Wind-light-storage energy system optimization scheduling method containing optimal dynamic grid-connected coefficient strategy

Abstract

The invention relates to a wind-light-storage energy system optimal scheduling method containing an optimal dynamic grid-connected coefficient strategy, which comprises the following steps of: 1, constructing an HDEED model, and designing an objective function and a constraint condition of the HDEED model; 2, designing an optimal dynamic grid-connected coefficient strategy to stabilize the fluctuation of new energy power generation; 3, establishing an improved thermal power test system comprising a photovoltaic power station and a wind power station, and setting test system parameters; 4, designing a GWOH algorithm, initializing parameters of the GWOH algorithm, and defining a target function Fobj as a fitness function Fit of the algorithm; and step 5, obtaining a Pareto optimal leading edge of the HDEED model by adopting a GWOH algorithm, obtaining a compromise solution by utilizing a fuzzy theory, and obtaining a scheduling scheme for comprehensively optimizing the economical efficiency and the environmental protection property of the system through an optimal dynamic grid-connected strategy on the premise of stabilizing the fluctuation of new energy power generation.

Description

Technical Field

[0001] The present invention belongs to the field of renewable energy power generation and aims to suppress the problem of renewable energy grid-connected power fluctuation. The present invention proposes a wind-solar-storage energy system optimization scheduling method with an optimal dynamic grid-connected coefficient strategy. The method adopts an optimal dynamic grid-connected coefficient strategy to smooth the volatility of renewable energy power generation, and uses a GWOH solution algorithm to solve the high-complexity, high-coupling HDEED problem. Background Art

[0002] With the continuous development of industry, the lack of fossil energy and the serious excess of carbon emissions have become urgent problems that countries around the world need to solve. The power industry is an important area of ​​economic development, and it is also a key area that consumes fossil energy and emits harmful gases. New energy such as solar energy and wind energy is the key to solving the above problems. However, new energy power generation has significant uncertainty and randomness. How to suppress the volatility of new energy power generation in the process of power system scheduling, and how to design a solution method suitable for solving high complexity and strong nonlinear optimization scheduling problems, has always been a hot topic and difficult problem in research. The power system is considered to be the most complex and largest system in the world, and its optimal scheduling has always been a research hotspot. Optimal scheduling refers to the process of achieving the optimal state of the system's comprehensive indicators by reasonably allocating unit output under specific constraints.

[0003] In terms of the number of optimization targets, optimal scheduling is divided into single-target optimization scheduling and multi-target optimization scheduling. Traditional economic scheduling is a form of single-target optimal scheduling. Decision makers only consider the economy of the system when formulating scheduling plans. As energy conservation and emission reduction have become an inevitable trend in the development of the power industry, traditional economic scheduling can no longer meet actual needs. Multi-target optimization scheduling of economic emissions that emphasizes environmental protection and economic coordination has become the pursuit and development direction of power scheduling. Considering the cross-section and time coupling during scheduling, optimal scheduling can be divided into static scheduling and dynamic scheduling. Static optimal scheduling only optimizes the output of the unit at a certain moment to obtain the optimal unit output allocation plan. Dynamic optimal scheduling considers time period coupling and load demand changes. The goal is to obtain the optimal output allocation plan for a scheduling cycle when the unit meets the characteristic constraints. Compared with static optimal scheduling, dynamic optimal scheduling considers the coupling between time periods, which enables it to have the ability to adjust the load in advance, so its modeling is also more complex. To obtain green and clean electricity, the transformation and upgrading of the power industry is imminent, so hybrid dynamic economic emission scheduling (HDEED) considering new energy generation has become a development direction.

[0004] In summary, existing studies have mostly solved the uncertainty problem of renewable energy generation by rotating reserve constraints or considering transmission losses when constructing HDEED models, and have not proposed a specific reliable strategy to smooth the fluctuations in renewable energy generation power. Summary of the invention

[0005] The technical problem to be solved by the present invention is to construct a wind-solar-storage energy system optimization scheduling method with an optimal dynamic grid-connected coefficient strategy. The present invention proposes a novel optimal dynamic grid-connected coefficient strategy, which uses the flexible energy time-shifting of energy storage units to suppress the fluctuation of new energy grid-connected power, and maximizes the economic benefits of the power system through the "low storage and high generation" strategy. At the same time, the present invention also proposes a gray wolf optimization algorithm with a hierarchical order weight allocation strategy, the GWOH algorithm, which is suitable for solving high-complexity and high-coupling HDEED problems. Compared with existing methods, the present invention helps to improve the utilization rate of traditional energy and new energy.

[0006] The present invention solves the technical problem by adopting the following scheme:

[0007] A method for optimizing the dispatching of a wind-solar-storage energy system with an optimal dynamic grid-connected coefficient strategy comprises the following steps:

[0008] Step 1: Construct the HDEED model and design its objective function and constraints;

[0009] Step 2: Design an optimal dynamic grid-connection coefficient strategy to smooth out the volatility of renewable energy generation;

[0010] Step 3: Establish an improved thermal power test system including photovoltaic power stations and wind farms, and set test system parameters;

[0011] Step 4: Design the GWOH algorithm, initialize its parameters, and transform the objective function F obj Defined as the fitness function Fit of the algorithm;

[0012] Fit=F obj

[0013] Step 5: Use the GWOH algorithm to obtain the Pareto optimal frontier of the HDEED model, use fuzzy theory to obtain a compromise solution, and use the optimal dynamic grid-connected strategy to obtain the best scheduling plan for system economy and environmental protection while smoothing the volatility of new energy power generation.

[0014] The process of GWOH algorithm includes:

[0015] First, the wolf pack reorganization coefficient p (p = 0.6) is set. By reorganizing the wolf pack positions in the late iteration, the wolf pack maintains diversity throughout the entire iteration process, improving the wolf pack's ability to balance global exploration and local development during the solution process. If the random number r < p, the individual positions in the wolf pack are updated according to the following expression:

[0016]

[0017] In the formula is the average position of the wolf pack; F i is a random number in the interval [0, 2].

[0018] In addition, in the GWO algorithm, the average position method is used to update the positions of other individuals in the wolf pack. The advantage of this update method is that the principle is simple and easy to implement, but it ignores the importance of the wolf pack's hierarchical order to the update of other individual positions. Specifically, when updating the positions of other individuals in the wolf pack, the hierarchical order of the wolf pack should be highlighted, especially the degree of influence of α, β, and δ wolves on other individuals. To this end, the present invention designs a hierarchical order weight allocation strategy, which constructs the allocation weight according to the fitness values ​​of α, β, and δ wolves, as shown in the following formula:

[0019]

[0020] Where w α 、w β 、w δ is the hierarchical order weight; Fit(α), Fit(β), and Fit(δ) are the fitness values ​​of α, β, and δ wolves respectively.

[0021] In order to make the algorithm better balance the global exploration and local search capabilities during the iteration process, the present invention designs a nonlinear decreasing coefficient k * . Relative to k1, k * This makes individual wolves more adaptable in the problem-solving process.

[0022] The specific steps of the wind-solar-storage HDEED model with the optimal dynamic grid-connected coefficient strategy and the GWOH solution method are as follows:

[0023] (1) Construct a HDEED model with an optimal dynamic grid-connected coefficient strategy and design the objective function and constraints of the HDEED model.

[0024] 1) Construct the objective function

[0025] The fuel cost function of thermal power units with valve point effect consists of two parts: one is the fuel cost function of thermal power units simulated by quadratic function, and the other is the cost generated by valve point effect simulated by sine function. The model is:

[0026]

[0027] Where N f is the number of thermal power units; T d is the scheduling period; d, e, f, g, h are the fuel cost coefficients; p f，min Indicates the minimum output of a thermal power unit.

[0028] For wind farms, photovoltaic power stations, and energy storage units, the present invention only considers their operation and maintenance costs, and the model is as follows:

[0029]

[0030] Where N w , N pv , N soc are the number of wind farms, photovoltaic power stations and energy storage units respectively; c w , c pv , c soc They are the operation and maintenance cost coefficients of wind farms, photovoltaic power stations and energy storage units respectively.

[0031] The cost function of the HDEED model is:

[0032]

[0033] Wind farms, photovoltaic power stations, and energy storage units do not emit polluting gases during operation, only thermal power units emit polluting gases during operation. The polluting gas emission model of the HDEED model is:

[0034]

[0035] In the formula, α, β, δ, η represent the pollution emission coefficients of the generator sets.

[0036] HDEED model contains F cost With F emission The two objectives belong to a multi-objective optimization problem. The present invention adopts a priori method to solve the HDEED model, and sets the weight coefficient C w , transform the multi-objective into a single objective and obtain the optimization objective F obj for:

[0037] F obj =C w ×F cost +(1-C w )×F emission

[0038] Where G w ∈[0, 1].

[0039] 2) Build constraints

[0040] In order not to significantly increase the complexity of the model and ensure the scientific nature of the decision-making scheme, the invention focuses on the following constraints:

[0041] ① Power balance constraints:

[0042]

[0043] In the formula Indicates the grid-connected power standard value of the new energy system at time i; represents the load demand at the i-th moment; represents the transmission loss at the i-th moment; Represents the power output of the energy storage unit at the i-th moment (when When , the energy storage unit discharges; on the contrary, when When the energy storage unit is charged); and is the power transmission loss coefficient.

[0044] ② Constraints on output power of thermal power units:

[0045]

[0046] In the formula represents the lower limit of the power output of the jth thermal power unit at time i, It represents the upper limit of the power output of the j-th thermal power unit at time i.

[0047] ③ Output power constraints of wind and photovoltaic power generation systems:

[0048]

[0049] Where p w，max Indicates the upper limit of wind power generation system output power; p pv，max Indicates the upper limit of the output power of the photovoltaic power generation system.

[0050] ④ Constraints on the ramp rate of thermal power units:

[0051]

[0052] Where ΔT (ΔT = 1) represents the scheduling time interval; represents the ramp rate of the jth thermal power unit, Represents the ramp rate of the j-th thermal power unit.

[0053] ⑤Positive and negative spinning reserve constraints:

[0054] The present invention copes with the uncertainty of the output of wind power and photovoltaic power generation systems through positive and negative spinning reserve constraints.

[0055] Positive spinning reserve constraints:

[0056]

[0057] In the formula The positive spinning reserve capacity provided for the j-th thermal power unit at time i; r f，upis the positive spinning reserve capacity factor of the thermal power unit to cope with unit outage or load forecast error; r w，up It represents the positive spinning reserve capacity factor of thermal power units to cope with the output power forecast error of wind power generation system; r pv，up It represents the positive spinning reserve capacity factor of the thermal power unit to deal with the output power prediction error of the photovoltaic power generation system.

[0058] Negative spinning reserve constraints:

[0059]

[0060] In the formula The negative spinning reserve capacity provided by the j-th thermal power unit at time i; r f，down is the negative spinning reserve capacity factor of the thermal power unit to cope with unit outage or load forecast error; r w，down It represents the negative spinning reserve capacity factor of thermal power units to cope with the output power forecast error of wind power generation system; r pv，down It represents the negative spinning reserve capacity factor of thermal power units to cope with the output power forecast error of photovoltaic power generation system.

[0061] ⑥Charge and discharge power constraints of energy storage units:

[0062]

[0063] In the formula is the maximum charging power of the energy storage unit at the i-th moment; is the maximum discharge power of the energy storage unit at the i-th moment.

[0064] In addition, in order to ensure the normal operation of the energy storage unit during operation, the mutual exclusion conditions of charging and discharging operations need to be met:

[0065]

[0066] ⑦Charge ratio constraints of energy storage units:

[0067]

[0068] In the formula is the lower limit of the charge ratio at the i-th moment; is the upper limit of the charge ratio at the i-th moment; is the charge ratio value of the energy storage unit at the i-th moment, and its calculation formula is as follows:

[0069]

[0070] In the formula is the remaining power of the energy storage unit at time i; is the maximum storage capacity of the energy storage unit at moment i.

[0071] To ensure the sustainability of the charging and discharging operations of the energy storage unit when participating in the dispatch, the total charging power of the energy storage unit should be greater than or equal to the total discharging power within a dispatch cycle, that is:

[0072]

[0073] (2) Constructing a comprehensive optimization model based on the optimal dynamic grid-connected coefficient strategy

[0074] In order to smooth out the volatility of grid-connected power of renewable energy power generation, the present invention proposes a dispatching strategy based on the optimal dynamic grid-connected coefficient. The biggest advantage of this strategy is that it can reduce the volatility of grid-connected power of renewable energy while improving the economic benefits of the system by utilizing the "low storage and high generation" strategy of energy storage units. Its basic principle is: divide the load curve into three time periods: valley, flat and peak. Set the grid-connected standard power values ​​of wind, solar and storage systems to A_Ongrid, B1_Ongrid, B2_Ongrid and C_Ongrid for different load periods, and connect renewable energy power generation to the grid according to the set grid-connected power values.

[0075] The definition of grid-connected standard value is as follows:

[0076]

[0077] Where c a (0≤c a ≤1), c b (c b ≥1) and c c (0≤c c ≤1) are all dynamic grid-connected power coefficients, P a，min is the minimum grid-connected power value during the load valley period, P c，min is the minimum grid-connected power value during the load flat period, P bx，max It is the maximum value of grid-connected power during the peak load period. Adjusting the dynamic grid-connected coefficient in different load periods allows decision makers to implement different dispatching strategies based on the different economic and environmental benefits of the system.

[0078] The specific operation is as follows: First, during the load valley and flat period, the wind and solar power are connected to the grid according to the corresponding grid-connected standard value, and the wind and solar power that is higher than the grid-connected standard value is stored in the energy storage unit, which can achieve "low storage" while completing the smooth grid connection of wind and solar power. Secondly, during the load peak period, the wind and solar power are connected to the grid according to the set grid-connected standard value, and the energy storage unit discharges to further reduce the load peak, which can achieve "high generation" while completing the smooth grid connection of wind and solar power. In this strategy, the stability of the power of the power grid can be ensured in different load periods, and the dynamic economic benefits of "low storage and high generation" can be achieved by shifting the power stored in the load valley and flat period to the load peak period.

[0079] (3) Establish an improved thermal power test system including photovoltaic power stations and wind farms, set test system parameters, and obtain wind and solar power generation power.

[0080] The present invention is based on the IEEE-10 unit 39-node thermal power system and establishes an improved test system including a photovoltaic power station and two wind farms.

[0081] According to the load demand at different dispatching times, the load at different dispatching times is divided into three parts: low valley area, flat valley area and peak area. At the same time, the electricity price at different dispatching times needs to be obtained by referring to the local electricity price.

[0082] (4) Design the GWOH algorithm, initialize its parameters, such as population size, number of iterations, etc., and set the objective function F obj Defined as the fitness function Fit of the algorithm.

[0083] Set the maximum number of iterations of the algorithm Max_iter, the population size of the algorithm Pop, and the wolf pack recombination coefficient p. According to the dominance of different wolves in the wolf pack, the wolves are divided into α, β, δ, and ω. The α wolf has the strongest dominance and the best fitness value, and is usually regarded as the optimal solution in the population;

[0084] Use the GWOH algorithm to solve the HDEED problem, obtain the Pareto frontier solution of the problem, and set the objective function F of HDEED obj Defined as the algorithm's fitness function Fit:

[0085] Fit=F obj

[0086] (5) The GWOH algorithm is used to obtain the Pareto optimal frontier of the HDEED model, and the fuzzy theory is used to obtain a compromise solution. The optimal dynamic grid-connected strategy is used to obtain the best scheduling plan that combines the best system economy and environmental protection while smoothing the volatility of renewable energy power generation.

[0087] The specific steps of the GWOH algorithm are:

[0088] (1) Initialize the parameters of the GWOH algorithm, determine the maximum number of iterations of the wolf pack Max_iter, and the population size of the algorithm Pop; divide the wolves into α, β, δ, and ω according to the fitness values ​​of different wolves in the wolf pack. The α wolf has the largest fitness value and is usually regarded as the optimal solution in the population. The dominance of the β, δ, and ω wolves decreases in turn.

[0089] (2) Set the wolf pack recombination coefficient p (p = 0.6). If the random number r < p, then update the individual positions in the wolf pack according to the following formula, and calculate the individual fitness values ​​of the wolves to update α, β, δ and ω:

[0090]

[0091] In the formula, is the average position of the wolf pack; F i is a random number in the interval [0, 2].

[0092] If the random number r>p, calculate the linear attenuation coefficient k * . k * is the nonlinear decreasing coefficient, which decreases nonlinearly with the increase of the number of iterations.

[0093] During the hunting process, the distance between individual wolves and α, β, and δ wolves is:

[0094]

[0095] In the formula, is the individual position in the wolf pack, represents the distance between the gray wolf and its prey, is the coefficient matrix.

[0096] The present invention designs a hierarchical order weight distribution strategy, which constructs distribution weights according to the fitness values ​​of α, β, and δ wolves to calculate the positions of other individuals in the population:

[0097]

[0098] In the formula, The expression of is obtained by the following formula:

[0099]

[0100] (3) Obtain the Pareto frontier solution of the HDEED problem.

[0101] Repeat process (2) until the maximum number of iterations is reached, and output the position of the wolf pack to obtain the Pareto front solution.

[0102] The beneficial effects of the present invention are:

[0103] (1) The method of the present invention uses the optimal dynamic grid-connected coefficient strategy to utilize the charging and discharging flexibility of the energy storage unit to smooth out the fluctuations in the power of the new energy grid-connected power. When the power generation of new energy is unstable or excessive, the excess power is stored, and when the power generation is insufficient, it is released to balance the load of the power grid. This control strategy can effectively suppress the fluctuations in the power of new energy and improve the stability and reliability of the power grid.

[0104] (2) The system cost obtained by the GWOH algorithm proposed in this invention under the optimal dynamic coefficient grid-connected strategy is 4.52% lower than that without new energy, and the system emission is reduced by 11.84%. The results show that the optimal dynamic coefficient grid-connected strategy can not only effectively smooth the fluctuation of new energy power generation, but also improve the economic benefits by using the low storage and high generation strategy.

[0105] (3) The method of the present invention can more accurately match the difference between grid demand and new energy supply by flexibly adjusting the charging and discharging strategy of the energy storage unit, thereby minimizing dependence on non-new energy sources such as traditional coal-fired power and improving overall energy utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Figure 1 It is a schematic flow chart of a method for optimizing the dispatching of a wind-solar-storage energy system including an optimal dynamic grid-connected coefficient strategy according to the present invention.

[0107] Figure 2 It is the Pareto frontier obtained by each algorithm in Example 1.

[0108] Figure 3 This is the output of each unit at each scheduling time obtained by the GWOH algorithm in Example 1.

[0109] Figure 4 It is the Pareto frontier obtained by each algorithm in Example 2.

[0110] Figure 5 This is the output of each unit at each scheduling time obtained by the GWOH algorithm in Example 2.

[0111] Figure 6 It is the Pareto frontier obtained by each algorithm in Example 3.

[0112] Figure 7 This is the output of each unit at each dispatching time obtained by the GWOH algorithm in Example 3.

[0113] Figure 8 It is the charging and discharging power and quantity of the energy storage unit at each scheduling time in Example 3.

[0114] Fig. 9 It is the standard power of new energy grid-connected at each scheduling time in Example 2 and Example 3.

[0115] Fig.10 It is the system power generation cost and pollutant gas emissions at each scheduling time of the scheduling scheme obtained by the GWOH algorithm when implementing three different cases. DETAILED DESCRIPTION

[0116] In order to illustrate the technical solution of the present invention, specific embodiments will be described below in conjunction with the accompanying drawings.

[0117] The present invention provides a wind-solar-storage energy system optimization scheduling method with an optimal dynamic grid-connected coefficient strategy. By constructing a HDEED model, designing its objective function and constraints, an optimal dynamic grid-connected coefficient strategy is proposed to smooth the volatility of renewable energy power generation, an improved thermal power test system including photovoltaic power stations and wind farms is established, test system parameters are set, a GWOH algorithm is designed, its parameters are initialized, and the objective function F is converted into a grid-connected coefficient. obj The fitness function Fit of the algorithm is defined as the algorithm. The GWOH algorithm is used to obtain the Pareto optimal frontier of the HDEED model. The fuzzy theory is used to obtain a compromise solution. The optimal dynamic grid-connected strategy is used to obtain the best scheduling plan for system economy and environmental protection under the premise of smoothing the volatility of renewable energy power generation.

[0118] Figure 1 The process of the wind-solar-storage energy system optimization scheduling method with the optimal dynamic grid-connected coefficient strategy of the present invention is as follows: constructing the HDEED model and improving the IEEE-10 unit thermal power generation test system → initializing the parameters and population of the GWOH algorithm → optimizing HDEED through the GWOH algorithm → obtaining the optimal scheduling scheme of HDEED.

[0119] Example 1

[0120] HDEED model without new energy

[0121] Figure 2 The Pareto frontier obtained by solving the HDEED model in Example 1 using the GWO, SSA, MFO, CSA and WOA algorithms is shown. The Pareto frontier solutions obtained by the GWOH, GWO, SSA and CSA algorithms have a wider distribution range, while the distribution range of the MFO and WOA algorithms is narrower. From the distribution performance of the optimal frontier, the GWOH, GWO and SSA algorithms perform well, and the continuity and smoothness of the obtained Pareto frontier solutions are good, indicating that the algorithms can adapt to the high complexity of the HDEED model. However, the distribution performance of the Pareto frontier obtained by the WOA algorithm is poor, indicating that the quality of the obtained solution is low and the algorithm cannot converge to the optimal Pareto frontier. Figure 2 As can be seen from (b), the Pareto optimal frontier obtained by the GWOH algorithm has the characteristics of wide distribution range, good continuity and smoothness, which can provide decision makers with more reasonable and diversified scheduling strategies.

[0122] Three solutions based on various algorithms in this embodiment are shown in Table 1:

[0123] Table 1 Solutions for HDEED without new energy

[0124]

[0125] Scheme A in Table 1 is the best compromise solution obtained based on fuzzy theory, and schemes B and C are boundary solutions that only consider system economic benefits and only consider environmental benefits, respectively.

[0126] As shown in Table 1, for the boundary solution, the result obtained by the GWOH algorithm is the most competitive. In a scheduling cycle, if only the economic benefits of the system are considered, the system power generation cost obtained by the GWOH algorithm is 2.465×10 6 $, which is 4000$, 14000$, 42000$, 85000$ and 100000$ lower than GWO, SSA, MFO, CSA and WOA algorithms respectively. If only the environmental benefits of the system are considered, the system pollutant gas emissions obtained by GWOH algorithm are 1.325×10 5 kg, which is 4.1×10 lower than that of GWO, SSA, MFO, CSA and WOA algorithms, respectively. 3 kg, 3.9×10 3 kg, 6.4×10 3 kg, 1.21×10 4 kg and 2.39×10 4 kg. For the compromise solution, the operating cost of GWOH is 2.470×106$, which is 11000$, 23000$, 49000$, 71000$ and 95000$ lower than that of GWO, SSA, MFO, CSA and WOA algorithms respectively. The pollutant gas emission is 1.416×10 5 kg, slightly higher than the SSA and MFO algorithms. However, the scheduling scheme obtained by the GWOH algorithm has the best comprehensive performance, that is, it maximizes the economic efficiency of the system while maintaining environmental benefits.

[0127] Figure 3 Table 2 shows the output of each unit at each dispatching time obtained by the GWOH algorithm in Example 1. Each stacked bar chart in Figure 4 represents the sum of the output of all units at their corresponding dispatching time, and the length of the bars of different colors represents the output size of each unit. It can be seen from the data in Table 2 that the output power of units U1 to U4 fluctuates greatly at each moment in a dispatching cycle. This is because units U1 to U4 are mainly responsible for peak load regulation. Units U5 to U10 are basically in full-load power generation operation within a dispatching cycle to meet load demand.

[0128] Table 2 Scheduling solutions obtained by GWOH for Case 1

[0129]

[0130]

[0131] Example 2

[0132] HDEED model with optimal dynamic grid-connected coefficient strategy.

[0133] Figure 4 The Pareto frontiers obtained by solving the HDEED model in Example 2 using the GWO, SSA, MFO, CSA and WOA algorithms are shown. Similar to the results obtained in Example 1, the Pareto optimal frontiers obtained by the GWOH, GWO and SSA algorithms in Example 2 still have good continuity and smoothness, indicating that the quality of the solution is high, while the Pareto optimal frontier obtained by the WOA algorithm has poor distribution, and the solution cannot converge to the optimal Pareto optimal frontier. Compared with Figure 2 The Pareto optimal frontier in (b) is Figure 4 The Pareto optimal frontier obtained by the GWOH algorithm in (b) is closer to the lower left end of the coordinate axis, indicating that the addition of new energy improves the economic and environmental benefits of the system. At the same time, the Pareto optimal frontier obtained by the GWOH algorithm can provide decision makers with rich decision possibilities, which is conducive to improving the environment and promoting the sustainable development of energy.

[0134] In this embodiment, the boundary solutions and compromise solutions obtained based on various algorithms are shown in Table 3:

[0135] Table 3 HDEED solution using the optimal dynamic grid-connected coefficient strategy

[0136]

[0137] It can be seen from Table 3 that the best compromise solutions obtained by the SSA and GWOH algorithms are the most competitive. Although the dispatch strategy obtained by the SSA algorithm is 0.21% lower than that of the GWOH algorithm in terms of power generation cost, the dispatch strategy obtained by the GWOH algorithm is 4.52% lower than that of the SSA algorithm in terms of pollutant emissions.

[0138] Compared with the results in Table 1, the power generation costs and pollutant emissions of the dispatching schemes obtained by each algorithm in Table 3 are significantly reduced, indicating that the grid connection of new energy sources helps to improve the economic and environmental benefits of the system. In addition, in the optimal dynamic grid connection coefficient strategy, at the load valley and flat moments, the wind and solar grid-connected power is adjusted according to the strategy, and the remaining wind and solar power at the load valley and flat moments is stored in the energy storage unit, and then released at the load peak moment. In this process, the wind, solar and storage system does not generate pollutants during operation. At the same time, thanks to the good benefits of the low storage and high generation strategy, this strategy can significantly reduce the cost and pollutant emissions of the system.

[0139] Figure 5 Table 4 shows the output of each unit at each dispatching time obtained by the GWOH algorithm in Example 2. As can be seen from Table 4, units U1 to U4 are mainly responsible for peak load regulation to meet the load demand at each moment during the dynamic dispatching process; units U5 to U10 are basically in full load power generation operation during the entire dispatching cycle. Figure 5 It directly reflects that the output values ​​of units U1 to U4 at each dispatching time fluctuate greatly, while the output values ​​of units U5 to U10 at each dispatching time are relatively stable. The optimal dynamic grid-connected coefficient strategy makes the output of the new energy system lower at the load level and valley time, and higher at the load peak time, achieving the dispatching purpose of "low storage and high generation" and "energy time shifting". At the same time, new energy is connected to the grid at a fixed value at different dispatching times, effectively smoothing the fluctuation of the new energy grid-connected power.

[0140] Table 4 Scheduling solutions obtained by GWOH for Case 2

[0141]

[0142] Note: Gcp is the grid-connected power of new energy

[0143] Example 3

[0144] Changes in key system performance parameters under dynamic grid-connected parameters.

[0145] In Example 3, the lowest c b Adjust the grid-connected power during peak hours. Minimum c b Under this circumstance, the strategy can still achieve the goal of smoothing the volatility of renewable energy grid-connected power, but it cannot realize the profitability of energy storage units.

[0146] Figure 6 All of them show the minimum c using GWOH, GWO, SSA, MFO, CSA and WOA algorithms respectively. b The Pareto optimal frontier obtained by solving the HDEED model. Figure 6It can be seen that the continuity and smoothness of the Pareto optimal frontiers obtained by the GWOH, GWO and SSA algorithms are the best. Compared with the Pareto optimal frontiers obtained by each algorithm in Example 1 and Example 2, the GWOH, GWO and SSA algorithms all show good adaptability in solving complex multi-objective dynamic optimization problems, among which the distribution of the Pareto optimal frontier obtained by the GWOH algorithm is the most competitive. In addition, comparing the Pareto optimal frontiers obtained in Example 1, Example 2 and Example 3, the solution in Example 3 is better than that in Example 1, but worse than that in Example 2, indicating that the lowest c b It can only ensure the stability of the grid-connected power of renewable energy, but cannot improve the system benefits.

[0147] In this embodiment, the boundary solutions and compromise solutions obtained based on various algorithms are shown in Table 5:

[0148] Table 5 Solution for Case 3

[0149]

[0150]

[0151] From Table 5, it can be seen that the scheduling solutions obtained by the GWOH algorithm are highly competitive, whether for the boundary solution or the compromise solution. It can provide decision makers with a scheduling solution that is both economical and environmentally friendly. Compared with the results in Tables 1 and 3, the results in Table 5 are better than those in Table 1, but worse than those in Table 3. This is because the minimum c b Under this condition, the arbitrage income of the energy storage unit is the lowest, which cannot guarantee the optimal economic benefits of the system. However, combined with the optimal c b , which can provide decision makers with rich and diverse scheduling solutions.

[0152] Figure 7 Table 6 shows the output of each unit at each dispatching time obtained by the GWOH algorithm in Example 3. Compared with the results in Table 4, it can be seen that the grid-connected power of new energy in Table 6 at the peak load moment is less than the grid-connected power at the corresponding moment in Table 4, which reflects the difference in dispatching strategies between Example 2 and Example 3. At each dispatching moment, the energy storage unit can intervene in time and connect the new energy power to the grid at a fixed value, which effectively smooths the volatility of the grid-connected power. Figure 7 It can be seen that the grid-connected power of new energy is low at load level and valley moments, and high at load peak moments, achieving the scheduling goals of "low storage and high generation" and "energy time shifting".

[0153] Table 6 Scheduling solutions obtained by GWOH for Case 3

[0154]

[0155] The dispatch strategies obtained in Example 2 and Example 3 show that the introduction of new energy can significantly reduce the output of thermal power units and effectively improve the economic and environmental benefits of the system. Under the optimal dynamic grid-connected coefficient strategy, all performance parameters of the system reach the optimal state.

[0156] Example 3: The charging and discharging power and quantity of the energy storage unit at each scheduling time are as follows: Figure 8 As shown. Figure 8 It can be seen that within a scheduling cycle, the total discharge power of the energy storage unit is 207.47MW, and the total charging power is 296.61MW. The charging power is greater than the discharge power, which is conducive to maintaining the sustainability of the energy storage unit's participation in system scheduling. In addition, the initial power of the energy storage unit at time 0 is 50MW.h, and the remaining power of the energy storage unit at time 24 is 139.15MW. Compared with the power of the energy storage unit in Example 2, the energy storage unit in Example 3 maintains sufficient energy storage power at each scheduling time, which is conducive to timely compensation when the power grid volatility is large, and smoothing the grid-connected power volatility. Minimum c b In this way, not only can the revenue and expenditure of the energy storage unit be balanced, but also the volatility of the grid-connected power can be minimized to the greatest extent possible, ensuring the stability of the grid operation. At the same time, after the end of this scheduling cycle, sufficient energy storage margin can be used to support decision makers in formulating more diversified scheduling strategies in the next scheduling cycle.

[0157] In Example 2 and Example 3, the standard power of renewable energy grid connection at each dispatching time is as follows: Fig. 9 Shown by Fig. 9 It can be seen that the grid-connected power of renewable energy is a single fixed value at different dispatching times, which effectively smooths the volatility of grid-connected power while taking into account the economy of system operation.

[0158] Under three different cases, the system power generation cost and pollutant gas emissions of the scheduling scheme obtained by the GWOH algorithm at each scheduling time are as follows: Fig.10 As shown. Fig.10 (a) It can be seen that in the load valley and flat period, the system power generation cost obtained in Example 2 and Example 3 is less than that in Example 1, indicating that the introduction of new energy is conducive to reducing the system power generation cost. Fig.10 (b) It can be seen that the system pollutant emissions obtained in Example 2 during the load valley and flat periods are the lowest, indicating that the optimal dynamic grid-connected coefficient strategy can reasonably allocate the output of each unit and help reduce system emissions. During the load peak period, there are large differences in the system power generation costs and pollutant gas emissions obtained under different cases. In summary, the intervention of new energy can significantly improve the economic and environmental benefits of the system. At the same time, the dispatching scheme obtained under the optimal dynamic grid-connected system strategy has the best economic and environmental performance.

[0159] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation manner. Any matters not described in the present invention that are obtained by those skilled in the art based on the technical solution of the present invention are applicable to the prior art.

Claims

1. An optimal dispatching method for wind-solar-storage energy system with an optimal dynamic grid-connected coefficient strategy: Step 1: Construct the HDEED model and design its objective function and constraints; Step 2: Design an optimal dynamic grid-connection coefficient strategy to smooth out the volatility of renewable energy generation; Step 3: Establish an improved thermal power test system including photovoltaic power stations and wind farms, and set test system parameters; Step 4: Design the GWOH algorithm, initialize its parameters, and transform the objective function F obj Defined as the fitness function Fit of the algorithm; Fit=F obj Step 5: Use the GWOH algorithm to obtain the Pareto optimal frontier of the HDEED model, use fuzzy theory to obtain a compromise solution, and use the optimal dynamic grid-connected strategy to obtain the best scheduling plan for system economy and environmental protection while smoothing the volatility of new energy power generation.

2. The method according to claim 1, characterized in that The optimal dynamic grid-connected coefficient strategy includes: In order to smooth out the fluctuation of the grid-connected power of renewable energy power generation, the present invention proposes a dispatching strategy based on the optimal dynamic grid-connected coefficient; the biggest advantage of this strategy is that it can improve the economic benefits of the system by utilizing the "low storage and high generation" of energy storage units while suppressing the fluctuation of the grid-connected power of renewable energy power generation. Its basic principle is: divide the load curve into three time periods: valley, flat and peak, set the grid-connected standard power values ​​of the wind, solar and storage systems to A_Ongrid, B1_Ongrid, B2_Ongrid and C_Ongrid for different load periods, and connect the renewable energy power generation to the grid according to the set grid-connected power values; The definition of grid-connected standard value is as follows: Where c a (0≤c a ≤1), c b (c b ≥1) and c c (0≤c c ≤1) are all dynamic grid-connected power coefficients, P a，min is the minimum grid-connected power value during the load valley period, P c，min is the minimum grid-connected power value during the load flat period, P bx，max The maximum grid-connected power value during the peak load period; adjusting the dynamic grid-connected coefficient in different load periods allows decision makers to implement different dispatching strategies based on the different economic and environmental benefits of the system; The specific operations are as follows: First, during the load valley and flat periods, the wind and solar power are connected to the grid according to the corresponding grid-connected standard value, and the wind and solar power that is higher than the grid-connected standard value is stored in the energy storage unit, which can achieve "low storage" while completing the smooth grid connection of the wind and solar power; secondly, during the load peak period, the wind and solar power are connected to the grid according to the set grid-connected standard value, and the energy storage unit discharges to further reduce the load peak, which can achieve "high generation" while completing the smooth grid connection of the wind and solar power; in this strategy, the stability of the grid power can be ensured in different load periods, and the dynamic economic benefits of "low storage and high generation" can be achieved by shifting the power stored in the load valley and flat periods to the load peak period for release.

3. The method according to claim 1, characterized in that The specific steps of the GWO algorithm with hierarchical order weight allocation strategy, namely the GWOH algorithm, are: First, the wolf pack reorganization coefficient p (p = 0.6) is set; by reorganizing the wolf pack positions in the late iteration, the wolf pack maintains diversity throughout the entire iteration process, improving the wolf pack's ability to balance global exploration and local development during the solution process; if the random number r < p, the individual positions in the wolf pack are updated according to the following expression: In the formula is the average position of the wolf pack; F i is a random number in the interval [0, 2]; In addition, in the GWO algorithm, the average position method is used to update the positions of other individuals in the wolf pack; the advantage of this updating method is that the principle is simple and easy to implement, but it ignores the importance of the hierarchy of the wolf pack to the update of the positions of other individuals; specifically, when updating the positions of other individuals in the wolf pack, the hierarchy of the wolf pack should be highlighted, especially the influence of α, β, and δ wolves on other individuals; for this reason, the present invention designs a hierarchy weight allocation strategy, and constructs the allocation weight according to the fitness value of α, β, and δ wolves, as shown in the following formula: Where w α 、w β 、w δ is the hierarchical order weight; Fit(α), Fit(β), and Fit(δ) are the fitness values ​​of α, β, and δ wolves respectively.

4. The method according to claim 1, characterized in that: In order to make the algorithm better balance the global exploration and local search capabilities during the iteration process, the present invention also designs a nonlinear decreasing coefficient k * ; relative to k1, k * This makes individual wolves more adaptable in the problem-solving process.

5. A method for optimizing the dispatching of a wind-solar-storage energy system with an optimal dynamic grid-connected coefficient strategy, the specific steps of the method are: (1) Construct the HDEED model and design its objective function and constraints 1) Construct the objective function The fuel cost function of thermal power units with valve point effect consists of two parts: one is the fuel cost function of thermal power units simulated by quadratic function, and the other is the cost generated by valve point effect simulated by sine function. The model is: Where N f is the number of thermal power units; T d is the scheduling period; d, e, f, g, h are the fuel cost coefficients; p f，min Indicates the minimum output of thermal power units; For wind farms, photovoltaic power stations, and energy storage units, the present invention only considers their operation and maintenance costs, and the model is as follows: Where N w , N pv , N soc are the number of wind farms, photovoltaic power stations and energy storage units respectively; c w , c pv , c soc are the operation and maintenance cost coefficients of wind farms, photovoltaic power plants, and energy storage units, respectively; The cost function of the HDEED model is: Wind farms, photovoltaic power stations, and energy storage units do not emit polluting gases during operation, and only thermal power units emit polluting gases during operation; the polluting gas emission model of the HDEED model is: In the formula, α, β, δ, η represent the pollution emission coefficients of the generator sets; HDEED model contains F cost With F emission The two objectives belong to a multi-objective optimization problem. The present invention adopts a priori method to solve the HDEED model, and sets the weight coefficient C w , transform the multi-objective into a single objective and obtain the optimization objective F obj for: F obj =C w ×F cost +(1-C w )×F emission Where C w ∈[0, 1]; 2) Build constraints In order not to significantly increase the complexity of the model and ensure the scientific nature of the decision-making scheme, the invention focuses on the following constraints: ① Power balance constraints: In the formula Indicates the grid-connected power standard value of the new energy system at time i; represents the load demand at the i-th moment; represents the transmission loss at the i-th moment; Represents the power output of the energy storage unit at the i-th moment (when When , the energy storage unit discharges; on the contrary, when When the energy storage unit is charged); and is the power transmission loss coefficient; ② Constraints on output power of thermal power units: In the formula represents the lower limit of the power output of the jth thermal power unit at time i, It represents the upper limit of the power output of the jth thermal power unit at time i; ③ Output power constraints of wind and photovoltaic power generation systems: Where p w，max Indicates the upper limit of wind power generation system output power; p pv，max Indicates the upper limit of the output power of the photovoltaic power generation system; ④ Constraints on the ramp rate of thermal power units: Where ΔT (ΔT = 1) represents the scheduling time interval; represents the ramp rate of the jth thermal power unit, represents the ramp rate of the jth thermal power unit; ⑤Positive and negative spinning reserve constraints The present invention copes with the uncertainty of wind and photovoltaic power generation system output through positive and negative spinning reserve constraints; Positive spinning reserve constraints: In the formula The positive spinning reserve capacity provided for the j-th thermal power unit at time i; r f，up is the positive spinning reserve capacity factor of the thermal power unit to cope with unit outage or load forecast error; r w，up It represents the positive spinning reserve capacity factor of thermal power units to cope with the output power forecast error of wind power generation system; r pv，up It represents the positive spinning reserve capacity factor of the thermal power unit to cope with the output power forecast error of the photovoltaic power generation system; Negative spinning reserve constraints: In the formula The negative spinning reserve capacity provided by the j-th thermal power unit at time i; r f，down is the negative spinning reserve capacity factor of the thermal power unit to cope with unit outage or load forecast error; r w，down It represents the negative spinning reserve capacity factor of thermal power units to cope with the output power forecast error of wind power generation system; r pv，down It represents the negative spinning reserve capacity factor of the thermal power unit to cope with the output power forecast error of the photovoltaic power generation system; ⑥Charge and discharge power constraints of energy storage units: In the formula is the maximum charging power of the energy storage unit at the i-th moment; is the maximum discharge power of the energy storage unit at the i-th moment; In addition, in order to ensure the normal operation of the energy storage unit during operation, the mutual exclusion conditions of charging and discharging operations need to be met: ⑦Charge ratio constraints of energy storage units: In the formula is the lower limit of the charge ratio at the i-th moment; is the upper limit of the charge ratio at the i-th moment; is the charge ratio value of the energy storage unit at the i-th moment, and its calculation formula is as follows: In the formula is the remaining power of the energy storage unit at time i; is the maximum storage capacity of the energy storage unit at time i; To ensure the sustainability of the charging and discharging operations of the energy storage unit when participating in the dispatch, the total charging power of the energy storage unit should be greater than or equal to the total discharging power within a dispatch cycle, that is: (2) Designing an optimal dynamic grid-connection coefficient strategy to smooth out the volatility of renewable energy generation In order to smooth out the fluctuation of the grid-connected power of renewable energy power generation, the present invention proposes a dispatching strategy based on the optimal dynamic grid-connected coefficient; the biggest advantage of this strategy is that it can reduce the fluctuation of the grid-connected power of renewable energy while utilizing the "low storage and high generation" of energy storage units to improve the economic benefits of the system. Its basic principle is: divide the load curve into three time periods: valley, flat and peak, set the grid-connected standard power values ​​of the wind, solar and storage systems to A_Ongrid, B1_Ongrid, B2_Ongrid and C_Ongrid for different load periods, and connect the renewable energy power generation to the grid according to the set grid-connected power values; The definition of grid-connected standard value is as follows: Where c a (0≤c a ≤1), c b (c b ≥1) and c c (0≤c c ≤1) are all dynamic grid-connected power coefficients, P a，min is the minimum grid-connected power value during the load valley period, P c，min is the minimum grid-connected power value during the load flat period, P bx，max The maximum grid-connected power value during the peak load period; adjusting the dynamic grid-connected coefficient in different load periods allows decision makers to implement different dispatching strategies based on the different economic and environmental benefits of the system; The specific operation is as follows: First, during the load valley and flat periods, the wind and solar power are connected to the grid according to the corresponding grid connection standard value, and the wind and solar power that exceeds the grid connection standard value is stored in the energy storage unit, which can achieve "low storage" while completing the smooth grid connection of wind and solar power; Secondly, during the peak load period, the wind and solar power are connected to the grid according to the set grid connection standard value, and the energy storage unit discharges to further reduce the load peak, which can achieve "high generation" while completing the smooth grid connection of wind and solar power. In this strategy, the stability of the power of the power grid can be ensured in different load periods. At the same time, by shifting the power stored in the load valley period and the flat period to the load peak period for release, the dynamic economic benefits of "low storage and high generation" can be achieved. (3) Establish an improved thermal power test system including photovoltaic power stations and wind farms, and set the test system parameters Based on the IEEE-10 unit 39-node thermal power system, the present invention establishes an improved test system including a photovoltaic power station and two wind farms; According to the load demand at different dispatching times, the load at different dispatching times is divided into three parts: low valley area, flat valley area and peak area; at the same time, the electricity price at different dispatching times needs to be obtained by referring to the local electricity price; (4) Design the GWOH algorithm, initialize its parameters, such as population size, number of iterations, etc., and set the objective function F obj Defined as the fitness function of the algorithm Fit Set the maximum number of iterations of the algorithm Max_iter, the population size of the algorithm Pop, and the wolf pack recombination coefficient p; divide the wolves into α, β, δ, and ω according to the dominance of different wolves in the wolf pack; α wolf has the strongest dominance and the best fitness value, and is usually regarded as the optimal solution in the population; Use the GWOH algorithm to solve the HDEED problem, obtain the Pareto frontier solution of the problem, and set the objective function F of HDEED obj Defined as the algorithm's fitness function Fit: Fit=F obj (5) The GWOH algorithm is used to obtain the Pareto optimal frontier of the HDEED model, and the fuzzy theory is used to obtain a compromise solution. The optimal dynamic grid connection strategy is used to obtain the best scheduling solution for system economy and environmental protection under the premise of smoothing the volatility of renewable energy generation. The specific steps of the GWOH algorithm are: 1) Initialize the GWOH algorithm parameters, determine the maximum number of iterations Max_iter of the wolf pack and the population size Pop of the algorithm; divide the wolves into α, β, δ and ω according to the fitness values ​​of different wolves in the wolf pack; the α wolf has the largest fitness value and is usually regarded as the optimal solution in the population; The dominance of β, δ, and ω wolves decreases in that order; 2) Set the wolf pack recombination coefficient p (p = 0.6); If the random number r < p, the individual positions in the wolf pack are updated according to the following formula, and the individual fitness values ​​of the wolves are calculated to update α, β, δ and ω: In the formula is the average position of the wolf pack; F i is a random number in the interval [0, 2]; If the random number r>p, calculate the linear attenuation coefficient k * ;k * is the nonlinear decreasing coefficient, which decreases nonlinearly with the increase of the number of iterations; During the hunting process, the distance between individual wolves and α, β, and δ wolves is: In the formula, is the individual position in the wolf pack, represents the distance between the gray wolf and its prey, is the coefficient matrix; The present invention designs a hierarchical order weight distribution strategy, which constructs distribution weights according to the fitness values ​​of α, β, and δ wolves to calculate the positions of other individuals in the population: In the formula, The expression of is obtained by the following formula: 3) Obtain the Pareto frontier solution of the HDEED problem; Repeat process 2) until the maximum number of iterations is reached, and output the position of the wolf pack to obtain the Pareto frontier solution.

6. The HDEED model according to claim 5, characterized in that: Under the optimal dynamic grid-connected coefficient strategy, the system cost obtained by the GWOH algorithm is 4.52% lower than when there is no renewable energy, and the system emissions are reduced by 11.84%; the results show that the optimal dynamic grid-connected coefficient strategy can not only effectively smooth the fluctuations of renewable energy electricity, but also utilize the advantages of low storage and high generation to improve economic benefits.

Images