Scheduling and active supporting method for village-level micro-grid containing supercapacitor

Abstract

The invention provides a scheduling and active supporting method for a village-level micro-grid containing a super capacitor, and belongs to the technical field of power grid scheduling supporting, and the method comprises the steps: building a scheduling model of a hybrid energy storage system by taking the minimum comprehensive cost of the village-level micro-grid, the minimum loss rate of a lithium battery and the minimum virtual inertia response speed of the super capacitor as objective functions; constructing constraint conditions of the scheduling model under the condition of considering photovoltaic, wind power and gas turbine output power constraints, lithium battery charging and discharging constraints, village-level micro-grid power balance constraints and power interaction constraints of the village-level micro-grid and an external power grid; and improving a South goose optimization algorithm, and solving the scheduling model by using the improved South goose optimization algorithm. The hybrid battery container has the advantages that the supercapacitor and the lithium battery are combined to form the hybrid battery container, excessive dependence of a village-level micro-grid on a traditional lithium battery is effectively avoided, and the problem that the service life of equipment is shortened due to frequent charging and discharging of the traditional lithium battery is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grid dispatching support, and in particular relates to a village-level microgrid dispatching and active support method containing a supercapacitor. Background Art

[0002] With the continuous rise of distributed power technology, the concept of microgrids composed of distributed power access has been proposed. Microgrids can connect sources, loads and storage to form a new type of power system, and have the characteristics of plug-and-play, high energy utilization, and energy self-consistency, which has attracted much attention. Photovoltaic and other new energy power generation has uncertainty and volatility. Compared with traditional integrated energy systems, the flexibility and stability of microgrids with a large amount of clean energy access are relatively poor. As a key platform connecting distributed clean energy and microgrid end users, energy storage systems can effectively alleviate problems such as supply and demand imbalance and system stability in village-level microgrids. At present, traditional energy storage systems often use lithium batteries for peak and frequency regulation, but frequent charging and discharging accelerates their life decay. In order to solve the above problems, hybrid energy storage composed of electrochemical energy storage and other forms of energy storage has emerged. Since new energy has both weak damping and inertia, the active support of village-level microgrids has decreased, while supercapacitors can achieve high power density, fast response, strong inertia and long cycle life. Therefore, they are often combined with lithium batteries to form hybrid energy storage for related applications and research to achieve active support and peak and frequency regulation of microgrids.

[0003] Currently, most microgrid energy storage systems rely on energy-type lithium batteries to balance supply and demand. Frequent charging and discharging can easily reduce the service life of the energy storage system. Summary of the invention

[0004] The technical problem to be solved by the present invention is to optimize the capacity configuration, energy scheduling and active support of a hybrid capacitor composed of a supercapacitor and a lithium battery in a village-level microgrid, and to extend the life of a lithium battery by a supercapacitor. In view of the deficiencies of the prior art, a village-level microgrid scheduling and active support method containing a supercapacitor that can support a chemical battery of a village-level microgrid is provided. To solve the above technical problems, the technical solution adopted by the present invention is to construct a scheduling model of a hybrid energy storage system with the minimum comprehensive cost of a village-level microgrid, the minimum loss rate of a lithium battery and the minimum virtual inertia response speed of a supercapacitor as the objective function;

[0005] Construct the constraints of the dispatch model by considering the output power constraints of photovoltaic, wind and gas turbines, the charging and discharging constraints of lithium batteries, the power balance constraints of village microgrids, and the power interaction constraints between village microgrids and external power grids;

[0006] An improved Snow Goose optimization algorithm is used to solve the scheduling model.

[0007] Furthermore, the objective function is in is the comprehensive cost of village microgrid, is the lithium battery loss rate, is the sum of the virtual inertia time constants of the supercapacitors during the scheduling period.

[0008] Furthermore, the comprehensive cost of village-level microgrid is:

[0009]

[0010] Among them, F penalty Penalty cost for imbalance of village microgrid, is the interaction cost between the village microgrid and the traditional large power grid, F maint Equipment operation and maintenance costs, The cost of treating pollutants emitted by gas turbines is The configuration cost of hybrid energy storage is is the gas turbine operating cost, ω1~ω3 are the weights obtained by the AHP method;

[0011] Among them, F penalty , F maint , as well as The calculation is as follows:

[0012]

[0013] Among them, P buy,t It represents the power shortage value when the internal energy of the village microgrid cannot support the load power demand, P sell,t It means that the internal energy of the village microgrid exceeds the electric energy value that can be absorbed by the village microgrid. and are the electricity purchase price and electricity selling price of the village microgrid to the external power grid at time t; gas is the unit price of natural gas; GT , β Bat , β plov , β SC , β WT are the maintenance cost coefficients of gas turbines, lithium batteries, photovoltaics, supercapacitors and wind turbines respectively; P GT,t is the gas turbine output power at time t, and They are respectively the charging and discharging power of the lithium battery at time t, is the output power of the ith photovoltaic power generation device at time t, P WT,t is the output power of the wind power generation device at time t, and are respectively the charging and discharging power of the supercapacitor in the hybrid energy storage, e poll , α poll , are the unit price of pollutant treatment, nitrogen oxide emission coefficient and the purchase price of carbon emission rights; σ GT is the carbon emission factor of the gas turbine; k penalty ,P van,t ,P waste,t They are the power imbalance penalty coefficient, the shortfall power and the power that cannot be absorbed;

[0014] Hybrid energy storage configuration costs for:

[0015]

[0016] In the above formula, The installation (configuration) cost of HESS; They are the capacity of lithium battery and the unit price of supercapacitor; They are the capacity of lithium battery and supercapacitor respectively; is the upper limit of the capacity of the supercapacitor during the capacity configuration process, is the annual value conversion factor; r is the power supply discount rate; T Y The life of the energy storage device.

[0017] Furthermore, the loss rate of lithium battery is:

[0018]

[0019] C bat (DOD)=28270·exp(-2.401DOD)+2.214·exp(5.901DOD);

[0020] in, is the loss rate of electrochemical energy storage in the entire battery dispatch, and DoD t >0.01 indicates all the moments when the lithium battery is charged and discharged; C bat (1) is the number of available battery cycles at the maximum discharge depth; C bat (DoD t ) is the number of available cycles of the battery under DODt after the battery is charged and discharged at time t; C bat It is the number of available cycles of lithium battery, and DOD is the depth of charge and discharge.

[0021] Furthermore, it is defined that the use of supercapacitors takes precedence over electrochemical energy storage;

[0022] The supercapacitor provides active support for the voltage and frequency fluctuations of the village microgrid. The sum of the virtual inertia time constants of the supercapacitors during the dispatch period is calculated as follows:

[0023]

[0024] In the formula, H B,t is the virtual inertia time constant of the supercapacitor at time t (unit: second); S SC is the capacity of the supercapacitor, W SC The energy stored in the supercapacitor energy storage device;

[0025] Among them, the calculation of the virtual inertia time constant of the supercapacitor includes:

[0026]

[0027] The rotational kinetic energy EG stored in the synchronous machine is described as:

[0028]

[0029] Among them, Δγ SOC With γ SOC W are the charge state change of the energy storage device at two moments and the charge state of the energy storage device at the current moment respectively; SC The energy stored in the supercapacitor energy storage device; E G The rotational kinetic energy stored in the rotor of a synchronous machine with the same energy storage capacity; Δω e With ω e are the electrical angular velocity changes of the synchronous generator at two moments and the electrical angular velocity of the synchronous generator at the current moment; p n is the number of pole pairs of the synchronous generator; J G is the inertia of a synchronous machine of equal capacity;

[0030] The virtual inertia time constant of the supercapacitor at time t is:

[0031]

[0032] Furthermore, the output power constraints of photovoltaic, wind and gas turbine are and are the upper limits of the power output of photovoltaic, wind and gas turbine respectively. and P GT,t are the output power of photovoltaic, wind and gas turbine at time t, P GT,t-1 is the output power of the gas turbine at time t-1, is the power output ramp rate of the gas turbine.

[0033] Furthermore, the lithium battery charge and discharge constraints are

[0034] In the formula, and They are the maximum charge and discharge power of lithium batteries respectively; and They are the charge and discharge signs of the lithium battery at time t. It indicates that the lithium battery is in charging state. When the battery is in the discharge state, it is used to prevent charging and discharging from being performed simultaneously. and They are respectively the lower and upper limit coefficients of the state of charge during the operation of the lithium battery.

[0035] Furthermore, the power balance constraint of the village-level microgrid is

[0036]

[0037] Among them, P Load,t is the power load of the village microgrid at time t; P is the power of the microgrid interacting with the external grid at time t, with positive values ​​representing purchases and negative values ​​representing sales; phvo,t , P WT,t and P GT,t are the output power of photovoltaic, wind and gas turbine at time t respectively; are the discharge power of lithium battery and supercapacitor respectively; They are the charging power of lithium battery and supercapacitor respectively.

[0038] Further, the method of improving the snow goose optimization algorithm includes:

[0039] The snow goose optimization algorithm is improved by enhancing the adaptive behavior of the population and introducing an external warehouse mechanism to generate a snow goose multi-objective optimization algorithm.

[0040] Parameter initialization: Initialize the population for the snow goose optimization algorithm, calculate the corresponding objective function value based on the initialization position, select the best individual and record the optimal solution;

[0041] Optimization stage selection: The snow goose formation is divided into two types: herringbone and straight. When the snow geese deflect too much in the herringbone formation and are ready to enter the development stage, the restoration of the normal formation is controlled by the parameter θ with greater randomness;

[0042] The asymptotic development and exploration control factor cycle is introduced to replace the original matrix angle parameter θ. Under the action of cycle, the probability of entering the exploration stage will decrease with the increase of the number of iterations T during the entire iteration process, while the development stage is the opposite, and it is repeated C times to balance the global search and local development intensity. The stage selection method is as follows:

[0043]

[0044] Where, T max is the maximum number of iterations; C is the number of loops; rand 2 is a random number between [0,1];

[0045] After stage selection, the specific modeling methods of the two queue improvement models are as follows:

[0046] The herringbone square model in the exploration phase:

[0047] In the algorithm exploration phase, snow geese are divided into leading geese, weak geese, and other groups with different health conditions:

[0048]

[0049] In the formula, They represent the position of the i-th snow goose and the position of the optimal snow goose at the t-th iteration respectively; They represent the center position of the population at the tth iteration and the position of the snow goose with the lowest ranking after the population is sorted; B, D, and E represent the weight coefficients of the corresponding components; represents the speed of the i-th snow goose at the t-th iteration;

[0050] In the exploration phase, the migration behavior of weak snow geese was simulated to make the snow geese fly at different speeds to the center of the population, thus enhancing the overall flight efficiency of the population.

[0051] Let F pro To follow the protection factor of the superior population, the snow geese in the bottom 1 / 5 of the simulated population are as follows:

[0052]

[0053] The linear array in the development stage:

[0054] When the angle of the snow goose herringbone array exceeds the cycle factor selected in the optimization stage, it enters the second stage of flight, that is, the algorithm development stage. In this stage, the idea of ​​high-quality population radius is introduced to guide the population, and the improved dynamic simulation snow goose flight speed is as follows:

[0055]

[0056] In the formula, rand 1 is a random number between [0,1], used to indicate whether the snow goose is in the local optimal solution; Brownian(dim) represents the standard Brownian motion, dim is the variable dimension; represents entry-wise multiplication, is the radius of the high-quality population; N is the physical decline coefficient caused by long-term flight of snow geese;

[0057] In the linear array, the physical decline coefficient N decreases with the increase of the number of algorithm iterations, simulating the decline in physical strength of snow geese in the later stage of migration, which causes the speed to decrease. At the same time, the radius of high-quality population is introduced. Enhance the ability of snow geese to follow high-quality flocks;

[0058] Non-dominated solution storage and maintenance: The external warehouse mechanism is introduced to improve the Snow Goose optimization algorithm. The proposed external warehouse mechanism mainly includes the non-dominated solution selection, non-dominated solution storage, and external warehouse update and maintenance stages;

[0059] Assume that when solution S1 dominates solution S2, let the corresponding objective function value of each solution be recorded as f, then we have:

[0060]

[0061] make represents the normalized goodness of the ith solution under the mth objective function, then the usage of the ith non-dominated solution can be calculated, thereby eliminating non-dominated solutions with high congestion; the calculation method is:

[0062]

[0063] Optimal solution update: The population position obtained by the t-th update is brought into the objective function to calculate the objective function value, and the local optimal solution obtained by the t-th iteration is compared with the global optimal solution obtained by the k-1 iteration, and the global optimal solution is updated;

[0064] Iterative update: Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, repeat the iterative update.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] (1) The combination of supercapacitors and lithium batteries to form a hybrid battery container effectively avoids the excessive dependence of village-level microgrids on traditional lithium batteries and solves the problem of reduced equipment life caused by frequent charging and discharging of traditional lithium batteries.

[0067] (2) It solves the problems of damping and low inertia of clean energy and improves the active support capacity of village-level microgrids.

[0068] (3) It supplements the in-depth research on the capacity configuration, energy scheduling and active support capabilities of hybrid energy storage combining supercapacitors and lithium batteries in small microgrids, as well as the impact of supercapacitors on the service life of lithium batteries.

[0069] (4) It can realize the optimal configuration of economic and active support for hybrid energy storage in village-level microgrids, and provide a multi-objective optimization scheduling model to achieve economic operation of village-level microgrids, internal energy autonomy, and improve the local energy consumption rate.

[0070] (5) Improving the existing single-objective snow goose optimization algorithm can effectively increase the iteration speed and global optimization accuracy, avoid falling into the local optimal solution, and improve the population fitness under complex models. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0072] Figure 1 : A schematic diagram of the topological structure of a village-level microgrid system in Embodiment 1 of the present invention;

[0073] Figure 2 : Flow chart of the improved snow goose multi-objective optimization algorithm in Example 2 of the present invention;

[0074] Figure 3 : Process diagram of improving the snow goose multi-objective optimization algorithm in embodiment 2 of the present invention;

[0075] Figure 4 : A schematic diagram of the structure of solving the VMC scheduling model in Embodiment 2 of the present invention;

[0076] Figure 5 : Schematic diagram of the test results of SGAR in Example 2 of the present invention;

[0077] Figure 6 : Radar chart of IGD and HV indicators in Example 2 of the present invention;

[0078] Figure 7 : SP indicator box plot and Pareto front diagram in Example 1 of the present invention. DETAILED DESCRIPTION

[0079] In order to better understand the present invention, the content of the present invention is further clearly described below in conjunction with the embodiments and the accompanying drawings, but the protection content of the present invention is not limited to the following embodiments. In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details.

[0080] The village-level microgrid system is mainly composed of photovoltaic, wind power, gas turbine and other power generation units, power loads, dispatching centers, and hybrid energy storage systems. The specific system topology is as follows: Figure 1 shown.

[0081] Embodiment 1: A method for dispatching and actively supporting a village-level microgrid containing a supercapacitor in this embodiment includes:

[0082] A hybrid energy storage system consisting of supercapacitors and lithium batteries is used for energy storage. The mathematical model of lithium batteries is as follows:

[0083]

[0084] Where ES t ,ES t-1 are the amount of electricity stored in the lithium battery at time t and t-1 respectively; ε Bat,e is the self-loss coefficient of the lithium battery; and are the charging power and discharging power of the lithium battery at time t respectively; are the charging efficiency and discharging efficiency of lithium batteries respectively; Δt is the scheduling period, which is 1h.

[0085] The mathematical model of supercapacitor energy storage is as follows:

[0086]

[0087] in, and are the charging power and discharging power of the supercapacitor at time t, respectively, and They are the charging and discharging efficiency of the supercapacitor respectively. SC is the self-loss coefficient of the supercapacitor.

[0088] The calculation method of supercapacitor charge state in hybrid energy storage system is as follows:

[0089]

[0090] In formula (3) SC soc (t) are the initial state of charge of the supercapacitor and the state of charge at time t, respectively; S SC is the capacity of the supercapacitor; V SC With V SC,max are the voltage value of the supercapacitor at time t and the maximum voltage value it can withstand; i SC is the current flowing through the supercapacitor at time t.

[0091] The scheduling model of the hybrid energy storage system is constructed with the minimum comprehensive cost of the village microgrid, the lowest lithium battery loss rate and the minimum virtual inertia response speed of the supercapacitor as the objective functions;

[0092] The objective function is in is the comprehensive cost of village microgrid, is the lithium battery loss rate, is the sum of the virtual inertia time constants of the supercapacitors during the scheduling period.

[0093] Construct the constraints of the dispatch model by considering the output power constraints of photovoltaic, wind and gas turbines, the charging and discharging constraints of lithium batteries, the power balance constraints of village microgrids, and the power interaction constraints between village microgrids and external power grids;

[0094] An improved Snow Goose optimization algorithm is used to solve the scheduling model.

[0095] 1. Hybrid energy storage configuration and optimized dispatch strategy

[0096] 1. Hybrid energy storage capacity configuration

[0097] Considering the economic cost of hybrid energy storage and the inertia provided by supercapacitors for microgrids, the capacity of supercapacitors and lithium batteries in hybrid energy storage is optimized to achieve economical operation of hybrid energy storage while improving the active support capacity of village-level microgrids.

[0098] First, the capacity configuration cost model is constructed based on the installation and maintenance costs of the hybrid energy storage composed of supercapacitors and lithium batteries. The specific formula is as follows:

[0099]

[0100] In the above formula, The installation (configuration) cost of HESS; They are the capacity of lithium battery and the unit price of supercapacitor; They are the capacity of lithium battery and supercapacitor respectively; is the upper limit of the capacity of the supercapacitor during the capacity configuration process, is the annual value conversion factor; r is the power supply discount rate; T Y The life of the energy storage device.

[0101] In addition to alleviating the attenuation of battery life, supercapacitors can also provide active support (operating inertia), which can alleviate the voltage and frequency fluctuations of village-level microgrids. Reasonable configuration of supercapacitor capacity can provide active support for microgrids. The virtual inertia time constant of supercapacitor energy storage is calculated as follows:

[0102]

[0103] Among them, Δγ SOC With γ SOC W are the charge state change of the energy storage device at two moments and the charge state of the energy storage device at the current moment respectively; SC The energy stored in the supercapacitor energy storage device; E G The rotational kinetic energy stored in the rotor of a synchronous machine with the same energy storage capacity; Δω e With ω e are the electrical angular velocity changes of the synchronous generator at two moments and the electrical angular velocity of the synchronous generator at the current moment; p n is the number of pole pairs of the synchronous generator; J G is the inertia of a synchronous machine of equal capacity.

[0104] According to formulas (5)-(6), the virtual inertia time constant of the supercapacitor at time t can be obtained as:

[0105]

[0106] Therefore, the sum of the virtual inertia time constants of the supercapacitors in the scheduling period within T periods is obtained as follows:

[0107]

[0108] In the formula, H B,t is the virtual inertia time constant of the supercapacitor at time t (unit: seconds); is the sum of the virtual inertia time constants of the supercapacitors during the scheduling period.

[0109] 2. Multi-objective scheduling considering hybrid energy storage configuration

[0110] This application is based on the configuration model constructed above, and comprehensively considers the social benefits of the village microgrid and the minimization of lithium battery losses in hybrid energy storage to optimize the configuration and scheduling of the source, grid, load and storage within the village microgrid.

[0111] (1) Energy storage loss rate

[0112] Existing studies have shown that the service life of supercapacitors is almost unaffected by the depth of discharge (DOD). Therefore, in this application, the use of supercapacitors is defined as having a higher priority than electrochemical energy storage, and supercapacitors are used for energy storage and release. Through supercapacitor buffering, the number of times the electrochemical energy storage is charged and discharged is reduced, thereby extending its service life.

[0113] For lithium batteries in hybrid energy storage, assuming C bat The number of available cycles of lithium batteries. The relationship between their service life and the depth of charge and discharge DOD is as follows:

[0114] C bat (DOD)=28270·exp(-2.401DOD)+2.214·exp(5.901DOD)(10)

[0115] It can be seen from formula (10) that as the depth of discharge increases, the number of available cycles of the battery continues to decrease. Therefore, the loss rate of electrochemical energy storage in the entire battery schedule (24h), that is, the loss rate of lithium battery, is defined as follows:

[0116]

[0117] In the formula, when DoD t >0.01 indicates all the moments of charging and discharging of lithium battery; C bat (1) is the number of available battery cycles at the maximum discharge depth; C bat (DoD t ) is the time when the battery is charged and discharged at time t, at DOD t The following battery life cycles.

[0118] (2) Social benefits of village microgrids

[0119] This application considers the interaction cost benefits between the village-level microgrid and the large power grid, the pollution treatment costs generated by the gas turbine, the equipment operation and maintenance costs, and the penalty costs for the imbalance of source, grid, load and storage as the social benefits of the village-level microgrid. The total social benefits of the village-level microgrid, that is, the comprehensive cost of the village-level microgrid, can be described as:

[0120]

[0121] Among them, F penalty Penalty cost for imbalance of village microgrid, is the interaction cost between the village microgrid and the traditional large power grid, F maint Equipment operation and maintenance costs, The cost of treating pollutants emitted by gas turbines is The hybrid energy storage configuration cost is formed by formula (4), is the gas turbine operating cost, ω 1 ~ω 3 are the weights obtained by the hierarchical analysis method.

[0122] In formula (12), F penalty , F maint , as well as The specific calculation method is as follows:

[0123]

[0124] Among them, P buy,t It represents the power shortage value when the internal energy of the village microgrid cannot support the load power demand, P sell,t It means that the internal energy of the village microgrid exceeds the electric energy value that can be absorbed by the village microgrid. and are the electricity purchase price and electricity selling price of the village microgrid to the external power grid at time t; gas is the unit price of natural gas; GT , β Bat , β plov , β SC , β WT are the maintenance cost coefficients of gas turbines, lithium batteries, photovoltaics, supercapacitors and wind turbines respectively; P GT,t is the gas turbine output power at time t, and They are respectively the charging and discharging power of the lithium battery at time t, is the output power of the ith photovoltaic power generation device at time t, P WT,t is the output power of the wind power generation device at time t, and are the charging and discharging power of the hybrid energy storage supercapacitor, e poll , α poll , are the unit price of pollutant control, nitrogen oxide emission coefficient and the purchase price of carbon emission rights; σ GT is the carbon emission factor of the gas turbine; k penalty ,P van,t ,P waste,t They are the power imbalance penalty coefficient, the shortage power and the power that cannot be absorbed.

[0125] 3. Constraints

[0126] (1) Output power constraints of photovoltaic, wind and gas turbines

[0127] Let the upper limits of photovoltaic, wind and gas turbine output power be and

[0128] The following constraints apply:

[0129]

[0130] Especially for gas turbines, it is necessary to constrain the power output ramp rate, and the maximum allowable ramp rate is The following constraints apply:

[0131]

[0132] (2) Hybrid energy storage system constraints

[0133] In the hybrid energy storage system, in order to prevent the lithium battery from overcharging and discharging, it is necessary to constrain the state of charge, charging and discharging power, and charging and discharging conditions during the charging and discharging process of the lithium battery. The specific constraints are as follows:

[0134]

[0135] In the formula and They are the maximum charge and discharge power of lithium batteries respectively; and They are the charge and discharge signs of the lithium battery at time t. It indicates that the lithium battery is in charging state. When the battery is in the discharge state, it is used to prevent charging and discharging from being performed simultaneously. and They are respectively the lower and upper limit coefficients of the state of charge during the operation of the lithium battery.

[0136] (3) Power balance constraints

[0137]

[0138] The power balance constraints of the village-level microgrid are as follows:

[0139] Among them, P Load,t is the power load of the village microgrid at time t; It is the power of the microgrid interacting with the external grid at time t. Positive values ​​represent purchases and negative values ​​represent sales.

[0140] (3) Power interaction constraints

[0141] The power interaction constraints between the village microgrid and the external power grid are as follows:

[0142]

[0143] in, is the maximum allowed interaction power.

[0144] Comprehensive objective function:

[0145] 2. Improved Snow Goose Multi-Objective Optimization Algorithm

[0146] Aiming at the shortcomings of the traditional snow goose single-objective optimization algorithm, such as falling into local optimality and poor population adaptability, this algorithm is optimized and improved based on enhancing the population adaptive behavior, and an external warehouse mechanism is introduced to improve it into a multi-objective optimization algorithm.

[0147] 1. Improved multi-objective snow goose optimization algorithm

[0148] The single-objective snow goose optimization algorithm solves the problem by simulating the migration behavior of snow geese. The specific modeling process and flow are as follows.

[0149] (1) Parameter initialization

[0150] First, the population is initialized for the snow goose optimization algorithm, and the corresponding objective function value is calculated according to the initialization position. The optimal individual is selected and the optimal solution is recorded.

[0151] (2) Optimization stage selection

[0152] The snow goose optimization algorithm divides the snow goose formation into two types: herringbone and straight line. When the snow geese deflect too much in the herringbone formation and are ready to enter the development stage, the restoration of the normal formation is controlled by the parameter θ with greater randomness. This application introduces the asymptotic development and exploration control factor cycle to replace the original array angle parameter θ. Under the action of cycle, during the entire iterative process, the probability of entering the exploration stage will decrease with the increase of the number of iterations T, while the development stage is the opposite, and it is repeated C times, balancing the intensity of global search and local development. The specific stage selection method is as follows:

[0153]

[0154] Where, T max is the maximum number of iterations; C is the number of loops; rand 2 is a random number between [0,1].

[0155] After stage selection, the specific modeling methods of the two queue improvement models are as follows:

[0156] a. Herringbone square model (exploration phase)

[0157] In the algorithm exploration phase, snow geese are divided into leading geese, weak geese, and other groups with different health conditions:

[0158]

[0159] In the formula, They represent the position of the i-th snow goose and the position of the optimal snow goose at the t-th iteration respectively; They represent the center position of the population at the tth iteration and the position of the snow goose with the lowest ranking after the population is sorted; B, D, and E represent the weight coefficients of the corresponding components; represents the speed of the i-th snow goose at the t-th iteration.

[0160] In the exploration phase, the migration behavior of weak snow geese was simulated to make the snow geese fly at different speeds to the center of the population, thus enhancing the overall flight efficiency of the population.pro In order to follow the protection factor of the high-quality population, the simulated snow geese in the last 1 / 5 of the population (weak snow geese) are as follows:

[0161]

[0162] b. Linear array (development stage)

[0163] When the angle of the snow goose herringbone array exceeds the cycle factor shown in formula (20), they will enter the second stage of flight, that is, the algorithm development stage. In this stage, the idea of ​​high-quality population radius is introduced to guide the population, and the improved dynamic simulation snow goose flight speed is obtained as follows:

[0164]

[0165] In the formula, rand 1 is a random number between [0,1], used to indicate whether the snow goose is in a local optimal solution; Brownian(dim) represents the standard Brownian motion (dim is the variable dimension); represents entry-wise multiplication, is the radius of high-quality population; N is the physical decline coefficient caused by long-term flight of snow geese.

[0166] In the linear array, the physical decline coefficient N decreases with the increase of the number of algorithm iterations, simulating the decline in physical strength of snow geese in the later stage of migration, which causes the speed to decrease. At the same time, the radius of the high-quality population is introduced. Enhance the ability of snow geese to follow high-quality populations.

[0167] (3) Storage and maintenance of non-dominated solutions

[0168] This application introduces an external warehouse mechanism to improve the Snow Goose optimization algorithm. The proposed external warehouse mechanism mainly includes several stages: non-dominated solution selection, non-dominated solution storage, and external warehouse update and maintenance.

[0169] When the solution 1 Dominant Solution S 2 When , let the objective function value corresponding to each solution be recorded as f, then we have:

[0170]

[0171] make Represents the normalized goodness of the ith solution under the mth objective function, then the usage of the ith non-dominated solution can be calculated, thereby eliminating non-dominated solutions with high congestion.

[0172]

[0173] (4) Optimal solution update

[0174] The population position obtained by the tth update is substituted into the objective function to calculate the objective function value, and the local optimal solution obtained by the tth iteration is compared with the global optimal solution obtained by the k-1th iteration, and the global optimal solution is updated.

[0175] (5) Iterative Update

[0176] Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, return to step (2) for repeated iterative updates.

[0177] The improved snow goose multi-objective optimization algorithm process proposed in this application is as follows Figure 2-3 shown.

[0178] VMG optimization scheduling solution process based on MSGAR

[0179] The VMG optimization scheduling solution process based on MSGAR, the solution process of the VMG scheduling model is as follows Figure 4 shown.

[0180] Depend on Figure 4 It can be seen that the VMG scheduling model solution process consists of the following parts:

[0181] (1) Initialize equipment parameters, algorithm parameters, and load and weather data.

[0182] (2) Obtain the output of the equipment through the algorithm and calculate the values ​​of each objective function.

[0183] (3) Based on the objective function value, evaluate and update the population.

[0184] (4) Determine whether the maximum number of iterations has been reached. If so, output the optimal population position, calculate the status of each VMG device, and finally output the scheduling result; otherwise, return to step (2).

[0185] Algorithm Testing and Performance Analysis,Case 1 verifies the effectiveness of SGAR and MSGAR through a standard test set.

[0186] Representative algorithms proposed in recent years are selected as comparison algorithms, such as Parrot Optimizer (PO) and Black-winged Kite Algorithm (BKA). The parameter settings of each algorithm during the test are shown in Table 1.

[0187] Table 1 Case 1 single-objective algorithm parameter settings

[0188]

[0189] The latest CEC-2022 test set was selected to perform performance testing on the algorithm, and each test function was tested 30 times in a row. The test results of SGAR are as follows: Figure 5 shown.

[0190] Depend on Figure 5 It can be seen that, taking function F1 as an example, the variances of SGA and PO are 477.2542 and 597.5121 respectively, and the results fluctuate significantly; the variances of SGAR and BKA are 6.9118 and 12.9655 respectively, and the solution stability of SGAR is significantly better than other algorithms. For the iterative curve, the optimal solutions of F1 obtained by SGAR and BKA are 300.564 and 318.443 respectively; for F2 and F7, SGAR also obtains the optimal solution with higher quality, showing the advantages of fast iterative convergence speed, strong iterative ability and strong robustness.

[0191] The classic multi-objective metaverse optimization algorithm (MOMVO) and multi-objective particle swarm optimization algorithm (MOPSO) are selected as comparison algorithms. The parameter settings of each algorithm are shown in Table 2.

[0192] Table 2 Case 1 multi-objective algorithm parameter settings

[0193]

[0194]

[0195] The CEC-2009 test set was selected to verify the effectiveness of the multi-objective algorithm, where UF1, UF2, UF6 and UF7 are dual-objective functions, and UF8 and UF9 are triple-objective functions. Each test function was tested 10 times in a row, and IGD, SP and HV were selected as evaluation indicators. The test results of the multi-objective algorithm are shown in Figure 2. Figure 6 shown.

[0196] from Figure 6 It can be seen that for MSGAR, the best results are obtained in terms of the mean of the IGD and HV indicators. For UF1 and UF7, the IGD mean is only 0.0195 and 0.0122; for UF9, the HV mean is 0.6099. For UF9, the IGD standard deviation is 0.0745, slightly higher than the other two algorithms, but overall, MSGAR still has high stability and strong competitiveness.

[0197] The results of the SP evaluation index are plotted as a box plot, and the Pareto frontiers of UF1 and UF8 are given, such as Figure 7 shown.

[0198] from Figure 7It can be seen that the SP index results obtained by MSGAR are better than those of MOMVO and MOPSO. For UF8, the mean and standard deviation of SP obtained by MSGAR are only 0.1025 and 0.0624. Although the test function complexity of CEC-2009 is relatively high, the distribution and uniformity of the Pareto front obtained by MSGAR are the best.

[0199] Beneficial effects:

[0200] (1) It effectively avoids the excessive reliance of village-level microgrids on traditional lithium batteries and solves the problem of reduced equipment life caused by frequent charging and discharging of traditional lithium batteries.

[0201] (2) Combining supercapacitors with lithium batteries to form a hybrid battery container solves the problems of low damping and low inertia of clean energy and improves the active support capacity of village-level microgrids.

[0202] (3) It supplements the in-depth research on the capacity configuration, energy scheduling and active support capabilities of hybrid energy storage combining supercapacitors and lithium batteries in small microgrids, as well as the impact of supercapacitors on the service life of lithium batteries.

[0203] (4) It can realize the optimal configuration of economic and active support for hybrid energy storage in village-level microgrids, and provide a multi-objective optimization scheduling model to achieve economic operation of village-level microgrids, internal energy autonomy, and improve the local energy consumption rate.

[0204] (5) Improving the existing single-objective snow goose optimization algorithm can effectively increase the iteration speed and global optimization accuracy, avoid falling into the local optimal solution, and improve the population fitness under complex models.

[0205] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Other modifications or equivalent substitutions made to the technical solution of the present invention by ordinary technicians in the field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for dispatching and actively supporting a village-level microgrid containing supercapacitors, characterized in that: include: The scheduling model of the hybrid energy storage system is constructed with the minimum comprehensive cost of the village microgrid, the lowest lithium battery loss rate and the minimum virtual inertia response speed of the supercapacitor as the objective functions; Construct the constraints of the dispatch model by considering the output power constraints of photovoltaic, wind and gas turbines, the charging and discharging constraints of lithium batteries, the power balance constraints of village microgrids, and the power interaction constraints between village microgrids and external power grids; An improved Snow Goose optimization algorithm is used to solve the scheduling model.

2. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: The objective function is in is the comprehensive cost of village-level microgrid, is the lithium battery loss rate, is the sum of the virtual inertia time constants of the supercapacitors during the scheduling period.

3. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors as claimed in claim 2, characterized in that: The comprehensive cost of village-level microgrid is: Among them, F penalty Penalty cost for imbalance of village microgrid, is the interaction cost between the village microgrid and the traditional large power grid, F maint Equipment operation and maintenance costs, The cost of treating pollutants emitted by gas turbines is The configuration cost of hybrid energy storage is is the gas turbine operating cost, ω1~ω3 are the weights obtained by the AHP method; Among them, F penalty , F maint , as well as The calculation is as follows: Among them, P buy,t It represents the power shortage value when the internal energy of the village microgrid cannot support the load power demand, P sell,t It means that the internal energy of the village microgrid exceeds the electric energy value that can be absorbed by the village microgrid. and are the electricity purchase price and electricity selling price of the village microgrid to the external power grid at time t; gas is the unit price of natural gas; GT , β Bat , β plov , β SC , β WT are the maintenance cost coefficients of gas turbines, lithium batteries, photovoltaics, supercapacitors and wind turbines respectively; P GT,t is the gas turbine output power at time t, and They are respectively the charging and discharging power of the lithium battery at time t, is the output power of the ith photovoltaic power generation device at time t, P WT,t is the output power of the wind power generation device at time t, and are respectively the charging and discharging power of the supercapacitor in the hybrid energy storage, e poll , α poll , are the unit price of pollutant control, nitrogen oxide emission coefficient and the purchase price of carbon emission rights; σ GT is the carbon emission factor of the gas turbine; k penalty ,P van,t ,P waste,t They are the power imbalance penalty coefficient, the shortfall power and the power that cannot be absorbed; Hybrid energy storage configuration costs for: In the above formula, The installation (configuration) cost of HESS; They are the capacity of lithium battery and the unit price of supercapacitor; They are the capacity of lithium battery and supercapacitor respectively; is the upper limit of the capacity of the supercapacitor during the capacity configuration process, is the annual value conversion factor; r is the power supply discount rate; T Y The life of the energy storage device.

4. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors as claimed in claim 2, characterized in that: The loss rate of lithium battery is: C bat (DOD)=28270·exp(-2.401DOD)+2.214·exp(5.901DOD); in, is the loss rate of electrochemical energy storage in the entire battery dispatch, and DoD t >0.01 indicates all the moments when the lithium battery is charged and discharged; C bat (1) is the number of available battery cycles at the maximum discharge depth; C bat (DoD t ) is the number of available cycles of the battery under DODt after the battery is charged and discharged at time t; C bat It is the number of available cycles of lithium battery, and DOD is the depth of charge and discharge.

5. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors as claimed in claim 2, characterized in that: Define that the use of supercapacitors takes precedence over electrochemical energy storage; The supercapacitor provides active support for the voltage and frequency fluctuations of the village microgrid. The sum of the virtual inertia time constants of the supercapacitors during the dispatch period is calculated as follows: In the formula, H B,t is the virtual inertia time constant of the supercapacitor at time t (unit: second); S SC is the capacity of the supercapacitor, W SC The energy stored in the supercapacitor energy storage device; Among them, the calculation of the virtual inertia time constant of the supercapacitor includes: Among them, the rotational kinetic energy E stored in the rotor of the synchronous machine G Described as: Among them, Δγ SOC With γ SOC W are the charge state change of the energy storage device at two moments and the charge state of the energy storage device at the current moment respectively; SC The energy stored in the supercapacitor energy storage device; E G The rotational kinetic energy stored in the rotor of a synchronous machine with the same energy storage capacity; Δω e With ω e are the electrical angular velocity changes of the synchronous generator at two moments and the electrical angular velocity of the synchronous generator at the current moment; p n is the number of pole pairs of the synchronous generator; J G is the inertia of a synchronous machine of equal capacity; The virtual inertia time constant of the supercapacitor at time t is:

6. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: The output power constraints of photovoltaic, wind and gas turbine are and are the upper limits of the power output of photovoltaic, wind and gas turbine respectively. and P GT,t are the output power of photovoltaic, wind and gas turbine at time t, P GT,t-1 is the output power of the gas turbine at time t-1, is the power output ramp rate of the gas turbine.

7. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: Lithium battery charging and discharging constraints are In the formula, and They are the maximum charge and discharge power of lithium batteries respectively; and They are the charge and discharge signs of the lithium battery at time t. It indicates that the lithium battery is in charging state. When the battery is in the discharge state, it is used to prevent charging and discharging from being performed simultaneously. and They are respectively the lower and upper limit coefficients of the state of charge during the operation of the lithium battery.

8. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: The power balance constraint of the village microgrid is Among them, P Load,t is the power load of the village microgrid at time t; P is the power of the microgrid interacting with the external grid at time t, with positive values ​​representing purchases and negative values ​​representing sales; phvo,t , P WT,t and P GT,t are the output power of photovoltaic, wind and gas turbine at time t respectively; are the discharge power of lithium battery and supercapacitor respectively; They are the charging power of lithium battery and supercapacitor respectively.

9. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: The power interaction constraint between the village microgrid and the external grid is: in, is the maximum allowed interaction power, is the interaction power at time t.

10. The method for dispatching and actively supporting a village-level microgrid containing supercapacitors according to claim 1, characterized in that: Methods to improve the Snow Goose Optimization Algorithm include: The snow goose optimization algorithm is improved by enhancing the adaptive behavior of the population and introducing an external warehouse mechanism to generate a snow goose multi-objective optimization algorithm. Parameter initialization: Initialize the population for the snow goose optimization algorithm, calculate the corresponding objective function value based on the initialization position, select the best individual and record the optimal solution; Optimization stage selection: The snow goose formation is divided into two types: herringbone and straight. When the snow geese deflect too much in the herringbone formation and are ready to enter the development stage, the restoration of the normal formation is controlled by the parameter θ with greater randomness; The asymptotic development and exploration control factor cycle is introduced to replace the original matrix angle parameter θ. Under the action of cycle, the probability of entering the exploration stage will decrease with the increase of the number of iterations T during the entire iteration process, while the development stage is the opposite, and it is repeated C times to balance the global search and local development intensity. The stage selection method is as follows: Where, T max is the maximum number of iterations; C is the number of loops; rand2 is a random number between [0,1]; After stage selection, the specific modeling methods of the two queue improvement models are as follows: The herringbone square model in the exploration phase: In the algorithm exploration phase, snow geese are divided into leading geese, weak geese, and other groups with different health conditions: In the formula, They represent the position of the i-th snow goose and the position of the optimal snow goose at the t-th iteration respectively; They represent the center position of the population at the tth iteration and the position of the snow goose with the lowest ranking after the population is sorted; B, D, and E represent the weight coefficients of the corresponding components; represents the speed of the i-th snow goose at the t-th iteration; In the exploration phase, the migration behavior of weak snow geese was simulated to make the snow geese fly at different speeds to the center of the population, thus enhancing the overall flight efficiency of the population. Let F pro To follow the protection factor of the superior population, the snow geese in the last 1 / 5 of the simulated population are as follows: The linear array in the development stage: When the angle of the snow goose herringbone array exceeds the cycle factor selected in the optimization stage, it enters the second stage of flight, that is, the algorithm development stage. In this stage, the idea of ​​high-quality population radius is introduced to guide the population, and the improved dynamic simulation snow goose flight speed is as follows: Where rand1 is a random number between [0,1], which is used to indicate whether the snow goose is in the local optimal solution; Brownian(dim) represents the standard Brownian motion, dim is the variable dimension; ⊕ represents the entry-wise multiplication, is the radius of the high-quality population; N is the physical decline coefficient caused by long-term flight of snow geese; In the linear array, the physical decline coefficient N decreases with the increase of the number of algorithm iterations, simulating the decline in physical strength of snow geese in the later stage of migration, which causes the speed to decrease. At the same time, the radius of high-quality population is introduced. Enhance the ability of snow geese to follow high-quality flocks; Non-dominated solution storage and maintenance: The external warehouse mechanism is introduced to improve the Snow Goose optimization algorithm. The proposed external warehouse mechanism mainly includes the non-dominated solution selection, non-dominated solution storage, and external warehouse update and maintenance stages; Assume that when solution S1 dominates solution S2, let the corresponding objective function value of each solution be recorded as f, then we have: make represents the normalized goodness of the ith solution under the mth objective function, then the usage of the ith non-dominated solution can be calculated, thereby eliminating non-dominated solutions with high congestion; the calculation method is: Optimal solution update: The population position obtained by the t-th update is brought into the objective function to calculate the objective function value, and the local optimal solution obtained by the t-th iteration is compared with the global optimal solution obtained by the k-1 iteration, and the global optimal solution is updated; Iterative update: Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, repeat the iterative update.

Images