CN-122001015-A - Power distribution network operation method and device based on reconstruction and cluster collaborative optimization, electronic equipment and storage medium

CN122001015ACN 122001015 ACN122001015 ACN 122001015ACN-122001015-A

Abstract

The invention discloses a power distribution network operation method, a device, electronic equipment and a storage medium based on reconstruction and cluster collaborative optimization, and belongs to the field of power grid operation optimization. And constructing a collaborative optimization model according to the initial data. And (3) through iterative optimization operation until the relative deviation is smaller than a convergence threshold value, obtaining the optimal network topology and photovoltaic output. In each round of optimization, a network reconstruction model is solved according to the current photovoltaic output to generate the current optimal network topology, the photovoltaic power stations are clustered, and the collaborative optimization model is solved to obtain the optimal power output of each photovoltaic power station. And calculating the difference value of the operation indexes of the whole network, generating relative deviation, and judging whether convergence is achieved. By implementing the method and the device, the problem that the optimization result in the prior art is in local optimum can be solved.

Inventors

LU JIANGANG
ZHAO RUIFENG
GUO WENXIN
LI HAOBIN
LI SHIMING

Assignees

广东电网有限责任公司电力调度控制中心

Dates

Publication Date: 20260508
Application Date: 20260122

Claims (10)

1. The power distribution network operation method based on reconstruction and cluster collaborative optimization is characterized by comprising the following steps of: acquiring initial network topology, initial photovoltaic output, branch parameters and load of a power distribution network; according to the branch parameters and the load, a network reconstruction model is built by taking the optimization of the whole network operation index as a target; constructing a collaborative optimization model by taking the optimization of operation indexes in a cluster as a target according to the initial network topology, the initial photovoltaic output, the branch parameters and the load; Repeatedly executing iterative optimization operation until the relative deviation is smaller than a preset convergence threshold value, and generating a target network topology and a target photovoltaic output; Scheduling the power distribution network according to the target network topology and the target photovoltaic output; The iterative optimization operation comprises the steps of solving a network reconstruction model according to current photovoltaic output to generate a current optimal network topology and a current whole network operation index, wherein the initial photovoltaic output is the initial photovoltaic output; according to the current optimal network topology, all photovoltaic power stations in the power distribution network are clustered to generate a photovoltaic cluster dividing result, and a leading node is determined for each photovoltaic cluster by analyzing the voltage sensitivity relationship among the photovoltaic power stations in each photovoltaic cluster; solving the collaborative optimization model according to the current optimal network topology, the dominant nodes and the dividing result of the photovoltaic clusters to generate optimal power output values of all photovoltaic power stations, and combining the optimal power output values of all photovoltaic power stations in the current round to generate the current optimal photovoltaic output; calculating the difference between the current full-network operation index and the full-network operation index of the previous round to generate a relative deviation; and judging whether the relative deviation is smaller than a preset convergence threshold, if so, recording the current optimal photovoltaic output as target photovoltaic output, recording the current optimal network topology as target network topology, and if not, updating the current optimal photovoltaic output as current photovoltaic output.
2. The power distribution network operation method based on reconstruction and cluster collaborative optimization according to claim 1, wherein the constructing a network reconstruction model according to the branch parameters and the load and with the optimization of the whole network operation index as a target comprises: constructing a voltage deviation term according to the difference relation between the voltage of each node and a preset first reference voltage, wherein the nodes are electrical connection points in the power distribution network, and comprise load nodes and grid-connected nodes where photovoltaic power stations are located; according to the branch parameters, an active network loss item is constructed by calculating the sum of active power losses generated by currents flowing through all branches of the whole network on corresponding branch resistances; The voltage deviation item and the active network loss item are subjected to weighted combination to generate a whole network economic and safe operation objective function; Constructing a load flow balance constraint according to the branch parameters and the load; constructing a radial network structure constraint of a power grid according to the relation between nodes and branches contained in the initial network topology; Constructing voltage and current safety constraint according to a preset power grid safety operation limit value; And generating a network reconstruction model according to the whole-network economic and safe operation objective function, the tide balance constraint, the power grid radial network structure constraint and the voltage and current safety constraint.
3. The power distribution network operation method based on reconstruction and cluster collaborative optimization according to claim 2, wherein constructing a collaborative optimization model according to an initial network topology, an initial photovoltaic output, a branch parameter and a load and with an operation index in a cluster as an optimal target comprises: according to the difference relation between the voltage of each cluster leading node and the preset second reference voltage, a leading node voltage deviation term is constructed; according to the branch parameters, active power loss of the branch in each photovoltaic cluster is calculated respectively, and loss of all photovoltaic clusters is summed, so that a cluster-level total network loss item is constructed; performing weighted combination on the voltage deviation item of the leading node and the total network loss item of the cluster level to construct an inter-cluster collaborative optimization objective function; According to the difference relation between the voltage of all nodes in each photovoltaic cluster and a preset third reference voltage, an internal node voltage deviation term is built; according to the branch parameters, an intra-cluster network loss item is constructed by calculating the sum of active power losses of all branches in a single photovoltaic cluster; The internal node voltage deviation item and the cluster internal network loss item are subjected to weighted combination to generate an intra-cluster autonomous optimization objective function; constructing a photovoltaic power station power constraint according to the initial photovoltaic output; and generating a collaborative optimization model according to the inter-cluster collaborative optimization objective function, the intra-cluster autonomous optimization objective function, the power flow balance constraint, the voltage and current safety constraint, the photovoltaic power station power constraint and a preset inter-cluster collaborative control constraint.
4. The method for operating a power distribution network based on collaborative optimization of reconstruction and clustering according to claim 3, wherein the performing cluster division on all photovoltaic power stations in the power distribution network according to the current optimal network topology to generate a division result of the photovoltaic clusters comprises: calculating a reactive voltage sensitivity matrix according to the current optimal network topology; according to the reactive voltage sensitivity matrix, calculating the electrical distance between each photovoltaic power station in the power distribution network; Constructing a cluster division evaluation function taking modularity maximization as a target according to the electrical distance; and solving the cluster division evaluation function to generate a division result of the photovoltaic clusters.
5. The method for operating a power distribution network based on collaborative optimization of reconstruction and clustering according to claim 4, wherein the calculating a reactive voltage sensitivity matrix according to the current optimal network topology includes: carrying out power flow calculation on the current optimal network topology to generate a corresponding jacobian matrix; Decomposing the jacobian matrix to generate an active power equation set and a reactive power equation set; And calculating and generating a reactive voltage sensitivity matrix according to the active power equation set and the reactive power equation set.
6. The power distribution network operation method based on reconstruction and cluster collaborative optimization according to claim 5, wherein the cluster partition evaluation function specifically comprises: in the formula, Is a modularity index; Is a node Sum node Weighting of the edges; half the sum of the weights of all edges in the network; For all and nodes The sum of the weights of the connected edges; For all and nodes The sum of the weights of the connected edges; as a function, when the node And node The value is 1 in the same cluster, and the value is 0 in the opposite way.
7. The method of operation of a power distribution network based on collaborative optimization of reconstruction and clustering according to claim 6, wherein a dominant node is determined for each photovoltaic cluster by: Aiming at each photovoltaic cluster, calculating the observability index of each photovoltaic power station according to the voltage sensitivity among the photovoltaic power stations in the current photovoltaic cluster; calculating the controllability index of each photovoltaic power station according to the reactive power-voltage sensitivity among the photovoltaic power stations in the current photovoltaic cluster; Carrying out weighted summation on the observability index and the controllability index of each photovoltaic power station to generate a comprehensive index of the corresponding photovoltaic power station; and taking the photovoltaic power station with the highest comprehensive index as a leading node of the current photovoltaic cluster.
8. The power distribution network operation device based on reconstruction and cluster collaborative optimization is characterized by comprising a power distribution network data acquisition module, a model construction module, a target network topology and target photovoltaic output generation module, an iteration optimization module and a power distribution network scheduling module; the power distribution network data acquisition module is used for acquiring initial network topology, initial photovoltaic output, branch parameters and load of the power distribution network; The model construction module is used for constructing a network reconstruction model by taking the optimization of the whole network operation index as a target according to the branch parameters and the load; The target network topology and target photovoltaic output generation module is used for repeatedly executing the iterative optimization module until the relative deviation is smaller than a preset convergence threshold value to generate target network topology and target photovoltaic output; the iterative optimization module is used for solving the network reconstruction model according to the current photovoltaic output to generate the current optimal network topology and the current whole network operation index; the method comprises the steps of obtaining initial photovoltaic output, carrying out cluster division on all photovoltaic power stations in a power distribution network according to a current optimal network topology, generating a photovoltaic cluster division result, determining a leading node for each photovoltaic cluster by analyzing a voltage sensitivity relation among all photovoltaic power stations in each photovoltaic cluster, solving a collaborative optimization model according to the current optimal network topology, the leading node and the photovoltaic cluster division result, generating an optimal power output value of each photovoltaic power station, combining the optimal power output values of all photovoltaic power stations in the current round, generating a current optimal photovoltaic output, calculating the difference between a current whole-network operation index and a whole-network operation index in the last round, generating a relative deviation, judging whether the relative deviation is smaller than a preset convergence threshold, if so, recording the current optimal photovoltaic output as a target photovoltaic output, and recording the current optimal network topology as the target network topology, otherwise, updating the current optimal photovoltaic output as the current photovoltaic output; The power distribution network scheduling module is used for scheduling the power distribution network according to the target network topology and the target photovoltaic output.
9. An electronic device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the reconstruction-and cluster-co-optimization-based power distribution network operation method according to any one of claims 1 to 7 when executing the computer program.
10. A storage medium comprising a stored computer program, wherein the computer program, when run, controls a device in which the storage medium is located to perform the method of operation of a power distribution network based on collaborative optimization of reconstruction and clustering according to any one of claims 1 to 7.

Description

Power distribution network operation method and device based on reconstruction and cluster collaborative optimization, electronic equipment and storage medium Technical Field The invention relates to the technical field of power grid operation optimization, in particular to a power distribution network operation method, device, electronic equipment and storage medium based on reconstruction and cluster collaborative optimization. Background The power distribution network is used as a terminal link of a power system, and safe, stable and economical operation is a basis for ensuring the power consumption of a society. In recent years, with the large-scale grid connection of new energy sources such as distributed photovoltaic and the like, the inherent randomness and fluctuation of the distributed photovoltaic and the like bring great challenges to the operation control of a power distribution network, and the problems of power flow feedback, voltage out-of-limit and the like are easily caused. Meanwhile, the traditional centralized regulation and control method faces bottlenecks in the communication and calculation aspects due to the mass distributed power supply, and the regulation and control requirements of the novel power distribution network are difficult to meet. In order to cope with the above challenges, a method of clustering distributed photovoltaics is often adopted in the prior art to realize hierarchical partition control and relieve regulation and control pressure. However, these cluster partitioning strategies are mostly formulated based on a static, fixed grid topology, ignoring the dynamic nature of the power distribution network operation. In actual operation, the power distribution network frequently performs network reconstruction due to reasons such as load adjustment, fault recovery or economic optimization, and the like, so that the network topology is changed. Once the topology is changed, the original cluster scheme based on the solidified topology is not applicable any more, and the optimal scheduling is carried out on the basis, so that the system is inevitably trapped into the trouble of local optimization, and the global safe and economic operation of the whole power distribution network under the dynamic change is difficult to realize. Disclosure of Invention The embodiment of the invention provides a power distribution network operation method, a device, electronic equipment and a storage medium based on reconstruction and cluster collaborative optimization, which can solve the problem that a static cluster division strategy in the prior art cannot adapt to dynamic reconstruction of a power distribution network, so that an optimization result falls into local optimum. The embodiment of the invention provides a power distribution network operation method based on reconstruction and cluster collaborative optimization, which comprises the following steps: acquiring initial network topology, initial photovoltaic output, branch parameters and load of a power distribution network; according to the branch parameters and the load, a network reconstruction model is built by taking the optimization of the whole network operation index as a target; constructing a collaborative optimization model by taking the optimization of operation indexes in a cluster as a target according to the initial network topology, the initial photovoltaic output, the branch parameters and the load; Repeatedly executing iterative optimization operation until the relative deviation is smaller than a preset convergence threshold value, and generating a target network topology and a target photovoltaic output; Scheduling the power distribution network according to the target network topology and the target photovoltaic output; The iterative optimization operation comprises the steps of solving a network reconstruction model according to current photovoltaic output to generate a current optimal network topology and a current whole network operation index, wherein the initial photovoltaic output is the initial photovoltaic output; according to the current optimal network topology, all photovoltaic power stations in the power distribution network are clustered to generate a photovoltaic cluster dividing result, and a leading node is determined for each photovoltaic cluster by analyzing the voltage sensitivity relationship among the photovoltaic power stations in each photovoltaic cluster; solving the collaborative optimization model according to the current optimal network topology, the dominant nodes and the dividing result of the photovoltaic clusters to generate optimal power output values of all photovoltaic power stations, and combining the optimal power output values of all photovoltaic power stations in the current round to generate the current optimal photovoltaic output; calculating the difference between the current full-network operation index and the full-network operation index of the previous round to ge