Source-load-storage coordinated optimization control method for promoting new energy consumption
Through improved time series similarity measurement and variational mode decomposition technology, combined with three-state division method and whale optimization algorithm, a source-load-storage coordination optimization model was established, and the power fluctuation of wind farms and energy storage resources were solved by wind power absorption and wind power consumption. The problem of insufficient configuration has achieved the improvement of wind power consumption capacity and system operation stability.
Patent Information
- Application Number
- CN202410926415.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-11
- Publication Date
- 2024-11-15
AI Technical Summary
The prior art is difficult to effectively curb the power fluctuations of wind farms during wind power consumption, resulting in unstable system operation and insufficient allocation of energy storage resources, affecting the wind power consumption capacity.
Adopting improved time series similarity measurement method and variational mode decomposition technology, combining three-state division method and whale optimization algorithm, establish a source-load-storage coordination optimization model, optimize wind power consumption and system operating costs, and improve system regulation flexibility.
Effectively suppress wind farm power fluctuations, improve system operation stability, optimize wind power consumption capacity, reduce operating costs, and avoid inconsistencies in measurement and conflicts between goals.
Smart Images

Figure CN118971178A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of wind power consumption in power grids, and in particular to a source-load-storage coordinated optimization control method for promoting consumption of new energy. Background Art
[0002] For wind power consumption, the traditional method is generally to establish a comprehensive optimization model, with the goal of minimizing cost or maximizing wind power consumption, but in reality, it is necessary to consider the impact of different wind power outputs at different times on the model itself. Dividing the source-load-storage coordinated operation mode according to the wind power consumption capacity situation can not only ensure that wind power is effectively consumed, but also smooth out wind farm power fluctuations and improve the safety of system operation. Therefore, dividing the operation mode is one of the key technologies to ensure the realization of the above goals, and actively promote related technologies so that they can better promote the efficient integration of source-load-storage systems.
[0003] Source-load-storage interaction is an important technical means for high-proportion renewable energy power systems to cope with insufficient flexibility. Quantitative analysis of the impact of energy storage on source-load-storage interaction to improve the system's regulation capacity is conducive to the lean configuration of energy storage resources. Due to grid constraints and conventional unit operation constraints, increasing the energy storage configuration capacity will lead to a relatively saturated state in improving the system's regulation capacity to cope with wind power output fluctuations, and the improvement of the system's wind power absorption capacity is affected by different capacity ratios of energy storage and flexible loads. Therefore, the rational planning of flexible resources and the quantitative evaluation of their impact on the system's wind power absorption capacity have become key issues in the construction of high-proportion renewable energy power systems. Summary of the invention
[0004] Purpose of the invention: The purpose of the present invention is to provide a source-load-storage coordinated optimization control method to promote the consumption of new energy and solve the problems existing in the background technology.
[0005] Technical solution: The source-load-storage coordinated optimization control method for promoting new energy consumption described in the present invention comprises the following steps:
[0006] (1) The improved time series similarity measurement method is used to obtain the source-load similarity index and curve fluctuation index of the load curve and the new energy output curve. The variational mode decomposition can be used to reconstruct the deviation sequence between the expected curve of the new energy and load and the original curve into a high-frequency power curve and a low-frequency power curve, which are used as the charging and discharging power constraints of the power battery and the energy battery in the optimization model respectively;
[0007] (2) Simulating the different operating states of the power system under different operating conditions, the three-state partitioning method is used to obtain the power system source-load-storage coordinated optimization dispatching model based on the three-state partitioning method;
[0008] (3) The source-load similarity index and curve fluctuation index are used to establish a source-load coordination multi-objective optimization model with the goal of maximizing wind power consumption and minimizing system operation cost, and the output of conventional power sources and the switching of high-energy loads are optimized;
[0009] (4) The multi-objective optimization model is transformed into a single-objective model by using the linear weighted method, ideal point and anti-ideal point method. The principle of swarm intelligence optimization is utilized to improve the whale optimization algorithm. The adaptive convergence factor strategy is combined to accelerate the convergence of the algorithm and solve the model.
[0010] Furthermore, step (1) includes the following steps:
[0011] (11) The numerical similarity of the source-load time series is measured by the improved Euclidean distance, and the morphological similarity of the source-load time series is measured by the improved DTW distance;
[0012] Suppose there are two time series X=(x1,x1,…,x n ) and Y=(y1,y1,…,y n ), the improved Euclidean distance expression is as follows:
[0013] Improved Euclidean distance expression:
[0014]
[0015] ΔXY=max(min(XY),min(YX)) (2)
[0016] Where: let σ = D'(X, Y) be the numerical similarity; D'(X, Y) be the Euclidean distance between time series X and Y; T be the length of the time series.
[0017] Construct the distance matrix D of n×m matrix n×m , where each element in the matrix has the following value:
[0018]
[0019] Formula (3) is point x i and i The improved Euclidean distance between them; ΔXY is the minimum point-to-point distance between two time series; the DTW method is used to find the optimal curved path P = (P1, P2, ..., P k ), so that the cumulative distance between sequences S and Q, that is, formula (4), is minimized. The formula is as follows:
[0020]
[0021] Where: P sIt represents the position of the path element in the distance matrix and usually needs to satisfy the boundary, continuity, and monotonicity.
[0022] In order to solve equation (4), we use the dynamic programming method to construct the cumulative cost matrix L, namely:
[0023]
[0024] Where: i = 1, 2, ..., n; j = 1, 2, ..., m; L (0, 0) = 0; L (i, 0) = L (0, j) = + ∞; the dynamic time warping distance of time series X and Y is DTW (X, Y) = L (n, m); let λ = DTW (X, Y) be the morphological similarity index of the two time series;
[0025] (12) Assume that the time series volatility index γ is as follows:
[0026]
[0027] Where: x i is the value of the time series X at time i; is the average value of the time series;
[0028] Assuming that the power change within 10 minutes should not exceed 1 / 3 of its corresponding installed capacity, the time series volatility index is as follows:
[0029]
[0030] Where: x i is the value of the time series X at time i; is the average value of the time series; T is the length of the time series;
[0031] (13) Based on variational mode decomposition, cloud energy storage power allocation is performed and price-based demand response is modeled:
[0032]
[0033] Where: ε ii , ε ij are the self-elasticity coefficient and the cross-elasticity coefficient respectively; i and j are the i-th and j-th time periods respectively; ΔP and Δρ represent the increase in electricity demand and electricity price respectively.
[0034] Furthermore, step (2) includes the following steps:
[0035] (21) Based on the amount of wind curtailment and its changing trend, the power system including thermal storage, electric energy storage equipment and flexible loads is divided into normal state, warning state and emergency state;
[0036] (22) Establish a coordinated optimization dispatch model of power system sources, loads and storage under three conditions.
[0037] Furthermore, the principles for dividing the system status are as follows: the power generation side and the load side jointly determine the system operating status, the system status is defined based on the amount of wind abandonment and its changing trend, the amount of wind abandonment corresponds to the wind power absorption margin, the two are opposite numbers, and the overlapping principles of economy, safety and priority absorption of wind power are considered. Whether economic dispatch should be adopted is determined based on whether the system retains a certain absorption margin for wind power.
[0038] Furthermore, step (3) includes the following steps:
[0039] (31) Construct an objective function that maximizes wind power consumption and minimizes operating costs based on system power balance constraints, conventional power supply operation constraints, and high-load load switching constraints;
[0040] (32) A multi-objective optimization model for source-load coordination is established to obtain the Pareto optimal solution set.
[0041] Furthermore, the improved whale optimization algorithm solution process is:
[0042] (S1) Initialize parameters and randomly generate initial individuals;
[0043] (S2) Calculate individual fitness and record the best individual;
[0044] (S3) Determine whether the maximum number of iterations is met; if so, output the optimal solution; if not, continue the loop.
[0045] (S4) Adjust the movement direction and distance of individual whales; update the adaptive convergence factor, the formula is as follows:
[0046] a=2-(D / D max )
[0047] Where: D max is the maximum possible value of diversity;
[0048] Update the group diversity metric, the formula is as follows:
[0049]
[0050] Where: N is the population size; X i is the i-th individual in the population; X is the average position of the population;
[0051] (S5) Update the coefficient matrix and distance vector matrix to control the extent to which the individual whale approaches or moves away from the optimal solution or the global optimal solution. The formula is as follows:
[0052]
[0053] Where: is the position of individual j at the kth iteration, i.e., the position of the current solution; w is the weight parameter, which determines the influence of the Gaussian variation term on the position update; A is the random coefficient matrix, which is used to adjust the amplitude of the whale individual approaching or moving away from the optimal solution or the global optimal solution; D k is a distance vector matrix, which indicates the distance difference between the current solution and the optimal solution or the global optimal solution; pN(0,1) is a Gaussian variation term, which is used to help the algorithm escape from the local optimum, increase the diversity of the search space, and speed up the convergence of the algorithm;
[0054] (S6) Generate a random number p. If p<0.5, select the spiral mechanism to update the individual position. If p≥0.5, |A|≥1, select the search and foraging mechanism to update the individual position. If p≥0.5, |A|<1, select the shrinking and surrounding mechanism to update the individual position.
[0055] (S7) Perform Gaussian mutation on each individual whale and add a random number to it with a certain probability to increase its randomness and diversity: the formula is as follows:
[0056]
[0057] (S8) Calculate the new objective function value of each individual whale after Gaussian mutation, and compare it with its original objective function value; if the new objective function value is better, update the objective function value and position; otherwise, keep it unchanged;
[0058] (S9) Find the optimal solution and the global optimal solution among all the current whale individuals, and compare them with the optimal solution and the global optimal solution of the previous iteration; if the current optimal solution and the global optimal solution are better, update the optimal solution and the global optimal solution; otherwise, keep them unchanged;
[0059] (S10) Return to step (S2) and repeat the iterative process until the termination condition is reached.
[0060] The source-load-storage coordinated optimization control system for promoting the consumption of new energy according to the present invention comprises:
[0061] Curve module: It is used to obtain the source-load similarity index and curve fluctuation index of the load curve and the new energy output curve by using the improved time series similarity measurement method. The deviation sequence of the expected curve of the new energy and load and the original curve can be reconstructed into high-frequency power curve and low-frequency power curve by using variational mode decomposition, which are used as the charging and discharging power constraints of the power battery and energy battery in the optimization model respectively;
[0062] Optimization and dispatching module: used to simulate the different operating states of the power system under different operating conditions, and obtain the coordinated optimization dispatching model of the power system source, load and storage based on the three-state partitioning method by using the three-state partitioning method;
[0063] Multi-objective optimization module: used to establish a multi-objective optimization model for source-load coordination based on source-load similarity index and curve fluctuation index, with the goal of maximizing wind power consumption and minimizing system operation cost, and to optimize the output of conventional power sources and the switching of high-energy loads;
[0064] Solution module: It is used to transform the multi-objective optimization model into a single objective by using the linear weighted method, ideal point and anti-ideal point method. It uses the principle of swarm intelligence optimization to improve the whale optimization algorithm, and combines the adaptive convergence factor strategy to accelerate the algorithm convergence method to solve the model.
[0065] An electronic device described in the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, any one of the source-load-storage coordinated optimization control methods for promoting new energy consumption is implemented.
[0066] The present invention provides a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any one of the source-load-storage coordinated optimization control methods for promoting new energy consumption.
[0067] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: by improving the time series similarity measurement method to obtain the source-load similarity index, and taking the average fluctuation degree of each adjacent time period of the curve as the curve fluctuation index, the load curve can be more effectively made to follow the change of the new energy curve, so that the output of the traditional unit tends to be stable, thereby reducing the spare capacity, frequent start and stop and climbing of the traditional unit; imitating the different operating states of the power system under different operating conditions, a power system source-load-storage coordinated optimization scheduling model based on the three-state division method is established, which can improve the system adjustment flexibility and reduce the system operation cost; with the maximum wind power consumption and the minimum system operation cost as the goal, a source-load coordination multi-objective optimization model is established, which takes into account the impact of source-load coordination on wind power consumption and economic operation, and avoids the problems of inconsistent measurement and mutual conflict between objectives; the improved whale optimization algorithm increases the population diversity by introducing the Gaussian mutation operator, avoids the algorithm from falling into the local optimum, and accelerates the convergence speed of the algorithm; combined with the adaptive convergence factor strategy, while ensuring the global search capability of the whale optimization algorithm, the local search performance of the algorithm is improved, and the optimization ability of the algorithm is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0068] Figure 1 is a flow chart of the present invention;
[0069] Figure 2 It is a variational mode decomposition flow chart of the present invention;
[0070] Figure 3 This is a flow chart of solving the improved whale optimization algorithm of the present invention. DETAILED DESCRIPTION
[0071] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.
[0072] like Figure 1 As shown, an embodiment of the present invention provides a source-load-storage coordinated optimization control method for promoting new energy consumption, comprising the following steps:
[0073] Step 1: By improving the Euclidean distance, the source-load similarity index of the load curve and the new energy output curve is obtained, and the curve volatility index is measured by the improved DTW distance; the deviation sequence of the expected curve of the new energy and load and the original curve can be reconstructed into a high-frequency power curve and a low-frequency power curve by using variational mode decomposition, which are used as the charging and discharging power constraints of the power battery and the energy battery in the optimization model respectively; including the following sub-steps:
[0074] Step 1.1, measure the numerical similarity of the source-load time series σ = D'(X, Y) by the improved Euclidean distance, and measure the morphological similarity of the source-load time series λ = DTW(X, Y) by the improved DTW distance;
[0075] Given two time series X = (x1, x1, ..., x n ) and Y=(y1,y1,…,y n ), then the traditional Euclidean distance and the improved Euclidean distance are expressed as follows:
[0076] Traditional Euclidean distance expression:
[0077]
[0078] Improved Euclidean distance expression:
[0079]
[0080] ΔXY=max(min(XY),min(YX)) (3)
[0081] Where: Let σ = D'(X, Y) be the numerical similarity; D(X, Y) be the traditional Euclidean distance between time series X and Y; T be the length of the time series.
[0082] Construct the distance matrix D of n×m matrix n×m , where each element in the matrix has the following value:
[0083]
[0084] Formula (4) is point x i and i The improved Euclidean distance between them. ΔXY is the same as formula (3), which is the minimum point-to-point distance between two time series. The purpose of DTW is to find an optimal curved path P = (P1, P2, ..., P from point P1 = (1, 1) to point P2 = (n, m). k ), so that the cumulative distance between sequences S and Q, i.e., formula (5), is minimized.
[0085]
[0086] Where: P s It represents the position of the path element in the distance matrix and usually needs to satisfy the boundary, continuity, and monotonicity.
[0087] In order to solve equation (5), a cumulative cost matrix L is constructed using the dynamic programming method, namely:
[0088]
[0089] Where: i = 1, 2, …, n; j = 1, 2, …, m; L(0, 0) = 0; L(i, 0) = L(0, j) = +∞; the dynamic time warping distance of time series X and Y is DTW(X, Y) = L(n, m); let λ = DTW(X, Y) be the morphological similarity index of the two time series.
[0090] Step 1.2: Taking 30-150MW wind power station as an example, the time series volatility index γ is proposed:
[0091]
[0092] Where: x i is the value of the time series X at time i; X T is the average value of the time series;
[0093] Taking a 30-150MW wind power station as an example, it is stipulated that the power change within 10 minutes should not exceed 1 / 3 of its corresponding installed capacity. In order to characterize the fluctuation degree of the new energy curve and the load curve, the time series fluctuation index proposed in this patent is as follows:
[0094]
[0095] Where: x i is the value of the time series X at time i; is the average value of the time series; T is the length of the time series;
[0096] Step 1.3: Based on variational mode decomposition, cloud energy storage power allocation is performed and price-based demand response is modeled:
[0097]
[0098] Where: ε ii , ε ij are the self-elasticity coefficient and the cross-elasticity coefficient respectively; i and j are the i-th and j-th time periods respectively, ΔP and Δρ represent the increase in electricity demand and electricity price respectively;
[0099] Signal decomposition methods include empirical mode decomposition (EMD) and variational mode decomposition (VMD). Since EMD has the disadvantage of mode aliasing and VMD has the characteristics of good robustness and high computational efficiency, VMD is gradually applied to frequency signal separation and reconstruction. VMD is a new signal analysis method that is completely non-recursive and quasi-orthogonal. In essence, it is the construction and solution process of the variational problem, and its purpose is to decompose the complex signal f(t) into k eigenmode functions u with center frequencies from low to high. k (t), referred to as submodal function. The construction and solution process of variational problem is as follows Figure 2 shown.
[0100] By using the above method, f(t) can be decomposed into k sub-modes corresponding to different center frequencies [u1(t),u2(t),…,u k (t)].
[0101] Mutual information entropy (MI) is used to reflect the correlation between two random variables. When the correlation between two random variables is greater, the mutual information entropy is also greater. For each sub-mode, the expression of its information entropy is as follows:
[0102]
[0103] In the formula, P(u k ) is the energy value of the kth submode.
[0104] Then the mutual information entropy of the two sub-modes can be expressed as:
[0105] MI(u k ,u k+1 )=H(u k )+H(u k+1 )-H(u k ,u k+1 ) (9)
[0106] Normalize the mutual information entropy:
[0107]
[0108] After the difference sequence of load curve or new energy output curve is decomposed by VMD, the mutual information entropy of its adjacent sub-modes will first decrease and then increase. The minimum point of mutual information entropy is selected as the dividing point between high-frequency component and low-frequency component, and the high-frequency component and low-frequency component are reconstructed to obtain the high-frequency component and low-frequency component of the load power difference sequence and the new energy power difference sequence respectively. The high-frequency component and low-frequency component are used as the charging and discharging power constraints for optimizing medium-power batteries and energy batteries, respectively.
[0109] The modeling of price-based demand response usually uses the demand price elasticity matrix to describe the impact of changes in electricity prices on users. Price demand response resources refer to flexible loads that can respond to changes in electricity prices. The expression of its demand elasticity coefficient is as follows:
[0110]
[0111] Where: ΔP and Δp represent the increase in electricity demand and electricity price respectively; P0 and p0 represent the original electricity demand and electricity price respectively.
[0112] The change of electricity price not only affects the load in this period, but also affects the load in other periods. Therefore, the price elasticity coefficient is divided into self-elasticity coefficient and cross-elasticity coefficient. Its expression is as follows:
[0113]
[0114] In the formula, ε ii , ε ij are the self-elasticity coefficient and the cross-elasticity coefficient respectively; i and j are the i-th and j-th time periods respectively; ΔP and Δρ represent the increase in electricity demand and electricity price respectively.
[0115] If the dispatch cycle is defined as 24 time periods, the mathematical model of price demand response for one day is as follows:
[0116]
[0117] The price elasticity coefficient matrix ε can be obtained by actually investigating and analyzing the impact of electricity prices on electricity demand. Therefore, setting electricity prices reasonably can change the electricity demand in different time periods.
[0118] Step 2: According to the amount of wind power abandoned and its changing trend, the cogeneration system including heat storage, electric energy storage and flexible load is divided into normal state, warning state and emergency state, a three-state division method for solving wind power consumption is proposed, and a coordinated optimization dispatch model of source, load and storage of the power system under different states is established;
[0119] Step 2.1, according to the amount of wind abandonment and its changing trend, the power system including thermal storage, electric energy storage equipment and flexible load is divided into normal state, warning state and emergency state;
[0120] Based on the fact that the power system has different operating states of systems and equipment under different operating conditions, a three-state classification method for solving wind power absorption is proposed. According to the amount of abandoned wind power and its changing trend, the power system including heat storage, electric energy storage equipment and flexible loads is divided into three states.
[0121] The principles for dividing the system status are as follows: the power generation side and the load side jointly determine the system operating status; the system status is defined based on the amount of wind abandonment and its changing trend; the amount of wind abandonment corresponds to the wind power absorption margin, and the two are opposite numbers; considering the overlapping principles of economy, safety and priority absorption of wind power, it is determined whether to adopt economic dispatch based on whether the system retains a certain absorption margin for wind power.
[0122] Under normal conditions, all wind power is absorbed, and the following conditions are met:
[0123]
[0124] Where: and are the minimum and maximum outputs of thermal power units during period t respectively; is the wind power in period t; is the output value of the CHP unit during period t; is the predicted value of electric load during period t; N c is the heat-to-electricity ratio of the CHP unit; is the heat load value during period t; is the heat release power of the electric boiler during period t, under normal conditions I is the heat release power of the heat storage device during period t; h is the number of thermal power units; is the active power output of thermal power unit i during period t; P h,i,min and P h,i,max are the minimum and maximum active output of thermal power unit i respectively; r i,d and r i,u are the downward and upward climbing rates of thermal power unit i respectively; T 15 It is the duration of the scheduling period (the value is 15 minutes).
[0125] According to the principle of system status classification, the wind power consumption margin is:
[0126]
[0127] The normal state is described in mathematical language as follows: there is a constant C n , so that It is always valid at any time period, and when Equal to C n When Constant is established. Among them, C n It is the corresponding wind power consumption margin when the system operating point is at the normal state boundary.
[0128] After the flexible load is transferred, the new load power expression is:
[0129]
[0130] Where: is the transfer load in period t; is the outgoing load in period t. In period t, it is not possible to transfer in and out at the same time. 0, when transferring out is 0.
[0131] Assuming that the working cycle of all flexible loads in the system is exactly equal to the duration of a scheduling period, we have:
[0132]
[0133] Where: M s is the number of flexible load types; x m,t”,t is the number of loads of the mth type transferred out / into the t″ period; x m,t,t’ is the number of loads of the mth type transferred out of period t / into period t′; P m is the mth type load power; T n A set of scheduling periods in normal state.
[0134] Transferring part of the flexible load to the off-peak period improves the economy of system operation, but it cannot affect wind power consumption, so there is the following inequality.
[0135]
[0136]
[0137] When the system operating point is at the common boundary of the normal state and the warning state, the wind power consumption margin tends to decrease. It is always valid at any time period, and when hour, Heng established.
[0138] When the system operating point is located at the common boundary of the alert state and the emergency state, it has a tendency to move toward the alert state area. hour, Heng established.
[0139] After the energy storage device is connected, considering that the response time of the energy storage device to smooth the fluctuation of wind power is in seconds, there is an inequality:
[0140]
[0141] Where: and They are the minimum and maximum output values of the thermal power unit at time k in period t; is the wind power at time k in period t; is the CHP unit output at time k in period t; is the predicted value of electric load at time k in period t; is the output power of the energy storage device at time k in period t, A positive value indicates that the device is discharging. A negative value indicates that the device is charging.
[0142] Since load forecasting, thermal power unit scheduling and CHP unit output are based on time periods, this patent makes the following processing when analyzing the problem.
[0143]
[0144] Where: k = 0, 1, ..., 60T 15 .
[0145] From equations (26) and (27), we can get the upper and lower limits of the output power of the energy storage device allowed by the system, that is,
[0146]
[0147] In an emergency, before the thermal storage electric boiler is put into use, the wind power consumption margin of the system should satisfy the inequality And when hour, Heng established.
[0148] After the large-capacity thermal storage electric boiler is put into use, the relationship between the conventional thermal power units, wind power, CHP units, electric load and thermal storage electric boiler in the system should meet the following requirements:
[0149]
[0150] Where: is the power of the thermal storage electric boiler during period t.
[0151] Step 2.2: Establish a coordinated optimization dispatch model of power system sources, loads and storage under three conditions.
[0152] (A) System optimization model under normal conditions
[0153] Operating costt Including the power generation cost of CHP units and thermal power units and the dispatching cost of flexible loads.
[0154] minC t =C c +C h +C d (32)
[0155] Where: C c is the fuel cost of the CHP unit; C h is the power generation cost of thermal power units; C d,s is the flexible load dispatching cost.
[0156]
[0157] Where: a i , b i , c i , d i , e i and f i is the coal consumption coefficient of the i-th CHP unit; is the output of the i-th CHP unit in period t; I c is the number of all CHP units; is the heating power of the i-th CHP unit in period t, which satisfies
[0158]
[0159] Where: α i , β i and γ i is the operating cost parameter of thermal power unit i; Indicates the operating status of the thermal power unit, its value is equal to 0 or 1; τ i is the downtime of thermal power unit i; δ 0i , δ 1i and τ 0i is the start-up and shutdown cost parameter of thermal power unit i.
[0160] The outflow of flexible load in a certain period is equal to the inflow in another period. If only the outflow cost is considered, the load dispatching cost is:
[0161]
[0162] Where: μ m It is the unit compensation coefficient of the mth type flexible load.
[0163] Ignoring the power balance constraint of network loss, the system constraints are
[0164]
[0165] Spinning reserve constraint,This patent uses positive and negative reserve capacity to overcome wind power forecast errors,,i.e.
[0166]
[0167] Where: T3 is the spinning reserve response time, which is 3 minutes; L is the percentage of load forecast error to positive spinning reserve demand; f u and f d are the percentages of wind power output forecast error on the demand for positive and negative spinning reserves; P f,max It is the maximum output value of wind power.
[0168] The constraints of thermal power units are:
[0169] 1) Output power constraints:
[0170]
[0171] 2) Climbing speed constraint:
[0172]
[0173] The flexible load constraints are:
[0174]
[0175] Where: x m,t is the number of flexible loads in the period t before transfer; x m,t,t Indicates the number of flexible loads remaining in period t after transfer.
[0176] Energy storage device constraints are:
[0177] 1) Heat release power constraint:
[0178]
[0179] Where: H out,max is the output power of the heat storage device during period t The maximum value of .
[0180] 2) Ignore heat loss, thermal storage device state constraints:
[0181]
[0182] Where: is the heat stored in the heat storage device during the period t+1.
[0183] (B) System optimization model in alert state
[0184] In the alert state, the safety of operation is improved by adjusting the electric energy storage equipment in the system. Under the premise of not abandoning wind power, the switching of electric energy storage equipment is optimized with the goal of smoothing wind power fluctuations and minimizing energy storage output. Since the constraint condition can include the fluctuation smoothing interval, the objective function is:
[0185]
[0186] Where: Z = 9 is the number of optimized sequences; T a Represents a collection of alert status time periods.
[0187] Ignoring network losses, the power balance constraints are:
[0188]
[0189] The output power constraints of electric energy storage devices are:
[0190]
[0191] Where: is the maximum charging power of the energy storage device at time k in period t; S is the maximum discharge power of the energy storage device at time k in period t; b,max and S b,min They are the upper and lower limits of the state of charge of the electric energy storage device respectively; C is the charge state of the energy storage device at time k-1 in period t; b is the rated capacity of the electric energy storage device; ΔT b is the charging and discharging time of the energy storage device; P ch,max and P dh,max They are the maximum values of charging and discharging power of the energy storage device respectively.
[0192] 3) The capacity constraints of electric energy storage equipment are:
[0193]
[0194]
[0195] In the formula, It is the energy of the electric energy storage device at time k in period t.
[0196] 4) Constraint interval for smoothing composite power fluctuations:
[0197] The synthesized output power meets the power fluctuation suppression index within 1 minute. To improve the convergence speed of the program, the output power at time k = 0 in period t The upper and lower limits of the allowed power fluctuation smoothing constraints are:
[0198]
[0199] Where: max and min are the maximum and minimum values of the synthetic power fluctuation range within 59 seconds before the current moment; γ1 is the initial given value; P f,rated is the total installed capacity of the wind farm; f1 is the ratio of the synthetic power fluctuation range within 59 seconds before the current moment to the installed capacity of the wind farm.
[0200] Therefore, the composite output power at time k in period t should satisfy:
[0201]
[0202] (C) System optimization model in emergency situations
[0203] After adding the thermal storage electric boiler, from the perspective of comprehensive electricity and heat, with the goal of maximizing wind power consumption, the optimal scheduling model is established as follows:
[0204]
[0205] Where, T e A collection of dispatch time periods for emergency situations.
[0206] The system power balance constraints are:
[0207]
[0208] The constraints of the thermal storage electric boiler model are:
[0209] 1) Electric heat conversion constraints:
[0210]
[0211] Where: N r is the electric-heat conversion coefficient of the electric boiler, which takes a value of 1; is the input heat of the thermal storage device.
[0212] 2) Electricity Constraints:
[0213]
[0214] 3) Thermal storage device capacity constraints:
[0215]
[0216] Where: is the heat stored in the heat storage device during period t; C s,max is the maximum heat storage capacity.
[0217] 4) Thermal storage power constraints of thermal storage devices:
[0218]
[0219] Where: H in,max is the maximum input power of the thermal storage device.
[0220] 5) Ignore heat loss, thermal storage device state constraints:
[0221]
[0222] Step 3: Based on step 2, combined with the source-load similarity index and the curve fluctuation index, the influence of source-load coordination on wind power consumption and system operation is comprehensively considered, and a multi-objective optimization model of source-load coordination is established with the goal of maximizing wind power consumption and minimizing system operation cost. At the same time, the output of conventional power sources and the switching of high-energy loads are optimized;
[0223] Step 3.1, construct an objective function with the maximum wind power consumption and the minimum operation cost, which is based on the constraints of system power balance, conventional power supply operation and high-load switching;
[0224] A multi-objective optimization model is established with the goals of maximizing wind power consumption and minimizing system operating costs, and the output of conventional power sources and the switching of high-energy loads are optimized.
[0225] 1) Wind power consumption is the largest:
[0226] Wind power consumption refers to the active power of wind power dispatched by the power grid under the source-load regulation capacity.
[0227]
[0228] Where: T is the number of time periods during the scheduling period; N W is the number of wind farms; is the active dispatch output of wind farm i in period t.
[0229] 2) Minimum operating cost:
[0230] minC GH =C gen +C high-load (64)
[0231]
[0232] Where: C gen is the power generation cost of conventional units, which consists of the operating cost and start-up and shutdown cost of conventional units; C high-load is the switching cost of high energy load; N G is the number of conventional generator sets; is the start-stop state variable of conventional unit j in period t, Indicates that the conventional unit is in shutdown state. Indicates that the conventional unit is in the on state; is the active power output of conventional unit j in period t; α j , β j , γ j is the operating cost parameter of conventional unit j; η 0j , η 1j , τ j is the start-up and shutdown cost parameter of conventional unit j; τ is the downtime of conventional unit; N H is the number of high energy load groups; Hk is the unit regulation cost of high energy load k; is the switching state of high energy load k in time period t, Indicates that the high energy load k stops running during period t. Indicates that high energy load k is put into operation in time period t; P Hk It is the unit switching capacity of high energy load k.
[0233] The constraints include system power balance constraints, conventional power supply operation constraints and high-energy load switching constraints.
[0234] 1) System constraints
[0235] ① Power balance constraints:
[0236]
[0237] In the formula is the original active load of the system during period t.
[0238] ② Rotating standby constraints.
[0239] The randomness of wind power leads to certain errors in wind power forecasting. In order to avoid the adverse impact of wind power output forecasting errors on the optimal operation of the system, this patent responds to the fluctuations of large-scale wind power output by increasing positive and negative reserve capacity.
[0240]
[0241] Where: and are the maximum available output and minimum available output of conventional unit j in period t respectively; and are the positive and negative spinning reserves required to cope with the load forecast error during period t, respectively; and They are respectively the positive and negative spinning reserves required to cope with wind power fluctuations during period t.
[0242] 2) Wind power output constraints
[0243]
[0244] In the formula is the predicted active power output of wind farm i in period t.
[0245] 3) Conventional power supply operation constraints
[0246] ① Output power upper and lower limits:
[0247]
[0248] Where P Gj,max and P Gj,min are the upper and lower limits of the output power of conventional unit j respectively.
[0249] ②Minimum start and stop time constraints:
[0250]
[0251] Where: and are respectively the startup duration and shutdown duration of conventional unit j in time period t; and are the minimum operating time and minimum downtime of conventional unit j respectively.
[0252] ③Climbing speed constraint:
[0253]
[0254] Where: is the output of conventional unit j in period t-1; P Gj,up and P Gj,down They are the rising output limit and falling output limit of conventional unit j respectively.
[0255] 4) High-energy load switching constraints
[0256] ① Input capacity constraints:
[0257]
[0258] In the formula and They are the upper and lower limits of the input capacity of the high energy load in period t respectively.
[0259] ② Switching frequency constraints:
[0260]
[0261] Where M HkThe maximum allowable switching times of high energy load k.
[0262] ③ Switching time constraints:
[0263]
[0264] Where: and are respectively the input duration and interruption duration of high energy load k in time period t; and are the minimum continuous input time and minimum continuous interruption time of high energy load k respectively.
[0265] Step 3.2: Establish a multi-objective optimization model for source-load coordination and obtain the Pareto optimal solution set.
[0266] According to step 3.1, the source-load coordination multi-objective optimization model can be described as:
[0267]
[0268] Where: f = (-E W ,C GH ) is the objective function; x is the decision vector consisting of optimization variables; h j (x) is the equality constraint function; g k (x) is the inequality constraint function.
[0269] In multi-objective optimization problems, the objective functions conflict with each other, and it is impossible to optimize multiple objectives at the same time. Therefore, the optimal solution to a multi-objective optimization problem is a set of solutions where the value of any objective function cannot be further optimized without degrading other objective functions, that is, the Pareto optimal solution set.
[0270] Step 4: The multi-objective optimization model is transformed into a single objective using the linear weighted method, ideal point and anti-ideal point method. The principle of swarm intelligence optimization is used to improve the whale optimization algorithm. The adaptive convergence factor strategy is combined to accelerate the convergence of the algorithm and solve the solution.
[0271] Step 4.1, use the linear weighting method, ideal point and anti-ideal point method to transform the multi-objective optimization model into a single objective;
[0272] Multi-objective optimization problems are difficult to solve directly. The linear weighted method is a simple method that adds multi-objective functions into a single objective function according to weight coefficients. It can reduce the complexity of solution and improve efficiency. In order to ensure the rationality of the weights, considering the preference or importance of each objective by the user or decision maker, the proportion of the optimal value of each single objective function in the sum of the optimal values of all single objective functions is used as the weight coefficient. The weights are as follows:
[0273]
[0274] Where: w i is the weight of the i-th objective function; F i,min(t) is the minimum value of the i-th objective function.
[0275] In multi-objective optimization, in order to avoid using inaccurate or unknown extreme values for normalization, this patent adopts a normalization method based on the ideal point method and the anti-ideal point method. For an ideal point, each element represents the optimal value of each objective function in the global range. For the minimized objective function, the ideal point is its global minimum value; for the maximized objective function, the ideal point is its global maximum value.
[0276]
[0277] Where: Z * For the ideal point.
[0278] The anti-ideal point method uses each element to represent the global worst value of the objective function. For the objective function to be minimized, the anti-ideal point is its global maximum value; for the objective function to be maximized, the anti-ideal point is its global minimum value.
[0279]
[0280] Where: Z – The anti-ideal point.
[0281] Finally, the ideal point and the anti-ideal point are used to obtain the normalized value.
[0282]
[0283] Where: F i norm (t) is the normalized value of the i-th objective function; and are the worst value and optimal value of the i-th objective function in the global range, respectively.
[0284] Multiply each normalized objective function by the corresponding weight coefficient and sum them up to get a total unified objective function. The single objective function is:
[0285]
[0286] Where: F Z (t) is the simplified objective function after the multi-objective function is converted.
[0287] Step 4.2: Introduce the Gaussian mutation operator and combine the adaptive convergence factor with the improvement strategy of the population diversity metric to improve the whale optimization algorithm and solve the model.
[0288] Although the traditional whale optimization algorithm has the advantages of fast convergence speed, less parameter control, and strong optimization ability, it also has problems such as easy to fall into local optimality and imbalance between global search ability and local development ability.
[0289] In response to the above problems, this patent has made some optimizations and improvements to the algorithm to improve its optimization effect on multi-objective functions under practical conditions.
[0290] 1) In order to overcome the problem that the whale algorithm is easily trapped in the local optimal solution, the Gaussian mutation operator is introduced and improved on the basis of the traditional whale algorithm. In each iteration process, the whale individuals are Gaussian mutated with a certain probability, so that the whale individuals can jump out of the local optimal solution and increase the diversity of the search space.
[0291] The standard Gaussian distribution function is shown in formula (89):
[0292]
[0293] The Gaussian function is a probability distribution with bell-shaped curve characteristics. Using this function to generate normally distributed random numbers can give whale individuals a certain degree of randomness and diversity. In addition, the peak of the function near the mean is at a higher level, and the transition from the peak to both sides is relatively gentle. This means that whale individuals have a greater probability of conducting local searches when they are close to the optimal solution, and a greater probability of conducting global searches when they are far away from the optimal solution. After incorporating the Gaussian mutation operator, the position update formula of the whale search can be expressed as:
[0294]
[0295] Where: is the position of individual j at the kth iteration, i.e., the position of the current solution; w is the weight parameter, which determines the influence of the Gaussian variation term on the position update; A is the random coefficient matrix, which is used to adjust the amplitude of the whale individual approaching or moving away from the optimal solution or the global optimal solution; D k is a distance vector matrix, which represents the distance difference between the current solution and the optimal solution or the global optimal solution; pN(0,1) is a Gaussian variation term, which is a random part introduced for the current solution to help the algorithm escape from the local optimum, increase the diversity of the search space, and speed up the convergence of the algorithm.
[0296] 2) In terms of the balance between global search and local development capabilities, the whale optimization algorithm often shows shortcomings, causing the algorithm to fall into local optimality. This problem mainly stems from the linear convergence factor inherent in the whale algorithm. The linear decrease of this factor from 2 to 0 fails to fully reflect the dynamic behavior characteristics of the whale population in the search space. In order to solve this problem, this patent proposes an improved strategy that combines an adaptive convergence factor with a population diversity metric.
[0297] Firstly, the population diversity index D is introduced, which aims to evaluate the degree of difference between individuals in the population in real time.
[0298] The diversity index can be calculated using formula (91).
[0299]
[0300] Where: N is the population size; X i is the i-th individual in the population; X is the average position of the population
[0301] A higher diversity index D means that the population is still exploring the solution space extensively, while a lower diversity index D means that the population is searching finely in a local area.
[0302] Then, based on the above diversity index, an adaptive convergence factor a is constructed, which is mathematically expressed as follows:
[0303] a=2-(D / D max ) (92)
[0304] Where: D max is the maximum possible value of diversity.
[0305] In the early stages of the algorithm, when diversity is high, the convergence factor a tends to a wider global search. As iterations progress and diversity decreases, the convergence factor a is gradually adjusted, prompting the population to converge faster and conduct in-depth local searches in the potential solution space, and find a more accurate optimal solution.
[0306] This method of combining adaptive convergence factor and population diversity measurement brings dual advantages to the whale optimization algorithm. On the one hand, the diversity measurement can intuitively reflect the current population status; on the other hand, the adaptive convergence factor ensures that the algorithm can make appropriate strategy adjustments at different search stages. This not only makes up for the shortcomings of the original algorithm in global search and local development capabilities, but also provides it with a more efficient and flexible search strategy.
[0307] The improved whale optimization algorithm solution process is as follows Figure 3 As shown, the specific implementation steps are as follows.
[0308] Step 1) Initialize parameters, randomly generate initial individuals, set the population size to 150, and the number of iterations to 200.
[0309] Step 2) Calculate individual fitness and record the best individual.
[0310] Step 3) Determine whether the maximum number of iterations is met. If so, output the optimal solution; if not, continue the cycle.
[0311] Step 4) Adjust the moving direction and distance of individual whales. Update the adaptive convergence factor according to formula (92), and update the group diversity metric according to formula (91).
[0312] Step 5) Update the coefficient matrix and distance vector matrix according to formula (90) to control the extent to which the individual whale approaches or moves away from the optimal solution or the global optimal solution.
[0313] Step 6) Generate a random number p. If p<0.5, select the spiral mechanism to update the individual position; if p≥0.5, |A|≥1, select the search and foraging mechanism to update the individual position; if p≥0.5, |A|<1, select the shrinking and surrounding mechanism to update the individual position.
[0314] Step 7) Perform Gaussian mutation. Perform Gaussian mutation on each individual whale, and add a random number that obeys formula (89) with a certain probability to increase its randomness and diversity.
[0315] Step 8) Calculate the new objective function value of each whale individual after Gaussian mutation and compare it with its original objective function value. If the new objective function value is better, update the objective function value and position; otherwise, keep it unchanged.
[0316] Step 9) Find the optimal solution and global optimal solution among all whale individuals at present, and compare them with the optimal solution and global optimal solution of the previous iteration. If the current optimal solution and global optimal solution are better, update the optimal solution and global optimal solution; otherwise, keep them unchanged.
[0317] Step 10) Return to step 2) and repeat the iterative process until the termination condition is reached.
[0318] The embodiment of the present invention further provides a source-load-storage coordinated optimization control system for promoting the consumption of new energy, including:
[0319] Curve module: It is used to obtain the source-load similarity index and curve fluctuation index of the load curve and the new energy output curve by using the improved time series similarity measurement method. The deviation sequence of the expected curve of the new energy and load and the original curve can be reconstructed into high-frequency power curve and low-frequency power curve by using variational mode decomposition, which are used as the charging and discharging power constraints of the power battery and energy battery in the optimization model respectively;
[0320] Optimization and dispatching module: used to simulate the different operating states of the power system under different operating conditions, and obtain the coordinated optimization dispatching model of the power system source, load and storage based on the three-state partitioning method by using the three-state partitioning method;
[0321] Multi-objective optimization module: used to establish a multi-objective optimization model for source-load coordination based on source-load similarity index and curve fluctuation index, with the goal of maximizing wind power consumption and minimizing system operation cost, and to optimize the output of conventional power sources and the switching of high-energy loads;
[0322] Solution module: It is used to transform the multi-objective optimization model into a single objective by using the linear weighted method, ideal point and anti-ideal point method. It uses the principle of swarm intelligence optimization to improve the whale optimization algorithm, and combines the adaptive convergence factor strategy to accelerate the algorithm convergence method to solve the model.
[0323] An embodiment of the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, it implements any one of the source-load-storage coordinated optimization control methods for promoting new energy consumption.
[0324] An embodiment of the present invention further provides a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any one of the source-load-storage coordinated optimization control methods for promoting new energy consumption.
Claims
1. A source-load-storage coordinated optimization control method to promote the consumption of new energy. It is characterized in that The following steps are involved: (1) The improved time series similarity measurement method is used to obtain the source-load similarity index and curve fluctuation index of the load curve and the new energy output curve. The variational mode decomposition can be used to reconstruct the deviation sequence between the expected curve of the new energy and load and the original curve into a high-frequency power curve and a low-frequency power curve, which are used as the charging and discharging power constraints of the power battery and the energy battery in the optimization model respectively; (2) Simulating the different operating states of the power system under different operating conditions, the three-state partitioning method is used to obtain the power system source-load-storage coordinated optimization dispatching model based on the three-state partitioning method; (3) The source-load similarity index and curve fluctuation index are used to establish a source-load coordination multi-objective optimization model with the goal of maximizing wind power consumption and minimizing system operation cost, and the output of conventional power sources and the switching of high-energy loads are optimized; (4) The multi-objective optimization model is transformed into a single-objective model by using the linear weighted method, ideal point and anti-ideal point method. The principle of swarm intelligence optimization is utilized to improve the whale optimization algorithm. The adaptive convergence factor strategy is combined to accelerate the convergence of the algorithm and solve the model.
2. A source-load-storage coordinated optimization control method for promoting new energy consumption according to claim 1, It is characterized in that Step (1) comprises the following steps: (11) The numerical similarity of the source-load time series is measured by the improved Euclidean distance, and the morphological similarity of the source-load time series is measured by the improved DTW distance; Suppose there are two time series X=(x 1 ,x 1 ,…,x n ) and Y=(y 1 ,y 1 ,…,y n ), the improved Euclidean distance expression is as follows: Improved Euclidean distance expression: ΔXY=max(min(XY),min(YX)) (2) Where: let σ = D'(X, Y) be the numerical similarity; D'(X, Y) be the Euclidean distance between time series X and Y; T be the length of the time series. Construct the distance matrix D of n×m matrix n×m , where each element in the matrix has the following value: Formula (3) is point x i and i The improved Euclidean distance between them; ΔXY is the minimum point-to-point distance between two time series; DTW method is used to find the distance from point P 1 =(1,1) to point P 2 =(n,m)The optimal curved path P = (P 1 ,P 2 ,…,P k ), so that the cumulative distance between sequences S and Q, that is, formula (4), is minimized. The formula is as follows: Where: P s It represents the position of the path element in the distance matrix and usually needs to satisfy the boundary, continuity, and monotonicity. In order to solve equation (4), we use the dynamic programming method to construct the cumulative cost matrix L, namely: Where: i = 1, 2, ..., n; j = 1, 2, ..., m; L (0, 0) = 0; L (i, 0) = L (0, j) = + ∞; the dynamic time warping distance of time series X and Y is DTW (X, Y) = L (n, m); let λ = DTW (X, Y) be the morphological similarity index of the two time series; (12) Assume that the time series volatility index γ is as follows: Where: x i is the value of the time series X at time i; is the average value of the time series; Assuming that the power change within 10 minutes should not exceed 1 / 3 of its corresponding installed capacity, the time series volatility index is as follows: Where: x i is the value of the time series X at time i; is the average value of the time series; T is the length of the time series; (13) Based on variational mode decomposition, cloud energy storage power allocation is performed and price-based demand response is modeled: Where: ε ii , ε ij are the self-elasticity coefficient and the cross-elasticity coefficient respectively; i and j are the i-th and j-th time periods respectively; ΔP and Δρ represent the increase in electricity demand and electricity price respectively.
3. A source-load-storage coordinated optimization control method for promoting new energy consumption according to claim 1, It is characterized in that Step (2) comprises the following steps: (21) Based on the amount of wind curtailment and its changing trend, the power system including thermal storage, electric energy storage equipment and flexible loads is divided into normal state, warning state and emergency state; (22) Establish a coordinated optimization dispatch model of power system sources, loads and storage under three conditions.
4. A source-load-storage coordinated optimization control method for promoting new energy consumption according to claim 3, It is characterized in that The principles for dividing the system status are as follows: the power generation side and the load side jointly determine the system operating status, the system status is defined based on the amount of wind abandonment and its changing trend, the amount of wind abandonment corresponds to the wind power absorption margin, the two are opposite numbers, and the overlapping principles of economy, safety and priority absorption of wind power are considered. Whether economic dispatch should be adopted is determined based on whether the system retains a certain absorption margin for wind power.
5. A source-load-storage coordinated optimization control method for promoting new energy consumption according to claim 1, It is characterized in that Step (3) comprises the following steps: (31) Construct an objective function that maximizes wind power consumption and minimizes operating costs based on system power balance constraints, conventional power supply operation constraints, and high-load load switching constraints; (32) A multi-objective optimization model for source-load coordination is established to obtain the Pareto optimal solution set.
6. A source-load-storage coordinated optimization control method for promoting new energy consumption according to claim 1, It is characterized in that The improved whale optimization algorithm solution process is: (S1) Initialize parameters and randomly generate initial individuals; (S2) Calculate individual fitness and record the best individual; (S3) Determine whether the maximum number of iterations is met; if so, output the optimal solution; if not, continue the loop. (S4) Adjust the movement direction and distance of individual whales; update the adaptive convergence factor, the formula is as follows: a=2-(D / D max ) Where: D max is the maximum possible value of diversity; Update the group diversity metric, the formula is as follows: Where: N is the population size; X i is the i-th individual in the population; X is the average position of the population; (S5) Update the coefficient matrix and distance vector matrix to control the extent to which the individual whale approaches or moves away from the optimal solution or the global optimal solution. The formula is as follows: Where: is the position of individual j at the kth iteration, i.e., the position of the current solution; w is the weight parameter, which determines the influence of the Gaussian variation term on the position update; A is the random coefficient matrix, which is used to adjust the amplitude of the whale individual approaching or moving away from the optimal solution or the global optimal solution; D k is a distance vector matrix, which indicates the distance difference between the current solution and the optimal solution or the global optimal solution; pN(0,1) is a Gaussian variation term, which is used to help the algorithm escape from the local optimum, increase the diversity of the search space, and speed up the convergence of the algorithm; (S6) Generate a random number p. If p<0.5, select the spiral mechanism to update the individual position. If p≥0.5, |A|≥1, select the search foraging mechanism to update the individual position; If p≥0.5, |A|<1, the shrinking and surrounding mechanism is selected to update the individual position. (S7) Perform Gaussian mutation on each individual whale and add a random number to it with a certain probability to increase its randomness and diversity: the formula is as follows: (S8) Calculate the new objective function value of each individual whale after Gaussian mutation, and compare it with its original objective function value; if the new objective function value is better, update the objective function value and position; otherwise, keep it unchanged; (S9) Find the optimal solution and the global optimal solution among all the current whale individuals, and compare them with the optimal solution and the global optimal solution of the previous iteration; if the current optimal solution and the global optimal solution are better, update the optimal solution and the global optimal solution; otherwise, keep them unchanged; (S10) Return to step (S2) and repeat the iterative process until the termination condition is reached.
7. A source-load-storage coordinated optimization control system to promote the consumption of new energy. It is characterized in that include: Curve module: It is used to obtain the source-load similarity index and curve fluctuation index of the load curve and the new energy output curve by using the improved time series similarity measurement method. The deviation sequence of the expected curve of the new energy and load and the original curve can be reconstructed into high-frequency power curve and low-frequency power curve by using variational mode decomposition, which are used as the charging and discharging power constraints of the power battery and energy battery in the optimization model respectively; Optimization and dispatching module: used to simulate the different operating states of the power system under different operating conditions, and obtain the coordinated optimization dispatching model of the power system source, load and storage based on the three-state partitioning method by using the three-state partitioning method; Multi-objective optimization module: used to establish a multi-objective optimization model for source-load coordination based on source-load similarity index and curve fluctuation index, with the goal of maximizing wind power consumption and minimizing system operation cost, and to optimize the output of conventional power sources and the switching of high-energy loads; Solution module: It is used to transform the multi-objective optimization model into a single objective by using the linear weighted method, ideal point and anti-ideal point method. It uses the principle of swarm intelligence optimization to improve the whale optimization algorithm, and combines the adaptive convergence factor strategy to accelerate the algorithm convergence method to solve the model.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the computer program is loaded into the processor, a source-load-storage coordinated optimization control method for promoting new energy consumption according to any one of claims 1 to 7 is implemented.
9. A storage medium storing a computer program. It is characterized in that When the computer program is executed by the processor, a source-load-storage coordinated optimization control method for promoting new energy consumption is implemented according to any one of claims 1 to 7.
Citation Information
Cited By
Flexible controllable resource control performance quantitative evaluation method
CN119623836A
Energy storage regulation marginal income calculation method and system, electronic equipment and medium
CN119695978A
