Power distribution network bearing capacity evaluation method based on Bayesian network

Abstract

The invention discloses a power distribution network bearing capacity assessment method based on a Bayesian network, and belongs to the field of power distribution networks, and the method comprises the following steps: S1, collecting power grid structure data, load data and distributed power supply data in a power distribution network to be assessed; s2, based on the power grid structure data, the load data and the distributed power supply data in the power distribution network acquired in the step S1, processing uncertainty factors in the power distribution network by using a Bayesian network model, and generating probability distribution; s3, based on the probability distribution in the step S2, aiming at different constraint conditions, utilizing a bearing capacity optimization model to carry out segmented solution; and S4, performing simulation verification. By adopting the power distribution network bearing capacity evaluation method based on the Bayesian network, the output condition of the distributed power supply can be predicted more accurately, so that resources are better managed and scheduled, and the stability and efficiency of the whole power grid are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network, and in particular to a distribution network carrying capacity evaluation method based on Bayesian network. Background Art

[0002] The operation of the distribution network covers many aspects such as grid security, reliable power supply and green economy. Its carrying capacity comprehensively reflects the effects of planning, operation strategy, scheduling and real-time control.

[0003] When a large number of new energy sources (such as solar energy and wind energy) are connected to the power system, their output is intermittent and uncertain because the production of these energy sources is highly dependent on factors such as weather conditions. For example, when there is sufficient sunlight or strong wind, there may be a surplus of electricity; on the contrary, on cloudy and windless days, there may be a risk of power shortage. Therefore, the large-scale integration of new energy generation has brought new challenges to maintaining the balance between power supply and demand and ensuring system stability.

[0004] To address the above problems, SSA Mohammad, J Ma, D Zhang, AS Ahmadyar and H Marzooghi proposed a daily economic dispatch model for an electrical multi-energy complementary system considering distributed renewable energy and backup energy types in the paper "Probabilistic Assessment of Hosting Capacity in Radial Distribution Systems", and adopted a novel second-order cone decomposition method of the Weymouth equation to solve the non-convex optimization problem.

[0005] At the same time, El-Shimy M, Sharaf A, Khairy H et al. designed a new strategy for coordinated scheduling of electric-thermal multi-energy complementary systems based on the heterogeneous characteristics between the power and thermal subsystems in "Reduced-order modeling of solar-PV generators for small-signal stability assessment of power systems and estimation of maximum penetration levels" using a heterogeneous decomposition algorithm.

[0006] However, with the continuous growth of the installed capacity of distributed renewable energy in power generation systems, traditional deterministic optimization evaluation methods are becoming increasingly inadequate to meet the requirements in dealing with this change, and their reliability is also declining. Summary of the invention

[0007] The purpose of the present invention is to provide a distribution network carrying capacity evaluation method based on Bayesian network to solve the above technical problems.

[0008] To achieve the above object, the present invention provides a method for evaluating the carrying capacity of a distribution network based on a Bayesian network, comprising the following steps:

[0009] S1. Collect grid structure data, load data and distributed power source data in the distribution network to be evaluated;

[0010] S2, based on the grid structure data, load data and distributed power data in the distribution network collected in step S1, using the Bayesian network model to process the uncertainty factors in the distribution network and generate a probability distribution;

[0011] S3, based on the probability distribution described in step S2, for different constraints, using the carrying capacity optimization model to adopt the enumeration method to solve the problem piecewise;

[0012] S4. Simulation verification.

[0013] Preferably, in step S1, the load data includes historical load data and predicted future load data.

[0014] Preferably, in step S2, the uncertainty factor data in the distribution network includes load variation, distributed energy output fluctuation, and line load transfer rate.

[0015] Preferably, in step S2, a Bayesian network is constructed to model the uncertain factor data, and parameters are learned from historical load data to predict probability distribution under various conditions.

[0016] Preferably, step S2 specifically includes the following steps:

[0017] S21. Use the maximum likelihood estimation method to learn the Bayesian network parameters. The maximum likelihood estimation method calculation formula is as follows:

[0018] L(θ|S′0 G)=P(S|θ,G) (1);

[0019] Where L(θ|S′0 G) represents the likelihood function of the model parameter θ given the data S′0 and the Bayesian network structure G; P(S|θ,G) represents the probability of the sample data set S composed of uncertainty factor data occurring given the model parameter θ and the Bayesian network structure G;

[0020] S22. Take the logarithm of θ|S′0 G and transform the problem into solving the maximum value of the log-likelihood function:

[0021] l(θS,G)=logL(θS,G)=logP(S∣θ,G) (2);

[0022] Where l(θS,G) represents the log-likelihood function;

[0023] S23. Assuming that the sample data set S is independent and distributed, according to the structural characteristics of the Bayesian network:

[0024]

[0025] In the formula, Represents the number of instances that satisfy the sample data set S; P(S″ t ,G) represents the probability of the tth instance in the sample data set S under the given Bayesian network structure G;

[0026] The maximum likelihood estimation expression is as follows:

[0027] θ jk =N jk / N j (4);

[0028] In the formula, θ jk N represents the probability of the jth variable in the sample data set S under the kth state; jk N represents the number of times the jth variable in the sample data set S appears in the kth state; j Represents the total number of times the jth variable appears in the sample data set S.

[0029] Preferably, in step S3, a carrying capacity optimization model is constructed based on multiple constraints, resources are optimally configured using the carrying capacity optimization model, and an optimal solution is searched by defining an objective function.

[0030] Preferably, step S3 specifically includes the following steps:

[0031] S31. Construct the objective function of the carrying capacity optimization model: Set the optimization goal as the maximum carrying capacity evaluation index that meets multiple constraints:

[0032]

[0033] In the formula, C n (m) represents the maximum load capacity of operation mode m; It represents the maximum access capacity of distributed generation at node i under operation mode m;

[0034] S32. Constructing the power balance equation of the load-bearing capacity optimization model:

[0035]

[0036] In the formula, and They represent the maximum active power and reactive power injected by the distributed generation at node i respectively; P 1,1 and Q 1,1 Respectively represent the active power and reactive power of node 1 load; G j , B j and S j They represent the conductance, susceptance and phase angle of node j respectively; S i represents the phase angle of node i; B ij represents the susceptance between node i and node j; S ij represents the phase difference between node i and node j; U j represents the voltage at node j; U i represents the voltage of node i; Q represents the set of all nodes.

[0037] Therefore, the present invention adopts the above-mentioned distribution network carrying capacity evaluation method based on Bayesian network, which has the following beneficial effects:

[0038] A multi-state output model of distributed generation is established using a multi-scenario probability analysis method, and the impact of load changes, distributed generation fluctuations and line load transfer rate on outage time and outage cost are considered. It has the advantages of fast, stable and efficient algorithm.

[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A flowchart of a method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to the present invention;

[0041] Figure 2 This is a distributed network structure diagram of the simulation experiment of the present invention. DETAILED DESCRIPTION

[0042] In the description of the present invention, it should be noted that the terms "upper", "lower", "inside", "outside", etc. indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, or the positions or positional relationships in which the invented product is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In the description of the present invention, it should also be noted that, unless otherwise clearly specified and limited, the terms "setting", "installation", and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0043] The embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0044] like Figure 1 As shown, a distribution network carrying capacity evaluation method based on Bayesian network includes the following steps:

[0045] S1. Collect grid structure data, load data and distributed power source data in the distribution network to be evaluated;

[0046] In step S1, the load data includes historical load data and predicted future load data.

[0047] In this embodiment, the grid structure data is obtained in the following manner: the physical topology information of the grid is obtained through the existing power system database or GIS (geographic information system). The physical topology information includes nodes (load points), branch circuits, contact switches, etc., and the connection relationship between each node and line, as well as the specific location and attributes of each node are obtained.

[0048] The load data (power consumption of each node in different time periods) is obtained as follows: the load data of each node is collected from historical records or real-time monitoring systems. The above load can come from SCADA (supervisory control and data acquisition system) or other monitoring equipment of the power system.

[0049] The distributed power source data is obtained in the following ways: through field measurements, historical data records or prediction models to obtain the location, type (such as solar energy, wind energy, etc.), capacity, output characteristics and other information of the distributed power source.

[0050] In this embodiment, other data such as meteorological data, such as light intensity and wind speed, can also be collected to facilitate the prediction of photovoltaic and wind power output, and system parameters, such as line impedance and transformer parameters, can be obtained from design drawings or system configuration files.

[0051] In this embodiment, the acquired data is preprocessed as follows: first, the collected data is cleaned and formatted to ensure the consistency and accuracy of the data, and then missing values ​​and outliers are processed to ensure data quality.

[0052] S2. Based on the grid structure data, load data and distributed power source data in the distribution network collected in step S1, the Bayesian network model is used to process the uncertainty factors in the distribution network and generate a probability distribution. Thus, the Bayesian network model can be used to define node variables and their probability distribution, establish dependencies between nodes, and then calculate the expected output and load of each node through probabilistic reasoning;

[0053] In step S2, the uncertainty factor data in the distribution network includes load changes, distributed energy output fluctuations, and line load transfer rates.

[0054] In step S2, the uncertain factor data is modeled by constructing a Bayesian network, and parameters are learned from historical load data to predict the probability distribution under various conditions.

[0055] Step S2 specifically includes the following steps:

[0056] S21. Use the maximum likelihood estimation method to learn the Bayesian network parameters. The maximum likelihood estimation method calculation formula is as follows:

[0057] L(θ|S′0 G)=P(S|θ,G) (1);

[0058] Where L(θ|S0′ G) represents the likelihood function of the model parameter θ given the data S′0 and the Bayesian network structure G; P(S|θ,G) represents the probability of the sample data set S consisting of uncertainty factor data occurring given the model parameter θ and the Bayesian network structure G;

[0059] S22. Take the logarithm of θ|S′0 G and transform the problem into solving the maximum value of the log-likelihood function:

[0060] l(θS,G)=logL(θS,G)=logP(S∣θ,G) (2);

[0061] Where l(θS,G) represents the log-likelihood function;

[0062] S23. Assuming that the sample data set S is independent and distributed, according to the structural characteristics of the Bayesian network:

[0063]

[0064] In the formula, represents the number of instances that satisfy the sample data set S; P(S t ,G) represents the probability of the tth instance in the sample data set S under the given Bayesian network structure G;

[0065] The maximum likelihood estimation expression is as follows:

[0066] θ jk =N jk / N j (4);

[0067] In the formula, θ jk N represents the probability of the jth variable in the sample data set S under the kth state; jk N represents the number of times the jth variable in the sample data set S appears in the kth state; j Represents the total number of times the jth variable appears in the sample data set S.

[0068] S3, based on the probability distribution described in step S2, for different constraints, using the carrying capacity optimization model to adopt the enumeration method to solve the problem piecewise;

[0069] In step S3, a carrying capacity optimization model is constructed based on multiple constraints, resources are optimally configured using the carrying capacity optimization model, and an objective function is defined to search for the optimal solution.

[0070] Step S3 specifically includes the following steps:

[0071] S31. Construct the objective function of the carrying capacity optimization model: Set the optimization goal as the maximum carrying capacity evaluation index that meets multiple constraints:

[0072]

[0073] In the formula, C n (m) represents the maximum load capacity of operation mode m; It represents the maximum access capacity of distributed generation at node i under operation mode m;

[0074] S32. Constructing the power balance equation of the load-bearing capacity optimization model:

[0075]

[0076] In the formula, and They represent the maximum active power and reactive power injected by the distributed generation at node i respectively; P 1,1 and Q 1,1 Respectively represent the active power and reactive power of node 1 load; G j , B j and S j They represent the conductance, susceptance and phase angle of node j respectively; S i represents the phase angle of node i; B ij represents the susceptance between node i and node j; S ij represents the phase difference between node i and node j; U j represents the voltage at node j; U i represents the voltage of node i; Q represents the set of all nodes.

[0077] S4. Simulation verification.

[0078] In this embodiment, the Figure 2 The 20-node double-fed radial 20kV line branch is shown, and the line has five distributed power sources, of which 3, 5, 7 and 10 have four crew members respectively, the voltage fluctuation limits (H and L) are +7% and -7% respectively, the voltage fluctuation limit DMAX is 3%, and the maximum three-phase short-circuit current limit IDMAX is 20KA.

[0079] Experimental Results

[0080] When the distributed generation resources are large, if the planning is carried out according to the maximum possible capacity of distributed generation, it is likely to exceed the constraints. Assuming that the maximum possible capacity of distributed generation at each node is 1.3 times the corresponding value, and all distributed generation is connected to the grid according to its maximum possible capacity, the three constraints are out of bounds: the maximum voltage occurs at node 21 (22.46kV), the maximum voltage fluctuation occurs at node 14 (3.826%), and the maximum short-circuit current occurs at the transformer low-voltage bus (21.01kA). When using the multi-constraint forward and backward methods described in the present invention, all constraints must be considered, and the results are shown in Table 1.

[0081] Table 1 Maximum carrying capacity of distributed power generation / MVA

[0082]

[0083] As can be seen from Table 1, the maximum bearing capacities obtained by the multi-constraint forward and reverse methods are similar: the differences between the single-constraint method and the multi-constraint method are 4.25% and 3.47%; the maximum bearing capacity obtained by the multi-constraint method is 7.5%-8.6% higher than that of the single-constraint method, thus verifying the effectiveness of the present invention.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A distribution network carrying capacity evaluation method based on Bayesian network, characterized by: The following steps are involved: S1. Collect grid structure data, load data and distributed power source data in the distribution network to be evaluated; S2, based on the grid structure data, load data and distributed power data in the distribution network collected in step S1, using the Bayesian network model to process the uncertainty factors in the distribution network and generate a probability distribution; S3, based on the probability distribution described in step S2, for different constraints, using the carrying capacity optimization model to adopt the enumeration method to solve the problem piecewise; S4. Simulation verification.

2. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 1, characterized in that: In step S1, the load data includes historical load data and predicted future load data.

3. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 2, characterized in that: In step S2, the uncertainty factor data in the distribution network includes load changes, distributed energy output fluctuations, and line load transfer rates.

4. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 3 is characterized in that: In step S2, the uncertain factor data is modeled by constructing a Bayesian network, and parameters are learned from historical load data to predict the probability distribution under various conditions.

5. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 4 is characterized in that: Step S2 specifically includes the following steps: S21. Use the maximum likelihood estimation method to learn the Bayesian network parameters. The maximum likelihood estimation method calculation formula is as follows: L(θ|S′0G)=P(S|θ,G) (1); Where L(θ|S′0G) represents the likelihood function of the model parameter θ given the data S′0 and the Bayesian network structure G; P(S|θ,G) represents the probability of the sample data set S composed of uncertainty factor data occurring given the model parameter θ and the Bayesian network structure G; S22. Take the logarithm of θ|S′0G and transform the problem into solving the maximum value of the log-likelihood function: l(θS,G)=logL(θS,G)=logP(S∣θ,G) (2); Where l(θS,G) represents the log-likelihood function; S23. Assuming that the sample data set S is independent and distributed, according to the structural characteristics of the Bayesian network: In the formula, Represents the number of instances that satisfy the sample data set S; P(S″ t ,G) represents the probability of the tth instance in the sample data set S under the given Bayesian network structure G; The maximum likelihood estimation expression is as follows: i jk =N jk / N j (4); In the formula, θ jk N represents the probability of the jth variable in the sample data set S under the kth state; jk N represents the number of times the jth variable in the sample data set S appears in the kth state; j Represents the total number of times the jth variable appears in the sample data set S.

6. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 5, characterized in that: In step S3, a carrying capacity optimization model is constructed based on multiple constraints, resources are optimally configured using the carrying capacity optimization model, and an objective function is defined to search for the optimal solution.

7. The method for evaluating the carrying capacity of a distribution network based on a Bayesian network according to claim 6, characterized in that: Step S3 specifically includes the following steps: S31. Construct the objective function of the carrying capacity optimization model: Set the optimization goal as the maximum carrying capacity evaluation index that meets multiple constraints: In the formula, C n (m) represents the maximum load capacity of operation mode m; It represents the maximum access capacity of distributed generation at node i under operation mode m; S32. Constructing the power balance equation of the load-bearing capacity optimization model: In the formula, and They represent the maximum active power and reactive power injected by the distributed generation at node i respectively; P 1,1 and Q 1,1 Respectively represent the active power and reactive power of node 1 load; G j , B j and S j They represent the conductance, susceptance and phase angle of node j respectively; S i represents the phase angle of node i; B ij represents the susceptance between node i and node j; S ij represents the phase difference between node i and node j; U j represents the voltage at node j; U i represents the voltage of node i; Q represents the set of all nodes.

Images