Sub-network scheme index evaluation method suitable for parallel simulation of power distribution system

Abstract

The invention discloses a sub-network scheme index evaluation method suitable for parallel simulation of a power distribution system, and the method comprises the steps: S1, quantifying the simulation calculation resource distribution of a sub-network, and defining a sub-network imbalance index; s2, considering the time consumption difference brought by different power electronic switch models, and proposing a calculation complexity valuing method of each branch and node in the power distribution system; and S3, establishing an IEEE34 node test system comprising a photovoltaic power generation system, and verifying the effectiveness of the sub-network imbalance index. A simulation result proves that the index effectively reflects simulation time consumption under different sub-network strategies, so that an index reference is provided for an electromagnetic transient parallel simulation sub-network scheme.

Description

Technical Field

[0001] The invention belongs to the field of electromagnetic transient simulation of power systems, and in particular, relates to a network division scheme index evaluation method suitable for parallel simulation of power distribution systems. Background Art

[0002] With the access of a large number of distributed power sources and power electronic devices, the "double high" characteristics of the distribution network are becoming more and more obvious, which increases the computational burden faced by the electromagnetic transient simulation of the system. To deal with this problem, parallel simulation has become an effective solution. This method divides the complex power system into multiple subsystems and distributes them to different CPU cores for parallel solution, thereby improving computational efficiency.

[0003] When performing subnet division for parallel simulation of distribution networks, reasonable load distribution is the key to ensuring simulation efficiency. Different types of new energy equipment and their control parts have significant differences in simulation calculation volume. Therefore, it is of great research value to reasonably introduce this factor into the balance evaluation index of the subnet scheme to improve the overall performance of parallel simulation. Summary of the invention

[0004] The present invention aims at the problems in the prior art and provides a network division scheme index evaluation method suitable for parallel simulation of distribution systems. It fully considers the differences in simulation calculation amounts under different power electronic switch models, quantifies the solution complexity based on the number of states of the control system, and proposes a network division imbalance index, which provides an effective balance evaluation index for network division schemes in distribution network scenarios.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A method for evaluating network sub-scheme indicators for parallel simulation of power distribution systems includes the following steps:

[0007] S1, quantify the simulation computing resource distribution of the sub-network and define the sub-network imbalance index;

[0008] S2, considering the time consumption differences caused by different power electronic switch models, a method for determining the computational complexity of each branch and node in the distribution system is proposed;

[0009] S3, build an IEEE34 node test system including a photovoltaic power generation system to verify the effectiveness of the sub-grid imbalance indicator.

[0010] Preferably, in step S1, the simulation computing resource distribution of the subnetwork is quantified according to an electromagnetic transient analysis method; the electromagnetic transient analysis method uses the node voltage as a state variable, discretizes the component-level model of the power distribution system to form a differential equation, and combines the differential equations into the following algebraic equations:

[0011] Gu=i

[0012] Wherein, G is the nodal admittance matrix of the distribution system, u is the nodal voltage vector, and i is the equivalent injected current source.

[0013] Preferably, in the step S1, the quantization of the simulation computing resources is represented by the following formula:

[0014]

[0015] Wherein, H is the simulation step size of the distribution system, m, n, and d are respectively the total number of branch circuits, the number of conventional nodes, and the number of new nodes in the distribution system, α i is the complexity value of the i-th branch circuit, is the correction term brought by the cascade of the total branch circuits, β j is the complexity value of the j-th conventional node, γ k is the complexity value of the k-th new node, and s is the complexity value of the control system in the new nodes.

[0016] Preferably, in the step S1, the distribution system is divided into 2 sub-networks, and the unbalance index θ of the network division is calculated by the following formula:

[0017]

[0018] Wherein, is the average value of the simulation computing amounts R p and R q of the sub-networks.

[0019] Preferably, in the step S2, the complexity values α i and β j of each branch circuit and traditional node are determined based on the number of independent nodes included in the component.

[0020] Preferably, in the step S2, the computational complexity s of the control part is determined by the number of states it contains.

[0021] Preferably, in the step S3, in the established IEEE34-node test system, calculate the network division unbalance indexes at different network division points, and perform parallel simulations respectively to verify the effectiveness of the network division unbalance indexes.

[0022] Advantages of the present invention: The present invention introduces the influence of the differences in the power electronic switch models in new energy devices and the control system on the simulation computing amount into the balance evaluation index of the network division scheme, which helps to construct a more accurate evaluation system for the distribution network division scheme, thereby giving full play to the advantages of parallel simulation. Description of the Drawings

[0023] Figure 1 It is a flow chart of S1 to S3 in the present invention.

[0024] Figure 2 It is a topological diagram of a single-stage photovoltaic power generation system in the present invention.

[0025] Figure 3 It is the improved IEEE34 node system topology diagram in the present invention.

[0026] Figure 4 This is a diagram of simulation results in Example 2 of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is described in detail in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation cases described herein are only used to explain the present invention and are not intended to limit the invention. The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments of the specification:

[0028] Embodiment 1:

[0029] An evaluation index for network partitioning schemes suitable for parallel simulation of power distribution systems, such as Figure 1 As shown, the following steps are included:

[0030] S1, quantify the simulation computing resource distribution of the sub-network according to the electromagnetic transient analysis method and define the sub-network imbalance index;

[0031] S1.1, considering that electromagnetic transient simulation is essentially to solve a set of mathematical equations related to the state variables of the system. Considering the simulation efficiency, most simulation solving programs are currently established under the framework of node analysis method.

[0032] This method uses node voltage as the state variable, discretizes the component-level model to form a differential equation, and combines the differential equations into the following algebraic equations:

[0033] Gu=i

[0034] Where G is the node admittance matrix of the system, u is the node voltage vector, and i is the equivalent injection current source.

[0035] S1.2, since the order of the equations formed by the node voltage method is closely related to the number of nodes in the system, and the distribution system usually also includes a control part, based on this, the quantification of simulation computing resources can be expressed by the following formula:

[0036]

[0037] Wherein, H is the simulation step size of the system, m, n, and d are the total number of branches, the number of conventional nodes, and the number of new nodes of the system respectively, and α i is the complexity value of the i-th branch, is the correction term brought by the cascade of the total branches, and β j is the complexity value of the j-th conventional node, and γ k is the complexity value of the k-th new node, and s is the complexity value of the control system in the new nodes.

[0038] S1.3. Assume that a large power distribution system is divided into 2 sub-networks. To reflect the difference in the calculation resource allocation between the two, define the unbalance index of the sub-networks:

[0039]

[0040] Wherein, is the average value of the simulation calculation amounts R p and R q of the 2 sub-networks.

[0041] S2. Consider the time-consuming difference brought by different power electronic switch models, and propose a method for determining the calculation complexity values of each branch and node;

[0042] S2.1. The complexity values of each branch and node are all based on the number of independent nodes included in the component. However, it should be noted that since the new nodes usually contain a large number of power electronic devices, the power electronic switch models therein will occupy more computing resources. For this reason, taking the single-stage photovoltaic power generation unit as an example, introduce the complexity values under its different switch models.

[0043] The typical topology of a single-stage photovoltaic power generation system is as Figure 2 shown. The common power electronic switch modeling methods mainly include the binary resistance model and the constant admittance model. The main difference between the two is that the binary resistance model needs to frequently update the node admittance matrix during the simulation, while the constant admittance model does not. According to experience, usually set the simulation calculation amount of the binary resistance model to 4 times that of the constant admittance model to reflect this influence. It can be seen from this that under the constant admittance model, the complexity value γ k of the single-stage photovoltaic power generation unit is 8, while under the binary resistance model, its complexity increases to 32.

[0044] S2.2. Since the order of the control system is usually directly related to the number of its state variables, the calculation complexity s of the control part can be determined by the number of states it contains.

[0045] S3. Build an IEEE 34-node test system including a photovoltaic power generation system to verify the effectiveness of the proposed method.

[0046] S3.1. In the improved IEEE 34 - node system established, calculate the index values at different sub - network points and conduct parallel simulations respectively to verify the effectiveness of the present invention.

[0047] Embodiment 2:

[0048] The distribution network established in this embodiment is an IEEE 34 - node system with a voltage level of 24.9 kV, and a photovoltaic power generation system is connected at node 840. The overall topological structure of the distribution network is as Figure 3 shown. Parallel simulations are respectively conducted for the following three sub - network schemes: 1) Sub - network at nodes 852 - 832; 2) Sub - network at nodes 850 - 816; 3) Sub - network at nodes 834 - 842.

[0049] Set two scenarios with simulation step sizes H = 10 μs and H = 5 μs, and adopt a fixed - step discrete solver. The simulation time is 2 s. The evaluation indicators under the three sub - network schemes and the measured simulation time consumption are shown in Table 1.

[0050] Table 1 Unbalance indicators of sub - network schemes and their parallel simulation time consumption

[0051]

[0052] In this embodiment, the corresponding relationship between the index values and the measured simulation time consumption is as Figure 4 shown. It can be seen that the higher the unbalance index value, the longer the time spent on parallel simulation. This is because the evaluation index proposed by the present invention can reflect the distribution of computing resources among sub - tasks. The higher the index value, the more uneven the load distribution, resulting in an increase in the synchronization waiting time during the parallel simulation process and affecting the simulation efficiency of the system.

[0053] In summary, the present invention provides an effective evaluation index for the sub - network scheme of the parallel simulation of a new - type distribution system, thereby providing a judgment basis for improving the parallel simulation performance of the system.

Claims

1. A method for evaluating the index of a network division scheme applicable to parallel simulation of a distribution system, characterized in that The steps include: S1, quantify the simulation computing resource distribution of the sub-network and define the sub-network imbalance index; S2, considering the time consumption differences caused by different power electronic switch models, a method for determining the computational complexity of each branch and node in the distribution system is proposed; S3, build an IEEE34 node test system including a photovoltaic power generation system to verify the effectiveness of the sub-grid imbalance indicator.

2. The index evaluation method for subnetting schemes applicable to parallel simulation of power distribution systems according to claim 1, characterized in that In the step S1, the simulation computing resource distribution of the sub-network is quantified according to the electromagnetic transient analysis method; The electromagnetic transient analysis method uses node voltage as the state variable, discretizes the component-level model of the distribution system to form a differential equation, and combines the differential equations into the following algebraic equations: Gu=i Where G is the node admittance matrix of the distribution system, u is the node voltage vector, and i is the equivalent injection current source.

3. The index evaluation method for subnetting scheme applicable to parallel simulation of distribution system according to claim 2, wherein In step S1, the quantification of simulation computing resources is expressed by the following formula: where H is the simulation step size of the power distribution system, m, n, and d are the total number of branch circuits, the number of conventional nodes, and the number of new nodes in the power distribution system, respectively, and α i is the complexity value of the i-th branch circuit, is the correction term brought by cascading in the total branch circuit, and β j is the complexity value of the j-th conventional node, and γ k is the complexity value of the k-th new node, and s is the complexity value of the control system in the new node.

4. The index evaluation method for subnetting scheme applicable to parallel simulation of distribution system according to claim 3, wherein In step S1, the power distribution system is divided into two sub-networks, and the unbalance index θ of the sub-networks is calculated by the following formula: In the formula, is the simulation calculation amount R of 2 sub-networks p and R q is the average value.

5. The index evaluation method for the subnetting scheme applicable to parallel simulation of the power distribution system according to claim 4, wherein In the step S2, the complexity values α i and β j of each branch and traditional node are determined based on the number of independent nodes included in the component.

6. The index evaluation method for subnetting scheme applicable to parallel simulation of distribution system according to claim 5, characterized in that In step S2, the computational complexity s of the control part is determined by the number of states it contains.

7. The index evaluation method for subnetting scheme applicable to parallel simulation of distribution system according to claim 6, wherein In the step S3, in the constructed IEEE 34-node test system, the network imbalance index at different network points is calculated, and parallel simulations are performed respectively to verify the validity of the network imbalance index.

Images