.

[0100]

[0101] The penalty factor is shown in formula (3-5) :

[0102] G punish = C × W1+ D × W2 (3-5)

[0103] Where C is the penalty factor corresponding to the wind or light abandonment behavior, D is the penalty factor corresponding to the load interruption behavior, W1 is the wasted power of the wind or light abandonment behavior, and W2 is the power not supplied by the load interruption behavior.

[0104] The investment party operation and maintenance cost is shown in formula (3-6) :

[0105] G run = E × G inv (3-6)

[0106] Where: E is the operation and maintenance cost coefficient, G inv is the total investment cost.

[0107] In combination with actual usage, it is necessary to restrict the number of devices installed. The number of devices is limited as shown in formula (3-7):

[0108] N min ≤N≤N max (3-7)

[0109] N min 、N max The lower and upper limits for the number of devices.

[0110] The following further explains the process of filtering through the Kalman filter algorithm after initializing the parameters:

[0111] Prediction stage:

[0112] State prediction: Based on the state equation of the system, the state estimate value at the previous moment k is used To predict the current state value For linear systems, the state equation is usually expressed as Among them F k It is the state transition matrix, which describes the transition law of the system state from time k-1 to time k; is the optimal state estimate after filtering at the previous moment; B k is the control input matrix, u k It is the control input at the current moment (if any). For example, in wind power generation parameter filtering, the state variables may include the changes in wind speed, wind direction and other related physical quantities over time, and their dynamic evolution laws are reflected through the state transfer matrix for prediction.

[0113] Error covariance prediction:

[0114] At the same time, the error covariance matrix P of the state estimation kk-1 It is also predicted, and its calculation formula is Here P k-1 is the error covariance matrix of the previous moment, which reflects the uncertainty of the state estimation at the previous moment; It's F k The transpose matrix, Q k It is the process noise covariance matrix, which is used to describe the uncertainty caused by random interference factors in the system during the state transition process. For example, the random fluctuations of natural factors such as wind speed and light in the actual environment have an impact on the power generation parameters. This formula can be used to predict the error covariance of the state prediction at the current moment.

[0115] Update stage:

[0116] Calculate Kalman gain: first, the Kalman gain K is calculated k , and the calculation formula is , wherein H k is the observation matrix, which establishes the relationship between the state variables and the observation variables, such as the corresponding relationship between the observation values such as the measured power generation power and the actual wind speed, illumination and other state variables through the sensor; R k is the measurement noise covariance matrix, which reflects the uncertainty caused by sensor measurement error and other factors in the observation process. The Kalman gain is determined by comprehensively considering the prediction error covariance, observation relationship and measurement error and other factors through the above formula, and the gain is used to weigh the weights of the predicted value and the observation value.

[0117] State update: the predicted state value k is updated using the Kalman gain K , to obtain the optimal state estimation value after filtering at the current time , and the calculation formula is , wherein z k is the actual observation value at the current time, that is, the actual measurement data such as power generation power, tie line power and the like collected through the sensor. The formula indicates that the difference between the observation value and the observation prediction value based on the predicted state is weighted by the Kalman gain to correct the predicted state, so as to obtain a more accurate state estimation value.

[0118] Error covariance update: the error covariance matrix is updated, and the updated error covariance matrix P k is calculated as P k = (I-K k H k )P k|k-1 , wherein I is the unit matrix. The error covariance update reflects the uncertainty degree of the state estimation at the current time after the state update, and prepares for the next filtering cycle.

[0119] As shown in Figure 1 , before the power generation parameters and tie line power parameters are filtered, any one of the state transition matrix, the observation matrix, the process noise covariance matrix and the measurement noise covariance matrix is used to perform initial processing on the parameters. After the initial processing of the parameters is completed, the parameters that have completed the initial processing are further filtered based on the Kalman filtering algorithm.

[0120] Through the above settings, further execution logic support is provided for the execution of the method in the embodiment, ensuring the effectiveness and reliability of the execution of the method steps in the embodiment.

[0121] In summary, the method in the above embodiments can reduce the load supply interruption rate and the wind and light abandonment rate as much as possible under the condition of ensuring the highest overall income of each part (wind power generation, photovoltaic power generation, hydrogen storage), can perform more reasonable capacity configuration on the wind, light, hydrogen storage and micro-grid system, can perform reliable measurement on the power generation of the device, the hydrogen storage system parameters, the load demand and other conditions according to the specific conditions, can establish an improved multi-objective optimization particle swarm algorithm model, and can take the overall income of each party, the economy and reliability of the micro-grid system operation as the optimization target to perform reliable capacity optimization configuration on the micro-grid system.

[0122] The above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method is characterized by: include: Setting a parameter collection cycle, collecting power generation parameters in the wind power generation device and the photovoltaic power generation device based on the collection cycle, and synchronously collecting the tie line power parameters based on the collection cycle; Filtering the collected power generation parameters and tie line power parameters; Obtain the filtered power generation parameters and tie-line power parameters, and combine them to build a learning model using the MPSO algorithm; Set the capacity configuration target of the wind-solar-hydrogen-storage microgrid system and import the configuration target parameters into the learning model; Start the learning model and repeat the iterative search process of the particle swarm to continuously find the optimal capacity configuration solution that meets the goal; The power generation parameters include wind power generation parameters and photovoltaic power generation parameters. Wind power generation parameters include wind speed, wind direction, wind rotor speed, power output, air density and turbulence intensity. Photovoltaic power generation parameters include light intensity, temperature, open circuit voltage, short circuit current, maximum power point power and fill factor of photovoltaic cells. The tie line power parameters include active power, power direction, power fluctuation rate and power factor. Among them, the parameter collection cycle is set based on historical power generation parameters; The setting logic of the parameter acquisition cycle is expressed as: ； Where: T collect is the parameter acquisition period; ω, , γ is the weight; F is the fluctuation factor of power generation equipment; R is the response factor of power generation equipment; S is the factor based on storage limitation; Among them, the weight ω, ,γ are all positive numbers, and the weights ω, The sum of ,γ is 1, and the weight value is customized by the user end. The weight is initially set to 0.5, 0.3, and 0.2; The calculation formulas for F, R, and S are: ； Where: α, β are weights; σ w , σ s is the standard deviation of wind power generation fluctuation and photovoltaic power generation fluctuation; is the average power of wind power generation in the historical period, and the average power of photovoltaic power generation in the historical period; T w 、T s is the cumulative response time of wind power generation equipment and photovoltaic power generation equipment; M max is the maximum acceptable data storage capacity per unit time; m is the average storage bytes of a single data point; T tolal is the corresponding duration of the union of the historical time periods of wind power generation and photovoltaic power generation; Among them, the weights α and β are user-defined by the user end, and the weights α and β are both in the range of [0, 1]. The wind power generation in the historical time period and the photovoltaic power generation in the historical time period are user-defined by the user end. max ,m comes from the storage device used after the power generation parameters are collected; When the filtered power generation parameters and tie-line power parameters are processed by the MPSO algorithm, they are optimized by introducing dynamic inertia weights or adding Gaussian disturbances. After the MPSO algorithm completes the optimization, the corresponding learning model is constructed. Among them, the optimization of the MPSO algorithm is preferably performed by introducing dynamic inertia weights. The optimization steps include: Set the dynamic inertia weight function: ω′=ω max -(oh max -oh min )*(iter / iter max ); Where: ω′ is the current inertia weight; ω max is the maximum value of the set inertia weight; ω min is the minimum value of the set inertia weight; iter is the current number of iterations; iter max is the total maximum number of iterations; The dynamic inertia weight obtained based on the dynamic inertia weight function is further embedded into the particle velocity update part of the MPSO algorithm, that is, introduced into the particle velocity update formula: ； Where: v[i][j] is the velocity of the i-th particle in the j-dimensional space; c1 and c2 are learning factors; rand() is the random number generator function; pbest[i][j] and gbest[i][j] are the components of the individual optimal position of the i-th particle in the j-dimensional space and the group optimal position in the j-dimensional space; x[i][j] is the component of the current position of the i-th particle in the j-dimensional space. Among them, rand( ), ω max 、ω min Customized by the user.

2. The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method according to claim 1 is characterized in that: Before filtering the power generation parameters and the tie line power parameters, the parameters are initially processed using any one of the matrices: the state transfer matrix, the observation matrix, the process noise covariance matrix, and the measurement noise covariance matrix. After the parameters are initialized, the parameters that have completed the initialization processing are further filtered based on the Kalman filter algorithm.

3. The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method according to claim 1 is characterized in that: The steps of building the learning model include: In the learning model, the particle position vector is set to represent the capacity of the components of the wind-solar-hydrogen-storage microgrid system. The corresponding particle velocity vector is used to control the movement of the particle in the search space. Based on the desired goal of system capacity configuration, a fitness function is constructed to determine the quality of the capacity configuration scheme represented by different particles. Set the value range of penalty factor and learning factor; Divide the preprocessed data into a training set and a validation set in appropriate proportions. Use the validation set to validate the model and check the fit between the capacity configuration output by the model and the actual expected results. Based on the validation results, adjust the penalty factor and learning factor of the model and finally output the learning model. Among them, the components of the wind, solar, hydrogen and storage microgrid system include wind turbines, photovoltaic panels, and hydrogen storage equipment. The goals that the system capacity configuration expects to achieve include cost control and power supply reliability. The cost control among the goals that the system capacity configuration expects to achieve includes electricity value and fees. The power supply reliability among the goals that the system capacity configuration expects to achieve includes: the failure frequency of the wind, solar, hydrogen and storage microgrid system and the average energy storage capacity of the wind, solar, hydrogen and storage microgrid system.

4. The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method according to claim 3 is characterized in that: The fitness function is: ； Where: F(x) is the fitness value; C(x) is the cost objective function; T l is the duration of power outage; T is the total duration of the system in a given time period; E out The effective electric energy output by the system in a certain period of time; E w is the total wind energy input; E s is the total solar energy input; Among them, the cost objective function C(x)=C inv +C om +C f , C inv 、C om 、C f Equipment investment cost, operation and maintenance cost, and fuel cost.

5. The MPSO-based wind-solar-hydrogen-storage microgrid system capacity optimization configuration method according to claim 1 is characterized in that: The capacity configuration targets of the wind-solar-hydrogen-storage microgrid system include: utilization rate, cost, power load, and energy loss rate; Among them, the user-side customized cycle or the master control iterates the capacity configuration target of the wind, solar, hydrogen and storage microgrid system. When the capacity configuration target of the wind, solar, hydrogen and storage microgrid system changes, the refresh step is executed and the learning model is synchronously iterated and rebuilt.

Images