Method for distributed photovoltaic optimization point distribution based on zero-approaching load center of gravity

Abstract

The invention discloses a distributed photovoltaic optimization stationing method based on the center of gravity of a zero-approaching load, which optimizes the access position and access capacity of distributed photovoltaic based on the center of gravity of a zero-approaching electrical load and reduces network loss. The method comprises the following steps: taking distributed photovoltaic node positions and PV capacity of a network merged into each distributed photovoltaic node as decision variables of a model; the method comprises the following steps: taking a maximum zero-approaching electrical load gravity center as a target function, setting constraint conditions as a voltage out-of-limit constraint based on a voltage qualification rate, a line current-carrying capacity constraint and a grid-connected maximum PV admission capacity constraint, and obtaining a PV access scheme for optimizing a network state through iterative calculation by using an optimization algorithm. The method is used for optimizing and planning the distributed photovoltaic access position and access capacity of an existing network.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of photovoltaic power generation and access technology, and in particular to a method for optimizing the layout of distributed photovoltaics based on a zero load center of gravity. Background Art

[0002] Under the "dual carbon" goal, the application of distributed photovoltaics in distribution networks is becoming more and more widespread. Due to insufficient initial planning and later transformation, the existing network may have problems such as unreasonable structure and poor operating status. In order to give full play to the advantages of distributed photovoltaics, it is possible to consider how to use it to improve the network operation status.

[0003] Since E.M. Hunter proposed the standardized design of load center unit substations in 1942, people have gradually realized that the closer the transformer is to the load center of gravity, the smaller the loss of the power grid. Ultimately, based on the experience summary and induction of long-term engineering practice, this feature was incorporated into the industrial standard.

[0004] The conventional load center of gravity is defined as follows:

[0005]

[0006] Where: GBc x With GBc y are the horizontal and vertical coordinates of the center of gravity of the obtained load, respectively. x and y are the horizontal and vertical coordinates of node i in the plane coordinate system, representing the geographic spatial position of node i. i Then is the power of node i. Since the load center of gravity is based on geographic coordinates, the essence is to estimate the line length through the geographic coordinate value of the load center of gravity to map the line impedance, so this paper calls it GBc (Geographic Barycenter), which can be extended from the plane coordinate system to the space coordinate system.

[0007] From the given definition, it can be seen that the load center of gravity is a network attribute. Its essence is the weighted average of the estimated line length of each node with the node load as the weight, which can reflect the state of the line impedance and load distribution of the network, that is, a way to describe the balance of the network load structure. According to the theory, the smaller the deviation between the substation coordinates and the coordinates of the geographical load center of gravity, the lower the network line loss.

[0008] The load center calculation based on geographic coordinates was originally used to "plan the location of the substation when the load is determined." For an existing network, the substation location and the load node location are already determined, so the electrical distance can be used to more accurately characterize the network structure. So the definition can be modified to

[21] :

[0009]

[0010] Where: R i is the line resistance from node i to the root node, P i Then is the power size of node i. At this point, further, in order to consider the complete influence of electrical distance, that is, line impedance, the definition of the load center of gravity concept is modified as follows:

[0011]

[0012] The electrical load gravity center EBc (Electrical Bary center) is obtained. Where: | Z i | is the impedance modulus of the line from node i to the root node, P i Then is the power of node i.

[0013] Since a PV access scheme is required and the PV output curve changes over time, we need to expand the definition of EBc to adapt to the calculation of time series scenarios. Summary of the invention

[0014] In order to overcome the above technical defects, the embodiment of the present application provides a method for optimizing the distribution of distributed photovoltaics based on the zero load gravity center. From the above analysis and verification, it can be seen that when EBc approaches 0, the network line loss is smaller. The zero load gravity center can be obtained based on the electrical load gravity center, and then the distribution network distributed photovoltaics can be optimized to reduce network line losses.

[0015] The main plans are as follows:

[0016] The node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node integrated into the network are used as decision variables of the model;

[0017] The objective function is to maximize the center of gravity of the zero electrical load OEBc, and the constraints are set as voltage limit constraints based on voltage qualification rate, line current carrying capacity constraints, and maximum grid-connected PV access capacity constraints.

[0018] Among them, the voltage qualification rate is the ratio of the voltage not exceeding the upper and lower limits within a given period of time. The voltage over-limit constraint based on the voltage qualification rate is that the optimized network voltage qualification rate should be greater than the voltage qualification rate of the original network. The line current carrying capacity constraint is that the line current carrying capacity is not greater than the maximum current carrying capacity allowed by the line used. The maximum PV access capacity constraint for grid connection is that the PV capacity of the distributed photovoltaic node connected to the network does not exceed 25% of the maximum load in the power supply area of ​​the upper transformer.

[0019] Using the optimization algorithm, the node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node connected to the network are used as the decision variables of the model. On the basis of satisfying the voltage over-limit constraint, the line current carrying capacity constraint and the maximum PV access capacity constraint of the grid, the optimization calculation is performed with maximizing the zero electrical load gravity center OEBc as the optimization goal and the maximum number of iterations as the termination condition to obtain the PV access plan for the optimized network state, which is used to optimize the planning of the access location and access capacity of distributed photovoltaics in an existing network.

[0020] Optionally, in some possible implementations, the maximum zero-approaching electrical load center of gravity OEBc is: maxOEBc, where:

[0021] Optionally, in some possible implementations, the expression of the voltage limit constraint based on the voltage qualification rate is: Q net_new >Q net_ori , where Q net_new is the qualified rate of network voltage after optimization, Q net_ori is the network voltage qualification rate before optimization, where the node voltage limit range is defined as x1V nom ≤V i ≤x2V nom , where V i is the voltage amplitude of node i, V nom is the system nominal voltage, x1 and x2 are the given voltage upper and lower limits; the voltage qualification rate is defined as: In the formula, Q net is the network voltage qualification rate, Q i is the voltage qualification rate of node i, N is the number of nodes, T upper is the time when the corresponding node voltage exceeds the upper limit, T lower is the time when the corresponding node voltage exceeds the lower limit, T total The total time measured for the corresponding node.

[0022] Optionally, in some possible implementations, the expression of the line current carrying capacity constraint is: lineload ≤I max_i_ka , where I lineload is the line current carrying capacity, I max_i_ka The maximum current carrying capacity allowed by the line used.

[0023] Optionally, in some possible implementations, the expression of the grid-connected maximum PV access capacity constraint is: S t_pv ≤0.25S trans , where S t_pv is the total PV capacity, S trans is the transformer capacity.

[0024] Optionally, in some possible implementations, the optimization algorithm uses a genetic algorithm. To maximize OEBc, a fitness evaluation function is set to return the fitness of each individual and take the maximum value. To define the chromosome encoding, a set of possible PV capacity values ​​C is used as the range of individual genes: C = [0, allowed_cap], step (13) where allowed_cap is the maximum capacity of a single PV allowed, and step is a random step. At this time, the maximum capacity of each PV is within the closed interval between 0 and allowed_cap, and the value randomly obtained by taking step as the random step is used as the chromosome encoding of each individual, that is, the PV access plan.

[0025] Optionally, in some possible implementations, after obtaining the PV access solution that optimizes the network status, the method further includes:

[0026] Use network loss index and network voltage qualification rate index to judge the optimization effect of the network before and after optimization;

[0027] Among them, the network loss index is defined as: Where P loss_new is the total network loss value after optimization, P loss_ori is the total loss value of the original network;

[0028] The network voltage qualification rate is defined as: In the formula, Q net is the network voltage qualification rate, Q i is the voltage qualification rate of node i, N is the number of nodes, T upper is the time when the corresponding node voltage exceeds the upper limit, T lower is the time when the corresponding node voltage exceeds the lower limit, T total The total time measured for the corresponding node.

[0029] Optionally, in some possible implementations, taking the maximum number of iterations as the termination condition means completing the evaluation of all individuals in the search space with the number of generations multiplied by the population size of each generation.

[0030] The embodiments of the present application adopt the above technical solution to achieve the following technical effects:

[0031] As a network attribute, the load center of gravity can describe the state of a network from a global perspective. The main idea of ​​the method proposed in this paper is to convert the angle, and reverse the method originally used to determine the location of the power supply point to determine the load state of the power consumption point, and then determine the access plan of distributed photovoltaics. Its essence is to express the balance of the network power supply structure. It is a process in which distributed photovoltaics are used as active power injection in a time sequence process to change the load center of gravity at each time point, that is, to adjust a group of load points that cause network imbalance, so as to achieve a balanced supply and demand relationship in the network. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0033] Figure 1 A flow chart of a method for optimizing the placement of distributed photovoltaics based on the zero load center of gravity in this application;

[0034] Figure 2 A schematic diagram of the topology of the Kerber Dorfnetz network in this application;

[0035] Figure 3 This is a schematic diagram comparing active power loss before and after optimization on a typical annual day in this application;

[0036] Figure 4 This is a schematic diagram comparing the typical node voltage before and after optimization on a typical day of the year in this application. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0038] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0039] In the description of the present application, it should be understood that the numerical labels before the steps do not indicate the order in which the steps are executed, but are only used to facilitate the description of the present application and to distinguish each step, and therefore should not be understood as a limitation on the present application.

[0040] The function of the load gravity center based on geographic coordinates is: planning the location of the power supply point, that is, changing the impedance from the large load node to the power supply point by changing the location of the power supply point, so as to ensure a better network state. However, for an existing network, the location of the power supply point and each load node is fixed, that is, the impedance is fixed, so if you want to change the load gravity center, you can only change the load size on each load node, that is, inject active power into each load node, and connecting PV to the load node can do this. Therefore, we can use the concept of load gravity center used to plan the location of the power supply point to plan the access location and access capacity of distributed photovoltaics in an existing network. Its essence is to change the large load node far away from the power supply point into a normal load node or a small load node.

[0041] For an existing network, the substation location and load node location have been determined, so the electrical distance can be used to more accurately characterize the network structure. The definition of the load center of gravity can be modified to:

[0042] Where: R i is the line resistance from node i to the root node, P i Then is the power of node i.

[0043] At this time, due to the definition modification, the numerical dimension of the load gravity center is reduced from the original spatial dimension to one dimension. Therefore, the original "the smaller the offset between the geographical load gravity center coordinates and the substation coordinates" should be modified to "the closer the absolute value of the resistance load gravity center RBc (Resistance Barycenter) is to 0, the lower the network line loss is."

[0044] Furthermore, in order to consider the complete influence of electrical distance, that is, line impedance, the definition of the load center of gravity concept is modified as follows: The electrical load center of gravity EBc (Electrical Barycenter) is obtained, where: i | is the impedance modulus of the line from node i to the root node, P i Then is the power of node i.

[0045] The purpose of this application is to construct a model based on the load center of gravity theory to optimize the network status with a distributed photovoltaic access scheme, and the access scheme will be obtained by solving the model. The model includes two main aspects: objective function and constraint conditions. Among them, the objective function will be obtained by expanding the concept of load center of gravity; for a distribution network, we usually mainly consider its voltage quality through the voltage qualification rate; consider its line safety through the line current carrying capacity; and limit the maximum PV access capacity to ensure network stability and power supply normality from a general point of view, and reduce the occurrence rate or severity of possible voltage fluctuations, harmonics, three-phase imbalance and other problems. Therefore, the constraints designed for this model are voltage qualification rate, line current carrying capacity, and maximum PV access capacity to ensure that the access scheme obtained by solving the model can enable the power grid to operate within a reasonable range of voltage, current carrying, electric reverse current and other values.

[0046] The output curve of photovoltaic power generation is a curve that changes over time. Therefore, we need to extend the concept of load center of gravity so that it can play a role in a time series process. To this end, we need to calculate the value of the electrical load center of gravity at each time point in the time series process and integrate it into an index value.

[0047] Combined with the previous text, we can know that, without considering the existence of negative node power, the smaller the EBc of a network, the smaller the network loss of the network will tend to be. When considering the situation of PV access, the PV output may be greater than the load, so negative power may appear. Then, the theory should be modified to: "EBc is closer to zero (or the smaller the absolute value)". Then, by calculating the EBc of each time node and adding them up, we can get a value that describes the entire time series process. Because the electrical load center of gravity value is required to tend to zero, it is called "zero electrical load center of gravity (OEBc)", which is defined as follows:

[0048]

[0049] In the formula, T represents the number of time points in the time series data; N represents the number of load nodes involved in the calculation; T*N can be used to normalize the value, eliminate the numerical differences caused by different network scales and time series scales, and facilitate the comparison of the performance of the value in different cases. j represents the sequence number of T time points in the traversed time series interval. If the sum of all EBcs is zero, OEBc is given a maximum value, which can be regarded as an impossible event in actual calculations (because it is almost impossible for a power network to have a complete balance between load and photovoltaic output, even if the sum of all EBcs is completely 0, so there is almost no division by zero problem).

[0050] The meaning of the zero electrical load gravity center OEBc is: the reciprocal of the sum of the absolute values ​​of EBc at each time point. At this time, OEBc represents the degree to which the load gravity center at each time point approaches 0 during the simulated time period. The higher the degree, the larger the value of this definition, and vice versa.

[0051] Based on the above definition of OEBc, we can build a model by solving the maximum OEBc, supplemented by voltage qualification rate constraints, line current carrying capacity constraints, and maximum PV access capacity constraints to obtain a PV access solution that can optimize the network status.

[0052] The method for optimizing the layout of distributed photovoltaics based on the zero load center of gravity in this application includes:

[0053] S1. The node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node incorporated into the network are used as decision variables of the model;

[0054] S2, taking the maximization of the zero-approaching electrical load center of gravity OEBc as the objective function, and setting the constraints as the voltage qualification rate constraint based on voltage exceeding the limit, the line current carrying capacity constraint and the maximum PV access capacity constraint for grid connection;

[0055] S3. Using the optimization algorithm, the node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node connected to the network are used as the decision variables of the model. On the basis of satisfying the voltage qualification rate constraints, line current carrying capacity constraints and the maximum grid-connected PV access capacity constraints, the optimization calculation is performed with maximizing the zero electrical load gravity center OEBc as the optimization goal to obtain the PV access plan that optimizes the network status, which is used to plan the access location and access capacity of distributed photovoltaics in an existing network.

[0056] Among them, the objective function is: maxOEBc(6) where,

[0057]

[0058] The node voltage limit range, for the user receiving end, the three-phase power supply voltage deviation of 10kV and below is ±7% of the nominal voltage, that is, where is the voltage amplitude of the node and is the system nominal voltage. Since this paper uses the voltage qualification rate as the voltage quality indicator, the node voltage constraint range should be used as the calculation condition of the voltage qualification rate, and the voltage qualification rate is the relevant constraint condition.

[0059] Specifically, the voltage qualification rate is the ratio of the voltage not exceeding the upper and lower limits within a given period of time. The voltage limit constraint based on the voltage qualification rate is that the voltage qualification rate of the optimized network should be greater than the voltage qualification rate of the original network, to indicate that the optimized configuration scheme has indeed optimized the network status. Its expression is: Q net_new >Q net_ori , where Q net_new is the qualified rate of network voltage after optimization, Q net_ori is the network voltage qualification rate before optimization.

[0060] For voltage over-limit constraints: it mainly lies in the general requirements of the distribution network, that is, the voltage qualification rate requirements. Voltage qualification means that the voltage value is within the allowable deviation range. Therefore, the following lists the voltage deviation range based on the regulations and how to calculate the voltage qualification rate. The voltage qualification rate requirement for the example kerber dorfnetsz used in this article is: after the network applies the distributed photovoltaic access solution obtained by the solution model, its voltage qualification rate should be higher than the original network, that is, the network operation has been improved in terms of voltage quality.

[0061] Regarding the node voltage limit range, the fifth article of Chapter 2 of the "Regulations on Voltage Quality and Reactive Power Management of the Power System of the State Grid Corporation of China" stipulates that for the user receiving end, the three-phase power supply voltage deviation of 10kV and below is ±7% of the nominal voltage, that is, 0.93V nom ≤V i ≤1.07V nom , where V i is the voltage amplitude of node i, V nom is the nominal voltage of the system. Since this paper takes the voltage qualification rate as the voltage quality indicator, the node voltage constraint range should be used as the calculation condition of the voltage qualification rate, and the voltage qualification rate should be used as the relevant constraint condition.

[0062] Regarding the voltage qualification rate, the “Regulations on Voltage Quality and Reactive Power Management of the Power System of State Grid Corporation of China” defines the network voltage qualification rate as: In the formula, Q net is the network voltage qualification rate, Q i is the voltage qualification rate of node i, N is the number of nodes, T upper is the time when the corresponding node voltage exceeds the upper limit, T loweris the time when the corresponding node voltage exceeds the lower limit, T total The total time measured for the corresponding node.

[0063] Specifically, the line current carrying capacity constraint is that the line current carrying capacity is not greater than the maximum current carrying capacity allowed by the line used, and its expression is: lineload ≤I max_i_ka , where I lineload is the line current carrying capacity, I max_i_ka The maximum current carrying capacity allowed by the line used.

[0064] For line current carrying capacity constraints: it is mainly to consider the general safety requirements of the distribution network, that is, to ensure that the line current carrying capacity does not exceed the line carrying limit, which also implies that the PV access plan should fall within a certain reasonable range to avoid connecting too large a capacity PV at a specific node, resulting in excessive reverse current during reverse power transmission, causing danger. The "maximum current carrying capacity of the adopted line model" applied by the constraint condition depends on the line type used in the example network.

[0065] Specifically, the maximum PV access capacity constraint for grid connection is that the PV capacity of the distributed PV node connected to the network does not exceed 25% of the maximum load in the power supply area of ​​the upper transformer, and its expression is: S t_pv ≤0.25S trans , where S t_pv is the total PV capacity, S trans is the transformer capacity.

[0066] The mechanism analysis of the network optimization and network loss reduction after optimizing the layout based on the zero electrical load gravity center + voltage over-limit constraint + line current carrying capacity constraint + grid-connected maximum PV access capacity constraint is as follows:

[0067]

[0068] In a power system, we have the following loss calculation formula: P T =e T ·Ge+f T ·Bf;

[0069] Where: P T is the network loss, e is the real part of the voltage vector, G is the node conductance matrix of the power system, e T is the transpose of the real part of the voltage vector; f is the imaginary part of the voltage vector, B is the node susceptance matrix of the power system, f T is the transpose of the imaginary part of the voltage vector.

[0070] When we consider it from the perspective of each node, the above loss calculation formula is transformed into: P i =e i2 ·G ii +f i 2 ·B ii ; Add up the loss values ​​of each node to get the overall network loss value, then:

[0071] Divide both sides by the total network power We can get: Among them, P ％ is the ratio of power loss to total network power.

[0072] By We can Combining the two equations, we get:

[0073]

[0074] At this time, for a certain network, each constant is determined, and it can be seen that P ％ Correlation with EBc.

[0075] If the influence of reactive power is ignored for low voltage distribution network, we can get:

[0076] And because: If only the real part of the voltage is taken, then:

[0077]

[0078] At this point, it can be observed that: the more network nodes there are, the stronger the correlation between the loss ratio and the load gravity center. Since the impedance or resistance in the formula comes from the definition of the load gravity center, which means the impedance from the root node to a certain load point, the remote node actually reuses the line impedance of the near-end node, that is, the low-impedance path undertakes the main power transmission. Therefore, it can be seen from the formula that the smaller the impedance value of the node is, the greater the impact of P ％ The correlation with RBc has a greater impact. However, in the actual analysis process, the node distribution problem and the accumulation effect also need to be considered.

[0079] For a time sequence process, that is, when distributed photovoltaics are connected and photovoltaics are output, it is equivalent to injecting active power into the connected nodes during the photovoltaic output period, that is, changing a series of EBcs formed by multiple nodes in the network at multiple time points, and synthesizing OEBc, and then obtaining a distributed photovoltaic access plan based on OEBc.

[0080] Regarding the maximum PV access capacity constraints for grid connection: According to the State Grid Corporation's corporate standard "Technical Regulations for Photovoltaic Power Station Access to the Grid" (Q / GDW 617-2011), the total capacity of a small photovoltaic power station (sent through a 380V voltage level) should not exceed 25% of the maximum load in the power supply area of ​​the upper transformer in principle. There are three main reasons for this, which can be summarized as: grid stability, ensuring normal power supply, and technical considerations.

[0081] The stability of the power grid lies in the fact that photovoltaic power generation is greatly affected by weather and day and night. If the capacity is too large, it may randomly and periodically impact the operation of the power grid with large state fluctuations. Therefore, generally speaking, it is reasonable to impose a certain degree of capacity restriction.

[0082] Ensuring normal operation of power supply means that in modern power grids, when the power supply capacity and load capacity are at a ratio of 25%, the power grid will not collapse and the relevant protection actions will not be activated. Therefore, even when the photovoltaic power station is not generating electricity, the power grid can still operate.

[0083] Technical considerations include: Photovoltaic power generation is connected to the grid through power electronic inverters, which is prone to harmonics and three-phase current imbalance; the randomness of output power is prone to grid voltage fluctuations and flicker. Building photovoltaics are directly connected to the grid on the user side, and power quality issues directly affect the safety of users' electrical equipment. Photovoltaic output is active output. In order to avoid excessive increase in network frequency due to photovoltaic output, the access capacity is restricted to generally avoid affecting grid operation. By limiting the total PV access capacity, the solution can also be kept in a relatively stable range.

[0084] For the optimization algorithm:

[0085] From the mechanism analysis above, we can know that at a certain time point, P ％ There is a clear correlation with RBc. If the direction is specified and negative power is not considered, that is, when the load node sends power to the network in the reverse direction, P ％ It is positively correlated with EBc, that is, the smaller the EBc, the smaller the proportion of network loss. If the photovoltaic reverse power transmission is considered, the situation will become complicated and difficult to analyze. At this time, P ％ It does not show a perfect positive correlation with EBc, but the two should still have a general positive correlation.

[0086] Furthermore, for a timing situation, when the active power of different nodes changes due to active power injection of photovoltaic output, even if the change of electrical distance caused by thermal effect is not considered, there is still a node load P i There is a multi-node coupling effect on the network loss, transmission direction, and node voltage, which shows that P ％There is a problem with OEBc that it has strong nonlinear characteristics. Therefore, we should consider using a reliable solution algorithm that can solve nonlinear problems.

[0087] To this end, the optimization algorithm can be selected from the genetic algorithm, which is an evolutionary algorithm used to solve optimization and search problems. In a long period of research and application, it has been confirmed that it has the advantages of strong global optimization ability and good convergence. In order to verify the effectiveness of the optimization model under highly complex situations, this paper uses the genetic algorithm to solve it.

[0088] Specifically, to maximize OEBc, we set the fitness evaluation function to return the fitness of each individual and take the maximum value. The fitness calculation here directly uses the calculation method of OEBc. Taking the maximum fitness is to take maxOEBc for the OEBc of a series of individuals. To define the chromosome encoding, the possible PV capacity value set C is used as the range of individual genes: C = [0, allowed_cap], step (13) where allowed_cap is the maximum capacity of a single PV allowed, and step is a random step. At this time, the maximum capacity of each PV is within the closed interval between 0 and allowed_cap, and the value randomly obtained with step as the random step is used as the chromosome encoding of each individual, that is, the PV access plan.

[0089] Furthermore, individuals in the population are updated through the following population operation method.

[0090] First, selection method: select individuals according to fitness, and the selection probability is proportional to the fitness value, that is, In the formula, f(X i ) is the individual fitness of whether to select, f(X j ) is the fitness of each individual.

[0091] Second, the crossover method selects two parent individuals X i and X j , by randomly exchanging the genes of both parties to generate a new individual X' i and X' j .

[0092] Third, mutation method: modify the nodes in the gene that have been connected to PV to be unconnected, modify the nodes that have not been connected to PV to be connected, and select an element from the possible capacity value set C as the capacity value of the node.

[0093] Fourth, replacement method: replace the newly generated individuals into the population to form a new generation of population. When replacing, we can also consider using the elite retention strategy, that is, we can choose to retain a certain number of individuals with the highest fitness in the previous generation for a more aggressive solution. If not used, it is set to retain 0 elite individuals.

[0094] Finally, in order to complete the iteration, we also need to set a termination condition. The termination condition given in this paper is: the maximum number of iterations is used as the termination condition, that is, the evaluation of all individuals in the search space with the number of generations multiplied by the population size of each generation is completed.

[0095] In this application, network loss indicators and voltage quality are mainly used to determine whether the network status has been optimized.

[0096] 1. Network loss index

[0097] For the network after applying the obtained PV configuration, its total loss should be less than the total loss of the original network. This paper defines the network loss index as: Where P loss_new is the total network loss value after optimization, P loss_ori is the total loss value of the original network. When the percentage is lower, it shows that the obtained PV configuration scheme has a better effect on reducing network losses. The loss values ​​before and after will be obtained by performing time-series power flow calculations on the new and old networks.

[0098] 2. Voltage quality indicators

[0099] For a low-voltage distribution network, the most common voltage problem is voltage drop. In addition, due to the voltage increase caused by PV connection and PV output at the grid connection point, we should also consider that the voltage cannot be raised too high. At this time, we mainly consider whether the network voltage qualification rate can be improved.

[0100] The reason for using "total network loss" and "voltage quality" as network status indicators is that for low-voltage distribution networks, these two indicators are relatively important indicators. When judging the overall status of a network, the application of these two indicators is a general practice, and there is no need to be innovative here. (Since the voltage level of the low-voltage distribution network is low, the network loss is more obvious, so it is worth considering; since the low-voltage distribution network is close to the user's power supply, the requirement for ensuring voltage quality is more prominent.)

[0101] Since the access plan is obtained by solving the "optimal distribution model", if the network loss after the network applies the access plan is indeed lower than the original network, and the voltage quality is indeed better than the original network, then it can be said that the "optimal distribution model" that this article wants to declare can indeed improve the network status.

[0102] The proposed model is applied to the example network kerberdorfnetsz provided in this article. After solving the model, a distributed photovoltaic access plan is obtained. After applying the access plan to the example network, the total network loss and voltage quality before and after the network is connected to the distributed photovoltaic can be compared.

[0103] The “elia_pv_power” PV output data set provided by EILA was used, and k-means clustering was performed to obtain four types of typical daily PV output data: “medium-high”, “high”, “medium-low”, and “low”. At the same time, the “UMass Smart*Dataset-2017release” load data set provided by the University of Massachusetts randomly extracted load days with the same number of cluster types to match the typical days. The Kerber Dorfnetz network (its topology is shown in the attached figure) was used to generate the four typical daily PV output data sets. Figure 2 ) verification, using the PV grid connection point and the integrated capacity of the typical day calculation example shown in Table 1 (*Table 1 shows the PV configuration scheme obtained after optimization), using the parameters of the line model 24_AL1_4_ST1A shown in Table 2, the distributed photovoltaic configuration scheme shown in Table 1 is obtained, and Figure 3 , Figure 4 As shown, the results show that its state indicators are better, proving that the distributed photovoltaic access scheme obtained by the proposed model can optimize the network state.

[0104] Table 1

[0105]

[0106]

[0107] Table 2

[0108] parameter value r_ohm_per_km / resistance per kilometer 1.2012 x_ohm_per_km / Reactance per kilometer 0.335 max_i_ka / maximum current carrying capacity of the line 0.14 type / line type ol(overhead line) q_mm2 / circuit diameter 24

[0109] After clustering the PV data into typical annual days, that is, setting them into four types according to the output size: "medium-high", "high", "medium-low", and "low", the status indicators verified by the example are as follows Figure 3 and Figure 4 It can be seen that during the verified time period, the total network loss of the optimized network is significantly reduced and the voltage qualification rate is improved.

[0110] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present application can be implemented by general-purpose computer devices, they can be concentrated on a single computer device, or distributed on a network composed of multiple computer devices, optionally, they can be implemented by executable program codes of computer devices, so that they can be stored in a storage device and executed by the computer device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0111] It should be noted that the above are only preferred embodiments of the present application, and the patent protection scope of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for optimizing the distributed layout of a distribution network based on the zero load center of gravity, characterized in that: include: The node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node integrated into the network are used as decision variables of the model; The objective function is to maximize the center of gravity of the zero electrical load OEBc, and the constraints are set as voltage limit constraints based on voltage qualification rate, line current carrying capacity constraints, and maximum grid-connected PV access capacity constraints. Among them, the voltage qualification rate is the ratio of the voltage not exceeding the upper and lower limits within a given period of time. The voltage over-limit constraint based on the voltage qualification rate is that the optimized network voltage qualification rate should be greater than the voltage qualification rate of the original network. The line current carrying capacity constraint is that the line current carrying capacity is not greater than the maximum current carrying capacity allowed by the line used. The maximum PV access capacity constraint for grid connection is that the PV capacity of the distributed photovoltaic node connected to the network does not exceed 25% of the maximum load in the power supply area of ​​the upper transformer. Using the optimization algorithm, the node location of distributed photovoltaics and the PV capacity of each distributed photovoltaic node connected to the network are used as the decision variables of the model. On the basis of satisfying the voltage over-limit constraint, the line current carrying capacity constraint and the maximum PV access capacity constraint of the grid, the optimization calculation is performed with maximizing the zero electrical load gravity center OEBc as the optimization goal and the maximum number of iterations as the termination condition to obtain the PV access plan for the optimized network state, which is used to optimize the planning of the access location and access capacity of distributed photovoltaics in an existing network.

2. The method according to claim 1, characterized in that The maximum zero electrical load center of gravity OEBc is: maxOEBc, where:

3. The method according to claim 1, characterized in that The expression of voltage limit constraint based on voltage qualification rate is: Q net_new >Q net_ori , where Q net_new is the qualified rate of network voltage after optimization, Q net_ori is the network voltage qualification rate before optimization, where the node voltage limit range is defined as x1V nom ≤V i ≤x2V nom , where V i is the voltage amplitude of node i, V nom is the system nominal voltage, x1 and x2 are the given voltage upper and lower limits; the voltage qualification rate is defined as: In the formula, Q net is the network voltage qualification rate, Q i is the voltage qualification rate of node i, N is the number of nodes, T upper is the time when the corresponding node voltage exceeds the upper limit, T lower is the time when the corresponding node voltage exceeds the lower limit, T total The total time measured for the corresponding node.

4. The method according to claim 1, characterized in that: The expression of line current carrying capacity constraint is: lineload ≤I max_i_ka , where I lineload is the line current carrying capacity, I max_i_ka The maximum current carrying capacity allowed by the line used.

5. The method according to claim 1, characterized in that The expression of the maximum PV access capacity constraint for grid connection is: S t_pv ≤0.25S trans , where S t_pv is the total PV capacity, S trans is the transformer capacity.

6. The method according to claim 1, characterized in that The optimization algorithm uses a genetic algorithm. In order to maximize OEBc, a fitness evaluation function is set to return the fitness of each individual and take the maximum value. To define the chromosome encoding, the possible PV capacity value set C is used as the range of the individual gene: C = [0, allowed_cap], step (13) where allowed_cap is the maximum capacity of a single PV allowed, and step is a random step. At this time, the maximum capacity of each PV is within the closed interval from 0 to allowed_cap, and the value randomly obtained with step as the random step is used as the chromosome encoding of each individual, that is, the PV access plan.

7. The method according to claim 1, characterized in that After obtaining the PV access solution that optimizes the network status, it also includes: Use network loss index and network voltage qualification rate index to judge the optimization effect of the network before and after optimization; Among them, the network loss index is defined as: Where P loss_new is the total network loss value after optimization, P loss_ori is the total loss value of the original network, and the loss values ​​before and after will be obtained by performing time series power flow calculation on the new and old networks; The network voltage qualification rate is defined as: In the formula, Q net is the network voltage qualification rate, Q i is the voltage qualification rate of node i, N is the number of nodes, T upper is the time when the corresponding node voltage exceeds the upper limit, T lower is the time when the corresponding node voltage exceeds the lower limit, T total The total time measured for the corresponding node.

8. The method according to claim 6, characterized in that The termination condition of iterating to the maximum number of times means that all individuals in the search space with the number of generations multiplied by the population size of each generation are evaluated.

Images