CN-119834242-B - Distributed energy storage intelligent scheduling method and system based on station area autonomy

CN119834242BCN 119834242 BCN119834242 BCN 119834242BCN-119834242-B

Abstract

The invention discloses a distributed energy storage intelligent scheduling method and system based on the autonomous of a platform area, comprising the following steps: firstly, acquiring a task indication that the current power enters the target platform area, and acquiring power load data according to the task indication. And then, evaluating the load data by using a pre-trained platform load evaluation model, and judging whether the load state is normal or not. If so, the main root cause of the abnormality is determined. And finally, selecting a proper scheduling strategy from the pre-configured energy storage scheduling strategies based on the load state evaluation and abnormal state analysis results, and issuing the scheduling strategy to the distributed energy storage equipment of the target station area for execution. By the design, stability and scheduling efficiency of the power system are improved, and the intelligent management method is suitable for intelligent management of a distributed energy system.

Inventors

LI DEXIN
LI BAOJU
ZHANG YIFU
LI DAYONG
MENG XIANGDONG
ZHANG HAIFENG
YANG JINGYING
TIAN CHUNGUANG
LV XIANGYU
WANG JIARUI
ZHANG JIAJUN

Assignees

国网吉林省电力有限公司电力科学研究院

Dates

Publication Date: 20260508
Application Date: 20241219

Claims (9)

1. The distributed energy storage intelligent scheduling method based on the station area autonomy is characterized by comprising the following steps: Acquiring a task instruction of the current power entering a target platform area; Acquiring power load data of the current power according to the task indication; inputting the power load data of the current power into a pre-trained platform load evaluation model to obtain a corresponding load state evaluation result; when the load state evaluation result is characterized as an abnormal load state, determining a target abnormal root cause with highest association degree with an abnormal phenomenon of the abnormal load state, and obtaining an abnormal state analysis result; Selecting a target scheduling strategy from the pre-configured energy storage scheduling based on the load state evaluation result and the abnormal state analysis result; issuing the target scheduling strategy to distributed energy storage equipment included in the target platform area for execution; The platform load assessment model is obtained by the following steps: Acquiring power load data recorded in the process of transmitting target power to a target platform area, and carrying out transformation equipment association on the power load data to obtain sub power load data respectively corresponding to a plurality of transformation equipment, wherein the sub power load data comprises instantaneous load data respectively corresponding to each sampling time node of the target power in the process of transmitting the target power to the corresponding transformation equipment; based on each piece of sub-power load data, determining a load state corresponding to the corresponding transformer equipment, wherein the load state comprises a safe load state and an abnormal load state; Taking each piece of sub-power load data as a sample instance, taking the corresponding load state as an instance target value, constructing a sample instance array for training the platform area load evaluation model, and respectively executing the following processing for each piece of sub-power load data in the sample instance array: Acquiring a load mean value, a root mean square load value, a load fluctuation degree, a load dispersion degree and peak value data in the instantaneous load data of each instantaneous load data; dividing the load dispersion by the load mean value to obtain the variation rate deviation of the sub-power load data, dividing the peak value data by the root mean square load value to obtain the peak-to-average ratio of the sub-power load data, and dividing the peak value data by the load mean value to obtain the impact ratio of the sub-power load data; carrying out data integration processing on the load mean value, the load fluctuation degree, the change rate deviation, the peak-to-average ratio and the impact ratio to obtain load time sequence conversion domain data corresponding to a load time sequence analysis domain; acquiring spectrum distribution intensity and power spectrum density average value corresponding to the sub-power load data, and determining spectrum center frequency of the sub-power load data on an energy spectrum based on the spectrum distribution intensity; Combining the spectrum center frequency and the spectrum distribution intensity, determining the spectrum fluctuation degree of the sub-power load data on an energy spectrum, and acquiring a spectrum upper limit frequency in the spectrum distribution intensity; Performing data integration processing on the spectrum distribution intensity, the power spectrum density average value, the spectrum center frequency, the spectrum fluctuation degree and the spectrum upper limit frequency to obtain energy spectrum conversion domain data corresponding to the energy spectrum analysis domain; Performing power quality fluctuation mapping processing on the sub-power load data from a power quality fluctuation analysis domain to obtain power quality fluctuation conversion domain data corresponding to the power quality fluctuation analysis domain; Respectively executing feature extraction operation on each conversion domain data to obtain conversion domain features respectively corresponding to each conversion domain data; acquiring characteristic domain attributes corresponding to each conversion domain characteristic respectively, and carrying out weighted average processing on each characteristic domain attribute to obtain a reference domain attribute; Comparing each characteristic domain attribute with the reference domain attribute to obtain an attribute comparison result corresponding to each characteristic domain attribute; when the attribute comparison result represents that the attribute of the feature domain is consistent with the attribute of the reference domain, determining the corresponding conversion domain feature as a target conversion domain feature corresponding to the conversion domain feature; when the attribute comparison result represents that the characteristic domain attribute is inconsistent with the reference domain attribute, updating the characteristic domain attribute of the corresponding conversion domain feature into the reference domain attribute to obtain a target conversion domain feature corresponding to the conversion domain feature; performing feature integration processing on each target conversion domain feature to obtain integrated features; performing platform load assessment on the target platform area based on the integrated characteristic by using the platform area load assessment model to obtain an inferred load state of the target platform area; Executing a training process on the platform region load assessment model by combining the inferred load state and the corresponding load state; The platform region load evaluation model is used for evaluating the load state of the platform region where the target power is located based on the power load data of the target power.
2. The method of claim 1, wherein the acquiring power load data recorded by the target power during transmission to the target site comprises: Acquiring voltage fluctuation and current change corresponding to each sampling time node respectively in the process of transmitting the target power to the target platform area; integrating the standardized value of the corresponding voltage fluctuation and the standardized value of the current variation aiming at each sampling time node to obtain the instantaneous load data corresponding to the sampling time node; and arranging each piece of instantaneous load data according to the time sequence of the sampling time node to obtain the power load data.
3. The method of claim 1, wherein the sub-power load data includes instantaneous load data corresponding to each sampling time node of the target power during transmission to the corresponding transformer device, and wherein determining the load state corresponding to the corresponding transformer device based on each sub-power load data comprises: The following processing is performed for each of the sub power load data: For each sampling time node corresponding to the sub-power load data, at least one adjacent sampling time node of the sampling time node is obtained, and difference value operation is carried out on the instantaneous load data corresponding to the sampling time node and the instantaneous load data corresponding to each adjacent sampling time node respectively to obtain the load fluctuation variation corresponding to each adjacent sampling time node respectively; Determining a target load safety level as a first load safety level when at least one of the load fluctuation amounts exceeds a preset change amount threshold; determining the target load security level as a second load security level when the load fluctuation variance exceeds the preset variance threshold; The first load security level is used for representing that the load of the platform area, which is passed by the sampling time node, of the target power is an abnormal load range, and the second load security level is used for representing that the load of the platform area, which is passed by the sampling time node, of the target power is a safe load range; When the target load security level represents that the platform load of the target platform, which is passed by the target power at the sampling time node, is a security load range, determining the sampling time node corresponding to the target load security level as a target sampling time node; and determining the load state corresponding to the corresponding transformation equipment based on the node number of the target sampling time nodes.
4. A method according to claim 3, wherein said determining a load state corresponding to the corresponding transformation device based on the node number of the target sampling time node comprises: comparing the node number of the target sampling time node with a node number threshold value to obtain a node number comparison result; When the node number comparison result represents that the node number of the target sampling time node exceeds the node number threshold, determining the load state as the safe load state; when the node number comparison result indicates that the node number of the target sampling time node does not exceed the node number threshold, determining the load state as the abnormal load state; the determining the load state corresponding to the voltage transformation equipment based on the node number of the target sampling time node further comprises: Dividing the node number of the target sampling time node by the node number of the sampling time node to obtain the load stabilization rate of the target platform region; When the load stabilization rate exceeds a load stabilization rate threshold, determining the load state as the safe load state, and when the load stabilization rate does not exceed the load stabilization rate threshold, determining the load state as the abnormal load state.
5. The method of claim 1, wherein after performing a training procedure on the site load assessment model in conjunction with the inferred load state and the corresponding load state, the method further comprises: Acquiring a power dispatching plan of target power, and searching a load state of a reference platform corresponding to the power dispatching plan from a platform load database to obtain a search result; When the searching result represents that the load state of the standard platform area is not stored in the platform area load database, acquiring target load data of the standard platform area in a similar time period, and executing feature extraction operation on the target load data to obtain target load features; Performing platform load assessment on the reference platform area based on the target load characteristics by using a platform load assessment model after training to obtain an inferred load state of the reference platform area; And updating the inferred load state of the reference platform region into the platform region load database to obtain an updated platform region load database.
6. The method of claim 5, wherein the method further comprises: When the search result represents that the load state of the standard platform area is stored in the platform area load database, and the load state of the standard platform area represents that the platform area load of the standard platform area is an abnormal load range, generating a first notification message; when the search result represents that the load state of the reference platform is stored in the platform area load database and the load state of the reference platform represents that the platform area load of the reference platform is in a safe load range, generating a second notification message; the first notification message is used for notifying that the power dispatching plan is adjusted on the platform load database, and the second notification message is used for notifying that the power load of the reference platform corresponding to the power dispatching plan is normal.
7. The method according to claim 1, wherein the determining the target abnormality root cause having the highest degree of association with the abnormality of the abnormal load state, to obtain the abnormality analysis result, includes: Acquiring an abnormal analysis instruction aiming at an abnormal load state, and extracting an abnormal phenomenon of the abnormal load state from the abnormal analysis instruction; determining a district system to which the abnormal load state belongs, and acquiring a preset hypergraph of the district system, wherein each entity of the preset hypergraph comprises an abnormal phenomenon entity in which the abnormal phenomenon is located and a potential abnormal root cause entity except the abnormal phenomenon entity; Determining a plurality of key abnormal root cause entities from each potential abnormal root cause entity according to the corresponding entity relation of each entity in the preset hypergraph; determining an abnormal factor characterized by the key abnormal root cause entity as a pending abnormal root cause of the abnormal load state; normalizing the characteristic parameters corresponding to each undetermined abnormal root cause and the abnormal phenomenon to obtain normalized data; performing feature extraction operation on the standardized data to obtain abnormal factor features meeting the training requirements of expert models; Constructing a sample instance array containing the abnormal factor characteristics; performing expert model training on the sample instance array to obtain a target expert model; Taking the characteristic parameters corresponding to each undetermined abnormal root cause as the input of the target expert model, taking the characteristic parameters of the abnormal phenomenon as the output of the target expert model, and calculating the input expected influence degree of the target expert model through reverse propagation; based on each expected superposition influence degree, respectively determining initial association coefficients of the undetermined abnormal root causes and the abnormal phenomena, wherein the initial association coefficients are matched with the expected superposition influence degrees of the undetermined abnormal root causes; Aiming at each undetermined abnormal root cause, carrying out characteristic shielding on characteristic parameters of the undetermined abnormal root cause to obtain shielded data; determining a model stability evaluation value of the undetermined anomaly root cause on the target expert model based on the characteristic parameters of the anomaly and the deviation between the output response of the target expert model on the masked data; Combining the initial association coefficient and the model stability evaluation value to determine the association coefficient of the undetermined abnormal root cause and the abnormal phenomenon; determining a preset association coefficient threshold associated with the current training period; comparing the association coefficient corresponding to each undetermined abnormal root cause with the preset association coefficient threshold value, and determining a key abnormal root cause of which the association coefficient accords with the preset association coefficient threshold value from each undetermined abnormal root cause; Based on the characteristic parameter distribution of the key abnormal root cause updating sample instance, determining an adjusted sample instance array, and repeatedly executing the step of expert model training on the sample instance array until the key abnormal root cause accords with a preset termination state to obtain a target abnormal root cause; the sample instance of the subsequent expert model training comprises the characteristic parameters of the abnormal phenomenon and the characteristic parameters of the key abnormal root cause determined by the previous expert model training, wherein the characteristic parameters of the key abnormal root cause determined by the previous expert model training are different from the characteristic parameter distribution of the characteristic parameters of the abnormal root cause in the sample instance of the previous expert model training; and taking the association and matching relation between the abnormal phenomenon and the target abnormal root cause as an abnormal state analysis result of the abnormal load state.
8. The method of claim 7, wherein the training the expert model on the sample instance array to obtain a target expert model comprises: Determining a characteristic parameter attribute corresponding to each characteristic parameter in the sample instance array; Performing expert model training on the sample instance array by adopting a core training architecture associated with each characteristic parameter attribute to obtain a target expert model; The training of the expert model on the sample instance array to obtain a target expert model further comprises: Dividing the sample instance array into a plurality of sub-sample instance arrays, wherein each sub-sample instance array comprises characteristic parameters of the abnormal phenomenon and characteristic parameters of at least a part of the undetermined abnormal root cause; and respectively performing expert model training on each sub-sample instance array to obtain a target expert model corresponding to each sub-sample instance array.
9. A server system comprising a server for performing the method of any of claims 1-8.

Description

Distributed energy storage intelligent scheduling method and system based on station area autonomy Technical Field The invention relates to the technical field of artificial intelligence, in particular to a distributed energy storage intelligent scheduling method and system based on station area autonomy. Background With the rapid development of distributed energy and smart grid technologies, intelligent scheduling of district autonomy and distributed energy storage systems is becoming increasingly important. Conventional power dispatching methods typically rely on a centralized control system, which may not be flexible and efficient in a distributed energy system. Disclosure of Invention The invention aims to provide a distributed energy storage intelligent scheduling method and system based on the autonomous region of a platform. In a first aspect, an embodiment of the present invention provides a distributed energy storage intelligent scheduling method based on a platform autonomous, including: Acquiring a task instruction of the current power entering a target platform area; Acquiring power load data of the current power according to the task indication; inputting the power load data of the current power into a pre-trained platform load evaluation model to obtain a corresponding load state evaluation result; when the load state evaluation result is characterized as an abnormal load state, determining a target abnormal root cause with highest association degree with an abnormal phenomenon of the abnormal load state, and obtaining an abnormal state analysis result; Selecting a target scheduling strategy from the pre-configured energy storage scheduling based on the load state evaluation result and the abnormal state analysis result; and issuing the target scheduling strategy to distributed energy storage equipment included in the target platform area for execution. In a second aspect, an embodiment of the present invention provides a server system, including a server, where the server is configured to perform the method described in the first aspect. Compared with the prior art, the distributed energy storage intelligent scheduling method and system based on the station area autonomy have the beneficial effects that the task indication that the current power enters the target station area is firstly obtained, and the power load data is obtained according to the task indication. And then, evaluating the load data by using a pre-trained platform load evaluation model, and judging whether the load state is normal or not. If so, the main root cause of the abnormality is determined. And finally, selecting a proper scheduling strategy from the pre-configured energy storage scheduling strategies based on the load state evaluation and abnormal state analysis results, and issuing the scheduling strategy to the distributed energy storage equipment of the target station area for execution. By the design, stability and scheduling efficiency of the power system are improved, and the intelligent management method is suitable for intelligent management of a distributed energy system. Drawings In order to more clearly illustrate the technical solution of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described. It is appreciated that the following drawings depict only certain embodiments of the invention and are therefore not to be considered limiting of its scope. Other relevant drawings may be made by those of ordinary skill in the art without undue burden from these drawings. Fig. 1 is a schematic step flow diagram of a distributed energy storage intelligent scheduling method based on the autonomous region of a platform according to an embodiment of the present invention; fig. 2 is a schematic block diagram of a computer device according to an embodiment of the present invention. Detailed Description In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. The following describes specific embodiments of the present invention in detail with reference to the drawings. In order to solve the technical problems in the foregoing background, fig. 1 is a schematic flow chart of a distributed energy storage intelligent scheduling method based on the autonomous region of a platform, which is provided in an embodiment of the present disclosure, and the detailed description of the distributed energy stora