Power distribution network energy storage-reactive power optimization configuration method considering demand side response

Abstract

The invention relates to a power distribution network energy storage-reactive power optimization configuration method considering demand side response, and the method comprises the following steps: 1, taking ESS configuration capacity and the positions and capacities of CB and SVC as decision variables, and taking the minimum investment maintenance operation cost of each device as a target to construct an upper layer planning model; 2, constructing a lower-layer optimization model by taking a CB gear and a change moment, SVC reactive power output, charge and discharge power of energy storage, electric quantity and load variation under a DR strategy as decision variables and taking minimum voltage deviation and a demand side response index as targets; 3, adopting an improved multi-target mayfly naiad algorithm; 4, considering the energy storage in the power distribution network, considering the operation economy of the power distribution network, and achieving the double-layer reactive power optimization of the power distribution network by taking the minimum network loss cost, demand side response and voltage deviation as targets; 5, example analysis is carried out; the method has the advantages that coordination optimization configuration is considered, the improved multi-target mayfly naiad algorithm is adopted, and falling into a globally optimal solution is avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network optimization, and particularly relates to a method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response. Background Art

[0002] The penetration rate of renewable energy sources such as photovoltaic and wind power in the distribution network has been continuously increasing. However, the large-scale access of distributed generation (DG) will bring problems such as two-way power flow, voltage over-limit, and uneven power distribution, which will further lead to an increase in the network loss of the distribution network, serious curtailment of wind and light, and a reduction in operation reliability. Installing reactive power compensation devices such as capacitor banks (CB) or static var compensators (SVC) at appropriate positions in the distribution network and controlling and dispatching them can improve the voltage distribution and reduce power losses. Therefore, the research on the reactive power configuration problem of the distribution network has received much attention in recent years. However, the existing technology only configures static fixed-capacity reactive power compensation devices and energy storage in the distribution network, and cannot simultaneously optimize the safety, stability, and economy of the distribution network. Therefore, it is of great significance to study the coordinated optimization of the configuration and operation of the distribution network.

[0003] Furthermore, the location and capacity configuration of energy storage devices (ESS) and reactive power compensation devices have a great impact on the reactive power optimization of the distribution network. When modeling under the condition of determining the candidate installation nodes and capacities of ESS, SVC, and CB, the optimization results have a large human factor and are not the optimal configuration optimization scheme for the distribution network. In this regard, although the existing technology optimizes the configuration of reactive power devices and reactive power compensation in the distribution network, further ensures the voltage stability of the distribution network and effectively reduces the network loss of the distribution network, and gives full play to the advantages of various voltage regulation means of the distribution network, due to the high complexity and multi-state operation of the distribution network, it is not enough to only improve the operation ability of the distribution network from the network side through ESS, etc. It is necessary to further develop its regulation ability from the demand side and enrich the regulation means of the distribution network.

[0004] Finally, demand response (DR) is to guide users to adjust their original electricity consumption arrangements through means such as price signals or incentive mechanisms, reduce or transfer the electricity load during a certain period, improve the matching degree between the power source side and the demand side, and is of great significance for reducing the operating cost of the power grid, reducing the power supply pressure of the power grid, and maintaining the stable operation of the power grid; when conducting coordinated optimization research on the planning and operation of the distribution network, the upper and lower layer models contain many conditions, have complex coupling relationships, and the solution environment is relatively complex. With the rapid development of artificial intelligence algorithms, heuristic search algorithms represented by genetic algorithms and particle swarm optimization (PSO) provide new ideas for solving the reactive power optimization problem of the distribution network. However, the existing technologies do not adequately consider the optimization of key nodes, and the optimization solution methods are prone to falling into local optima. Due to their own defects, the existing heuristic algorithms are extremely prone to falling into local optima, increasing the difficulty of algorithm solution; therefore, it is very necessary to provide a distribution network energy storage-reactive power optimization configuration method considering demand response that takes into account coordinated optimization configuration, uses an improved multi-objective mayfly algorithm, and avoids falling into the global optimal solution. Summary of the Invention

[0005] (I) Technical Problems

[0006] In view of the above-mentioned current situation of the existing technology, the present application mainly addresses the following technical problems:

[0007] 1. The existing technology only configures static fixed-capacity reactive power compensation equipment and energy storage for the distribution network, resulting in insufficient static configuration and unable to simultaneously optimize the safety, stability, and economy of the distribution network;

[0008] 2. For the distribution network with high complexity and operation polymorphism, it is not enough for the existing technology to only improve the operation ability of the distribution network from the grid side through ESS, etc. It is necessary to further develop its regulation ability from the demand side and enrich the regulation means of the distribution network;

[0009] 3. For demand response, the existing technology does not adequately consider the optimization of key nodes, and the optimization solution method is prone to falling into local optima due to its own defects, increasing the difficulty of algorithm solution.

[0010] (II) Technical Solutions

[0011] The purpose of the present invention is to overcome the deficiencies of the existing technology and provide a distribution network energy storage-reactive power optimization configuration method considering demand response that takes into account coordinated optimization configuration, uses an improved multi-objective mayfly algorithm, and avoids falling into the global optimal solution.

[0012] The purpose of the present invention is achieved as follows: A distribution network energy storage-reactive power optimization configuration method considering demand response, which considers the coordinated optimization of active and reactive power of the distribution network with energy storage and demand response in the scenario of high proportion of new energy access, includes the following steps:

[0013] Step 1: Taking the ESS configuration capacity, the positions and capacities of CB and SVC as decision variables, an upper-layer planning model is constructed with the goal of minimizing the investment, maintenance, and operation costs of each device.

[0014] Step 2: Taking the CB gear and change time, the reactive power output of SVC, the charge and discharge power and electricity of energy storage, and the load change amount under the DR strategy as decision variables, a lower-layer optimization model is constructed with the goals of minimizing voltage deviation and demand-side response indicators.

[0015] Step 3: Improve the multi-objective mayfly algorithm. First, use Tent mapping for population initialization, and then use a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, thereby improving the search ability of the mayfly algorithm and avoiding falling into the global optimal solution.

[0016] Step 4: Based on the above models, considering the energy storage in the distribution network and the operation economy of the distribution network, in order to enhance the reliability of the distribution network operation, with the goals of minimizing network loss cost, demand-side response, and voltage deviation, a two-layer reactive power optimization of the distribution network considering energy storage and demand-side response is designed.

[0017] Step 5: Finally, through simulation verification on the IEEE 33-node system, the improved mayfly algorithm is used to solve the two-layer model of reactive power compensation optimization for the distribution network.

[0018] Furthermore, the upper-layer configuration model of the distribution network in Step 1 is specifically as follows: The upper-layer configuration model aims to minimize the system comprehensive cost, considering the investment, maintenance, and operation costs of ESS, CB, and SVC, converted to an annual basis. In addition, the network loss during the operation of the distribution network belongs to the system operation technical index, which is converted into an economic measure. The objective function of the upper-layer model of the distribution network planning model is: minC total = C ESS + C CB + C SVC + C op (6), where C ESS is the investment and maintenance cost of ESS; C CB is the investment and maintenance cost of CB; C SVC is the investment and maintenance cost of SVC; C op is the operation cost.

[0019] Furthermore, the constraint conditions of the upper-layer configuration model of the distribution network are: where P ESS,min , P ESS,max are the upper and lower limits of the rated power of ESS respectively; E ESS,min , E ESS,max are the upper and lower limits of the rated capacity of ESS respectively; Q CB,min , QCB,max are the upper and lower limits of the CB rated power; Q SVC,min and Q SVC,max are the upper and lower limits of the SVC rated power respectively.

[0020] Furthermore, the operating cost in the upper-layer configuration model of the distribution network is: C op = C loss + C buy (4), wherein, C loss and C buy are the network loss cost and the main network power purchase cost respectively; C Ploss and C Price are the network loss electricity price and the power purchase price respectively; P t g is the main network power purchase power at time t; I l,t is the current of branch l at time t; R l is the resistance of line l.

[0021] Furthermore, the lower-layer optimization model in step 2 is specifically: the lower-layer optimization model aims to minimize the voltage deviation, considers the flexibility of the load under price incentives, takes the scheduling schemes of various devices in the distribution network as decision variables, and makes the distribution network safe and stable reach the optimum by invoking demand-side response, energy storage and reactive power compensation devices; therefore, the objective function F of the lower-layer model is defined as shown in the following formula: F = min f vol + 0.5 min(1 - f con ) + 0.5 min(1 - f eco ) (13), where f vol is the voltage deviation; f con is the user comfort; f eco is the user economy.

[0022] Furthermore, the voltage deviation in the lower-layer optimization model is: wherein, V i rated is the rated voltage amplitude of node i; V i,max and V i,min are the upper and lower limits of the voltage of node i respectively; V i,t is the actual voltage value of node i at time t.

[0023] Furthermore, the constraint conditions of the lower-layer optimization model include: power flow constraint, energy storage state of charge constraint, and SVC and CB constraints.

[0024] Furthermore, in step 3, Tent mapping is used for population initialization, specifically: the expression of the Tent chaotic sequence is as follows: where \(i\) is the particle serial number, \(i = 1, 2, \cdots, N\); the Tent map is expressed as follows after Bernoulli shift transformation: \(z\) i+1 = 2z i mod 1(18), where mod 1 means taking the modulus with respect to 1, making \(z\) i+1 always lie within the interval \([0, 1)\); to avoid it falling into small periodic points or unstable periodic points, and without destroying the three major characteristics of chaotic variables, a random variable is introduced into the original Tent map expression, and the improved expression is as follows: where \(N\) is the number of particles in the sequence; \(rand(0, 1)\) is a random number in the range \([0, 1]\).

[0025] Furthermore, in step 3, a fuzzy controller is used to find the optimal values of the crossover probability \(P\) c and the mutation probability \(P\) m specifically: three triangular membership functions are generated from the fitness values of the overall set, namely the best fitness \(BF\), variance \(V\), and standard deviation \(SD\) as fuzzy inputs; the fuzzy set theory is used to calculate the \(BF\) among all individuals, and \(V\) and \(SD\) represent the variation and standard deviation of the fitness value in the \(i\)-th iteration; \(SD\) is selected as the measure of population diversity; the mathematical expressions of these three input fuzzy variables are as follows: where represents the fitness value relative to the \(n\)-th optimal compromise solution; \(UN\) represents the variation of the fitness value; \(N\) obj represents the total number of fitness functions; \(N\) pop represents the total number of mayfly populations.

[0026] Furthermore, the multi-objective mayfly algorithm in step 3 uses the centroid method to defuzzify the fuzzy controller, giving the determined values of the two probabilities; the optimal solution is determined in the final population using the fuzzy membership function, where \(F\) j is the \(i\)-th objective function value; and are the upper and lower limits of the objective function respectively; its normalized satisfaction value is solved by the following formula, and the solution with the largest satisfaction value is the optimal compromise solution, where \(n\) is the number of objective functions to be optimized.

[0027] (III) Beneficial Effects

[0028] 1. The present invention is a two-layer optimization strategy that collaboratively considers the distribution network equipment configuration and operation under the background of large-scale access of high-proportion distributed power sources to the distribution network, and is applied to the distribution network optimization technology field;

[0029] 2. The upper layer of the present invention considers the optimal configuration of energy storage and reactive power, and configures energy storage and reactive power compensation equipment. The lower layer considers demand response, and the solution uses an improved mayfly algorithm;

[0030] 3. In the case of high - proportion new - energy access, the present invention considers a two - layer optimal configuration based on energy storage and demand response, which can improve the economy and reliability of the operation of the distribution network;

[0031] 4. The present invention proposes a coordinated active - reactive power optimal configuration for a distribution network considering energy storage and demand - side response in the scenario of high - proportion new - energy access. When solving, an improved multi - objective mayfly solution algorithm is used, and a fuzzy logic controller is used to find the optimal dynamic values of the crossover probability and mutation probability, thereby improving the search ability of the mayfly algorithm and avoiding falling into the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a two - layer optimization model diagram of the reactive power compensation device for the distribution network of the present invention.

[0033] Figure 2 It is a flow chart for solving the reactive power optimization model of the distribution network of the present invention.

[0034] Figure 3 It is a topology diagram of the 33 - node system of the distribution network of the present invention.

[0035] Figure 4 It is a comparison diagram of the convergence curves of different algorithms of the present invention.

[0036] Figure 5 It is a comparison diagram of voltage deviations of different schemes of the present invention.

[0037] Figure 6 It is a comparison diagram of network losses of different schemes of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0038] In summary, at present, for the reactive power optimization problem, the modeling is still focused on the case where the candidate installation nodes and capacities of ESS, SVC, and CB have been determined. The optimization results have a large human factor and are not the optimal operation scheme for the distribution network, and the configuration of ESS, reactive power compensation devices, etc. is not carried out. Therefore, in the case of high - proportion new - energy access, the present invention considers a two - layer optimal configuration based on energy storage and demand response, which can improve the economy and reliability of the operation of the distribution network.

[0039] The following further describes the present invention in conjunction with embodiments and / or drawings.

[0040] Embodiment 1

[0041] As Figure 1-6As shown in the figure, a method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response. The method considers the coordinated optimization of active and reactive power in a distribution network with high proportion of new energy access and energy storage and demand-side response, and includes the following steps:

[0042] Step 1: Taking the configuration capacity of ESS, and the location and capacity of CB and SVC as decision variables, an upper-layer planning model is constructed with the goal of minimizing the investment, maintenance and operation costs of each device;

[0043] In the present invention, the upper-layer configuration model of the distribution network: The upper-layer configuration model aims to minimize the overall system cost, considering the investment, maintenance and operation costs of ESS, CB and SVC, converted to an annual basis. In addition, the network loss during the operation of the distribution network belongs to the system operation technical index, which is converted into an economic measurement to construct the upper-layer planning model of the distribution network.

[0044] ① Investment and maintenance cost of ESS: In the formula, N ESS is the installation quantity of ESS; r is the discount rate; c main is the equipment maintenance cost coefficient; y ESS is the service life of ESS; are the unit power and unit capacity investment costs of energy storage respectively; P i ESS 、 are the rated power and rated capacity of the i-th energy storage respectively.

[0045] ② Investment and maintenance cost of CB: In the formula, N CB is the installation quantity of CB; y CB is the service life of CB; is the unit power investment cost of CB; is the rated power of the i-th CB.

[0046] ③ Investment and maintenance cost of SVC: In the formula, N SVC is the installation quantity of SVC; y SVC is the service life of SVC; is the unit power investment cost of SVC; is the rated power of the i-th SVC.

[0047] ④ Operation cost: In the formula, C loss 、C buy are the network loss cost and the main network power purchase cost respectively; C Ploss 、C Price are the network loss electricity price and the power purchase price respectively; P t gis the main online shopping power at time t; I l,t is the current of branch l at time t; R l is the resistance of line l.

[0048] In summary, the upper-layer model objective function is: minC total = C ESS + C CB + C SVC + C op (6), the constraint conditions of the upper-layer model are: In the formula, P ESS,min , P ESS,max are respectively the upper and lower limits of the rated power of the ESS; E ESS,min , E ESS,max are respectively the upper and lower limits of the rated capacity of the ESS; Q CB,min , Q CB,max are respectively the upper and lower limits of the rated power of the CB; Q SVC,min , Q SVC,max are respectively the upper and lower limits of the rated power of the SVC.

[0049] Step 2: Taking the CB gear and change time, the reactive power output of the SVC, the charge and discharge power and electricity of the energy storage, and the load change amount under the DR strategy as decision variables, construct a lower-layer optimization model with the minimum voltage deviation and the demand-side response index as the objectives;

[0050] In the present invention, the lower-layer optimization model of the distribution network: The lower-layer optimization model aims at the minimum voltage deviation, considers the flexibility of the load under price incentives, takes the scheduling schemes of various devices in the distribution network as decision variables, and makes the distribution network safe and stable reach the optimum by invoking the demand-side response, energy storage and reactive power compensation devices.

[0051] ① Demand-side response: In the case of large-scale access of DGs, DR can guide the electricity consumption behavior of users in the form of time-of-use electricity prices, transfer some of the load during peak load periods to off-peak load times, achieve the purpose of peak shaving and valley filling, balancing the load, thereby reducing network losses and improving the economic operation of the distribution network; The model of the present invention considers two types of typical loads, commercial loads and residential loads, and adopts the control means of transferable loads, and its constraints are:

[0052] Mainly considering the demand-side response affected by time-of-use electricity prices, the load transfer amount is related to the electricity prices at different times, and users adjust their electricity consumption habits with reference to the electricity prices at different times to improve the load curve; The transferable load in the demand-side response changes within a certain range and ensures that the overall change load amount within a day is maintained within a certain small range, and its constraints are: In the formula, is the change amount generated by the demand-side response at node n at time t; and They are the upper and lower limits of the change in demand response at the nth node at time t, respectively. is the load before demand response at the nth node at time t; α is the allowable variable range of electricity consumption of users within one cycle; N b is the number of nodes in the distribution network.

[0053] The user economy and user comfort are introduced to represent the economy and comfort of users after considering demand response. In the formula, C t is the electricity price at time t.

[0054] ② Voltage deviation: In the formula, V i rated is the rated voltage amplitude of node i; V i,max and V i,min are the upper and lower limits of the voltage of node i, respectively; V i,t is the actual voltage value of node i at time t.

[0055] To sum up, the objective function F of the lower-layer model is defined as follows: F = min f vol + 0.5 min(1 - f con ) + 0.5 min(1 - f eco )(13).

[0056] Constraint conditions: ① Power flow constraint: To ensure the power quality and safe operation of the distribution network, the variables need to meet certain constraint conditions. The equality constraint of the model is the power flow constraint, and the specific calculation model is as follows: In the formula, P i and Q i are the active power and reactive power at node i, respectively; U i and U j are the voltage amplitudes of nodes i and j, respectively; G ij and B ij are the conductance and susceptance of branch k connecting nodes i and j; δ ij is the phase difference between nodes i and j.

[0057] ② Energy storage state of charge constraint: In the formula, is the power of the ith ESS at time t; S SOC,i (t) is the state of charge of the ith ESS at time t; S SOC,i,min and S SOC,i,max are the upper and lower bounds of the state of charge of the ith ESS, respectively; η is the charge-discharge efficiency.

[0058] ③ SVC and CB constraints: The tap changes of CB cannot be too frequent, so the number of tap changes of CB is constrained: In the formula, is the reactive power at node i at time t; is the number of CB groups connected at node i at time t; is the capacity of a single CB; is the total number of CB groups installed at node i; is the maximum switching times of the CB; is the reactive power output of the SVC at node i at time t; is the upper limit of the SVC reactive power at node i.

[0059] Step 3: Improve the multi-objective mayfly algorithm. First, use Tent mapping for population initialization, and then use a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, so as to improve the search ability of the mayfly algorithm and avoid falling into the global optimal solution;

[0060] Step 4: Based on the above model, considering the energy storage in the distribution network and the operation economy of the distribution network, in order to enhance the reliability of the distribution network operation, with the objectives of minimizing the network loss cost, demand-side response, and voltage deviation, design a two-layer reactive power optimization of the distribution network considering energy storage and demand-side response;

[0061] Step 5: Finally, conduct simulation verification through the IEEE 33-node system, and use the improved mayfly algorithm to solve the two-layer model of the distribution network reactive power compensation optimization.

[0062] The present invention is a method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response, specifically a two-layer optimization strategy that collaboratively considers the equipment configuration and operation of the distribution network under the background of large-scale access of high-proportion distributed power sources to the distribution network. In use, the present invention considers the optimization of energy storage and reactive power configuration under the condition of high-proportion new energy access, configures energy storage and reactive power compensation equipment, and considers a two-layer optimization configuration based on energy storage and demand response, which can improve the economy and reliability of the distribution network operation; when solving, the present invention uses an improved multi-objective mayfly solution algorithm, and uses a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, so as to improve the search ability of the mayfly algorithm and avoid falling into the global optimal solution; the present invention has the advantages of considering coordinated optimization configuration, using an improved multi-objective mayfly algorithm, and avoiding falling into the global optimal solution.

[0063] Embodiment 2

[0064] As Figure 1-6 shown, a method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response, the method considers the coordinated optimization of active and reactive power of the distribution network with energy storage and demand-side response in the scenario of high-proportion new energy access, and includes the following steps:

[0065] Step 1: Construct an upper-layer planning model with the ESS configuration capacity, the positions and capacities of the CB and SVC as decision variables, aiming to minimize the investment, maintenance, and operation costs of each device.

[0066] Step 2: Construct a lower-layer optimization model with the CB gear and change time, the SVC reactive power output, the charge and discharge power and electricity of the energy storage, and the load change under the DR strategy as decision variables, aiming to minimize the voltage deviation and the demand-side response index.

[0067] Step 3: Improve the multi-objective mayfly algorithm. First, use the Tent map for population initialization, and then use a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, thereby improving the search ability of the mayfly algorithm and avoiding falling into the global optimal solution.

[0068] In the present invention, the planning of reactive power compensation devices is further considered. Since the upper and lower layer models contain many conditions and the solution environment is relatively complex, the particle swarm algorithm is used in the upper layer of the present invention to solve the planning problem, and the improved multi-objective mayfly algorithm is used in the lower layer to solve the optimization problem.

[0069] ① Improved multi-objective mayfly algorithm: The mayfly algorithm has high search accuracy, fast convergence speed, good stability, and strong robustness, but it is prone to falling into the local optimal problem. For this, the Tent map is used for population initialization. The Tent chaotic sequence has the characteristics of randomness, ergodicity, and regularity. Using these characteristics for optimization search can effectively maintain the population diversity, inhibit the algorithm from falling into the local optimum, and improve the global search ability. The expression of the Tent chaotic sequence is as follows: In the formula, i is the particle serial number, i = 1, 2,..., N.

[0070] The Tent map is expressed as follows after Bernoulli shift transformation: z i+1 = 2z i mod1(18), where mod1 means taking the modulus of 1, so that z i+1 is always in the interval [0, 1).

[0071] Analysis shows that there are small cycles and unstable periodic points in the Tent chaotic sequence. To avoid it falling into small periodic points or unstable periodic points and at the same time not destroying the three major characteristics of the chaotic variable, a random variable is introduced into the original Tent map expression, and the improved expression is as follows: In the formula, N is the number of particles in the sequence; rand(0, 1) is a random number in the range of [0, 1].

[0072] ② Another limitation of the multi-objective mayfly algorithm is the weak mating process and limited search ability. Therefore, a fuzzy controller is used to find the crossover probability P cand the mutation probability P m to improve the search ability, convergence speed, and diversity.

[0073] The fuzzy controller consists of three steps: fuzzification, establishment of a fuzzy rule base, and defuzzification:

[0074] In the first step, the fitness values of the overall set generate three triangular membership functions, namely the best fitness (BF), variance (V), and standard deviation (SD), as fuzzy inputs; using fuzzy set theory, the BF, V, and SD among all individuals are calculated, where V and SD represent the variation and standard deviation of the fitness values at the i-th iteration; SD is selected as a measure of population diversity, which has a special influence on the true Pareto optimal solution; the mathematical expressions of these three input fuzzy variables are as follows: In the formula, represents the fitness value relative to the n-th optimal compromise solution; UN represents the variation of the fitness value; N obj represents the total number of fitness functions; N pop represents the total number of the mayfly population.

[0075] The algorithm uses the centroid method for defuzzification, giving two definite values of probability; the optimal solution is determined in the final population using the fuzzy membership function, In the formula, F j is the i-th objective function value; and are the upper and lower limits of the objective function, respectively; its normalized satisfaction value is solved by the following formula, and the solution with the largest satisfaction value is the optimal compromise solution, In the formula, n is the number of objective functions to be optimized.

[0076] Step 4: Based on the above model, considering the energy storage in the distribution network and the operation economy of the distribution network, in order to enhance the reliability of the distribution network operation, a two-layer reactive power optimization of the distribution network considering energy storage and demand-side response is designed with the objectives of minimizing the network loss cost, demand-side response, and voltage deviation;

[0077] Step 5: Finally, through simulation verification using the IEEE 33-node system, the improved mayfly algorithm is used to solve the two-layer model of the reactive power compensation optimization of the distribution network.

[0078] In the present invention, case analysis: Taking the IEEE 33-node distribution network system for case analysis, the network structure is as Figure 3As shown; the base voltage of the distribution network is 12.66 kV and the base capacity is 100 MV·A; in the simulation, 3 groups of ESS are connected, with the rated power range of [20, 60] kW and the rated capacity range of [40, 120] kW·h; 2 groups of CB are connected, with the unit power of 100 kvar for each group, and at most 6 groups can be connected at each location; 1 group of SVC is connected, with the unit power of 50 kvar; the maximum ramp rate of the main grid is 500 kW / h.

[0079] Table 1 shows the configuration results, and Table 2 shows the optimization results under different strategies; compared with the fixed configuration, the double-layer configuration reduces the network loss cost and the investment, maintenance and operation costs by 18.25% and 16.07% respectively, and the main grid power purchase cost is reduced by 15.49%; it can be concluded that considering ESS and DR can effectively reduce the network loss cost and the main grid power purchase cost of the distribution network and promote the economic and stable operation of the distribution network.

[0080] Table 1 Optimal configuration results

[0081]

[0082] Table 2 Target results of different optimization models

[0083]

[0084] In order to verify the advantages of the proposed optimization strategy in improving the voltage quality of the distribution network, reducing network losses, and reducing the load peak-valley difference, the following 3 schemes are used for comparative analysis: Scheme 1: Consider the coordinated optimization of CB and SVC; Scheme 2: Consider DR on the basis of Scheme 1; Scheme 3: Consider ESS on the basis of Scheme 2.

[0085] Scheme 1 considers CB and SVC, which effectively improves the network loss and voltage level of the distribution network, but still does not solve the matching problem between the renewable energy output and the load demand; Scheme 2 adds DR on the basis of Scheme 1, considering both the safe and economic operation of the distribution network and the user comfort and user economy. Compared with Scheme 1, Scheme 2 reduces the network loss and voltage deviation by 19.66% and 40.38% respectively; Scheme 3 adds ESS on the basis of Scheme 2, balances the active power flow, reduces the line loss, and ESS provides a certain amount of reactive power support to improve the voltage quality; the addition of ESS further improves the load curve. Compared with Scheme 2, Scheme 3 reduces the network loss by 45.03% and the voltage deviation by 72.56%; it can be seen that the proposed optimization strategy reduces the network loss and voltage deviation and improves the economy and security of the distribution network operation.

[0086] Table 3 Time-of-use electricity price

[0087]

[0088] Table 4 Simulation Parameter Settings

[0089]

[0090] The present invention relates to a method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response, specifically a two-layer optimization strategy that collaboratively considers the configuration and operation of distribution network equipment in the context of large-scale access of a high proportion of distributed power sources to the distribution network. In use, the present invention considers the optimization configuration of energy storage and reactive power in the case of high-proportion new energy access, configures energy storage and reactive power compensation equipment, and considers a two-layer optimization configuration based on energy storage and demand response, which can improve the economy and reliability of the operation of the distribution network; when solving, the present invention adopts an improved multi-objective mayfly algorithm, uses a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, thereby improving the search ability of the mayfly algorithm and avoiding falling into the global optimal solution; the present invention has the advantages of considering coordinated optimization configuration, adopting an improved multi-objective mayfly algorithm, and avoiding falling into the global optimal solution.

Claims

1. A method for optimal configuration of energy storage and reactive power in a distribution network considering demand-side response, characterized in that: The method considers the active and reactive power coordinated optimal configuration of a distribution network with energy storage and demand-side response in the scenario of high proportion of new energy access, including the following steps: Step 1: Taking the configured capacity of ESS, the positions and capacities of CB and SVC as decision variables, an upper-layer planning model is constructed with the goal of minimizing the investment, maintenance and operation costs of each device; Step 2: Taking the CB gear and change time, the reactive power output of SVC, the charge and discharge power and electricity of energy storage, and the load change amount under the DR strategy as decision variables, a lower-layer optimization model is constructed with the goals of minimizing voltage deviation and demand-side response index; Step 3: Improve the multi-objective mayfly algorithm. First, use Tent mapping for population initialization, and then use a fuzzy logic controller to find the optimal dynamic values of the crossover probability and mutation probability, so as to improve the search ability of the mayfly algorithm and avoid falling into the global optimal solution; Step 4: Based on the above model, considering the energy storage in the distribution network and the operation economy of the distribution network, a two-layer reactive power optimization of the distribution network considering energy storage and demand-side response is constructed with the goals of minimizing network loss cost, demand-side response and voltage deviation; Step 5: Finally, through simulation verification with the IEEE 33-node system, the improved mayfly algorithm is used to solve the two-layer model of the distribution network reactive power compensation optimization.

2. The method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response according to claim 1, wherein: The distribution network upper-layer configuration model in step 1 is specifically as follows: The upper-layer configuration model aims to minimize the overall system cost. Considering the investment and maintenance costs of ESS, CB, and SVC, which are converted to an annual basis. In addition, the network losses during the operation of the distribution network are system operation technical indicators, which are converted into economic measurements. The upper-layer planning model of the distribution network is constructed. The objective function of the upper-layer model is: minC total = C ESS + C CB + C SVC + C op (6), where C ESS is the investment and maintenance cost of ESS; C CB is the investment and maintenance cost of CB; C SVC is the investment and maintenance cost of SVC; C op is the operation cost.

3. A method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response according to claim 2, characterized in that: The constraint conditions of the upper-layer configuration model of the distribution network Namely: In the formula, P ESS,min , P ESS,max are respectively the upper and lower limits of the rated power of the ESS; E ESS,min , E ESS,max are respectively the upper and lower limits of the rated capacity of the ESS; Q CB,min , Q CB,max are respectively the upper and lower limits of the rated power of the CB; Q SVC,min , Q SVC,max are respectively the upper and lower limits of the rated power of the SVC.

4. The method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response according to claim 3, characterized in that: The operating cost in the upper-layer distribution network configuration model is: C op = C loss + C buy (4), wherein, C loss , C buy are the network loss cost and the main grid power purchase cost respectively; C Ploss , C Price are the network loss electricity price and the power purchase price respectively; is the main grid power purchase power at time t; I l,t is the current of branch l at time t; R l is the resistance of line l.

5. The method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response according to claim 1, characterized in that: The lower-layer optimization model in Step 2 is specifically as follows: The lower-layer optimization model aims to minimize the voltage deviation, considers the flexibility of the load under price incentives, takes the scheduling schemes of various devices in the distribution network as decision variables, and enables the optimal safety and stability of the distribution network by invoking demand-side response, energy storage, and reactive power compensation devices. Therefore, the objective function F of the lower-layer model is defined as shown in the following formula: F = min f vol + 0.5 min(1 - f con ) + 0.5 min(1 - f eco )(13), where f vol is the voltage deviation; f con is the user comfort level; f eco is the user economy.

6. The method for optimal configuration of distribution network energy storage and reactive power considering demand side response according to claim 1, characterized in that: The voltage deviation in the lower-layer optimization model is as follows: In the formula, is the rated voltage amplitude of node i; V i,max , V i,min are the upper and lower voltage limits of node i respectively; V i,t is the actual voltage value of node i at time t.

7. A method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response as described in claim 6, characterized in that: The constraint conditions of the lower-layer optimization model include: power flow constraint, energy storage state of charge constraint, and SVC and CB constraints.

8. A method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response according to claim 1, characterized in that: In step 3, Tent mapping is used for population initialization, specifically: The expression of the Tent chaotic sequence is as follows: In the formula, i is the particle serial number, i = 1, 2,..., N; The Tent mapping is expressed as follows after Bernoulli shift transformation: z i+1 = 2z i mod1(18), in the formula, mod1 represents taking the modulus of 1, so that z i+1 is always within the interval [0, 1); To avoid it falling into small periodic points or unstable periodic points, and at the same time not destroying the three major characteristics of the chaotic variable, a random variable is introduced into the original Tent mapping expression, and the improved expression is as follows: In the formula, N is the number of particles in the sequence; rand(0, 1) is a random number in the range of [0, 1].

9. The method for optimizing the configuration of distribution network energy storage and reactive power considering demand-side response according to claim 8, characterized in that: In step 3, a fuzzy controller is used to find the optimal values of the crossover probability P c and the mutation probability P m , specifically: The fitness values of the overall set generate three triangular membership functions, namely the best fitness BF, variance V, and standard deviation SD as fuzzy inputs; the fuzzy set theory is used to calculate the BF among all individuals, and V and SD represent the variation and standard deviation of the fitness values in the i-th iteration; SD is selected as the measure of population diversity; the mathematical expressions of these three input fuzzy variables are as follows: In the formula, represents the fitness value relative to the n-th optimal compromise solution; UN represents the variation of the fitness value; N obj represents the total number of fitness functions; N pop represents the total number of mayfly populations.

10. A method for optimizing the configuration of energy storage and reactive power in a distribution network considering demand-side response as described in claim 9, characterized in that: In the multi-objective mayfly algorithm in step 3, the centroid method is used to defuzzify the fuzzy controller, and two definite values of probability are given; the optimal solution is determined in the final population by using the fuzzy membership function. where F j is the value of the i-th objective function; and are the upper and lower limits of the objective function respectively; its normalized satisfaction value is solved by the following formula, and the solution with the largest satisfaction value is the optimal compromise solution. where n is the number of objective functions to be optimized.

Images