Time-sharing segmented regulation and control method linked with charging load of distribution transformer system

Abstract

The invention discloses a time-sharing segmented regulation and control method linked with a charging load of a distribution transformer system, which relates to the technical field of intelligent power grids, and comprises the following steps: dynamically segmenting the time linked with the charging load of the distribution transformer system all day long, and providing a time division basis for time-sharing regulation and control; the output power and the access sequence in the charging process are dynamically adjusted, and closed-loop linkage regulation and control between the distribution transformer system and the charging load are completed; and generating a regulation and control decision adaptive to the current operation environment, and completing adaptive adjustment in the time-sharing and segmented regulation and control process. According to the method, accurate prediction of the charging load of the distribution transformer system and fine division in the time dimension are realized, the perspectiveness and adaptability of load regulation and control are improved, the equipment operation safety and the user charging satisfaction degree are improved, a regulation and control decision can be dynamically evolved according to the actual operation environment, and the user experience is improved. The intelligent linkage and fine management of the charging load and the distribution transformer system are realized, and the operation efficiency of the power distribution network and the utilization benefit of electric energy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart grid, in particular to a time-division and section-regulation method linked with charging load of distribution transformer system. BACKGROUND

[0002] With the rapid development of new energy vehicles and distributed energy systems, the role of distribution network in power system is increasingly important. As a key link of distribution network, the operation efficiency and load regulation capacity of distribution transformer system directly affect the stability and economy of the whole power grid. Especially under the background of rapid growth of electric vehicle charging load, the fixed regulation mode of traditional distribution network has been difficult to meet the demand of dynamic load change. Therefore, an intelligent regulation method linked with charging load is urgently needed to realize fine management and optimal scheduling of distribution network operation state.

[0003] The existing regulation technology of distribution transformer system mostly adopts fixed strategy in a unified time period, which lacks dynamic response ability to load change characteristics. Especially when facing charging load such as electric vehicle with obvious space-time distribution characteristics, it is easy to cause problems such as overload of distribution transformer equipment or waste of resources. In addition, the existing technology fails to fully realize the collaborative regulation between time-division and section-regulation and charging load, resulting in low efficiency of power grid operation and affecting power supply quality and user satisfaction. Therefore, it is urgent to propose a time-division and section-regulation method linked with charging load of distribution transformer system to improve the adaptability and regulation precision of distribution network to dynamic load. SUMMARY

[0004] In view of the problems existing in the prior art time-division and section-regulation method linked with charging load of distribution transformer system, the present application is proposed.

[0005] Therefore, the problem to be solved by the present application is how to realize dynamic collaborative regulation between distribution transformer system and charging load to cope with the problem of accurate time-division and section-regulation caused by large fluctuation of charging load, diverse user demand and complex system operation constraints.

[0006] To solve the above technical problems, the present application provides the following technical scheme:

[0007] In the first aspect, the present application provides a time-division and section-regulation method linked with charging load of distribution transformer system, which comprises: collecting multi-source data of charging load, establishing fusion features of extracted multi-source data, analyzing time series load prediction mechanism of fusion features of multi-source data, dynamically segmenting time linked with charging load of distribution transformer system throughout the day on the basis of time series load prediction mechanism, and providing time division basis for time regulation.

[0008] Based on dynamic segmentation and time division, the distribution transformer system is monitored to obtain priority information of the charging process, and according to the load prediction value of the current time period and the priority information, the output power and access sequence of the charging process are dynamically adjusted to complete the closed-loop linkage regulation and control between the distribution transformer system and the charging load.

[0009] During the execution of the closed-loop linkage regulation and control strategy, the regulation target is modeled, the priority information is optimized and the weight of the priority information is adjusted to generate a regulation decision suitable for the current operating environment, and adaptive adjustment in the time-of-use and segmented regulation process is completed.

[0010] As a preferred scheme of the time-of-use and segmented regulation method linked with the charging load of the distribution transformer system, wherein: the collection of the multi-source data of the charging load includes constructing a collection framework of multi-source data, deploying an edge collection cloud data platform in the collection framework, synchronously collecting the multi-source data of the charging load in the distribution transformer system, and completing time sequence alignment through a unified clock synchronization mechanism;

[0011] The multi-source data fusion feature extraction includes processing the collected multi-source data of the charging load in the distribution transformer system to generate a data stream of the multi-source data of the charging load, constructing a feature encoder for different types of multi-source data of the charging load in the distribution transformer system, vectorizing and preprocessing the multi-source data of the charging load in different types of distribution transformer systems, and extracting local time sequence features of the data stream through a trainable linear transformation layer.

[0012] The feature encoder construction includes inputting the multi-source data of the charging load in different types of distribution transformer systems into a multi-scale attention fusion network, wherein the multi-scale attention fusion network is composed of several parallel attention heads, each attention head focuses on the interaction of different types of multi-source data of the charging load in the distribution transformer system, and generates a representation of the multi-source data fusion feature.

[0013] As a preferred scheme of the time-of-use and segmented regulation method linked with the charging load of the distribution transformer system, wherein: the time sequence load prediction mechanism for analyzing the multi-source data fusion feature includes using a time sequence modeling method to model the multi-source data fusion feature based on the representation of the multi-source data fusion feature.

[0014] The time sequence modeling of the multi-source data fusion feature includes:

[0015] The multi-source data fusion feature and the vector of the preprocessed multi-source data of the charging load in the distribution transformer system are arranged in time stamp order to form a high-dimensional time sequence feature sequence, which is input into a hybrid time sequence modeling network.

[0016] The dynamic segmentation of the time of the power load of the distribution system throughout the day includes generating a predicted load curve under the time scale of the whole day after time series modeling of the multi-source data fusion features, the predicted load curve including predicted load values at time points, and based on the predicted load curve, using a dynamic density clustering algorithm to segment the curve;

[0017] The dynamic density clustering algorithm calculates the load change rate between adjacent time points through a sliding window, and dynamically identifies the inflection point of the load change by combining the local density and the distance threshold to divide the whole day into several time periods with load change characteristics.

[0018] As a preferred scheme of the time segmentation and regulation method linked with the charging load of the distribution system, wherein: the time division of the time segmentation includes introducing a dynamic adjustment mechanism, the dynamic adjustment mechanism refers to training a Q network model with the state between the current predicted load curve and the historical load mode as input, the Q network model is used to adjust the parameters of the distance threshold and the local density in the dynamic density clustering algorithm, so that the time period division result is self-adaptively optimized according to the real-time load change trend, and the dynamic segmentation of the time of the distribution system charging load linkage is completed.

[0019] As a preferred scheme of the time segmentation and regulation method linked with the charging load of the distribution system, wherein: the monitoring of the distribution system includes deploying an edge computing unit based on the time series load prediction and dynamic segmentation time division based on the multi-source data fusion features;

[0020] The running parameters including the load rate of the distribution transformer, the three-phase current unbalance degree, the voltage fluctuation, the environmental temperature and the historical load trend are collected, and the time stamp is aligned through a unified clock mechanism and a time segmentation label to form a distribution running state data stream with a time period label;

[0021] The user interaction platform connected through the communication interface obtains the priority information of the charging process, the priority information including the current state of charge of the battery, the user's scheduled charging time, the user's historical credit score and the charging type, the priority information is processed by a feature encoder to generate a priority feature vector, and is fused with the predicted load value of the current time period as an input of the regulation decision;

[0022] The fused predicted load value of the current time period is input into a dynamic regulation decision engine of a hybrid strategy, and the engine is composed of two parallel running subsystems:

[0023] The priority sorting subsystem;

[0024] The power distribution and access control subsystem.

[0025] As a preferred scheme of the time-sharing and segmenting regulation method linked with the charging load of the distribution system, in the application, the modeling of the regulation target comprises determining a set of interrelated regulation targets according to the operation characteristics of the distribution system and the charging load management demand in the closed-loop linkage regulation strategy execution process, and the regulation targets comprise a distribution load safety constraint target, a three-phase load balance target, a user charging demand satisfaction degree target, a power regulation smoothness target, and a device operation energy efficiency target.

[0026] According to the unified regulation target modeling formula of the regulation target, an optimized target function is constructed through a weighted normalization and a multi-target coupling mechanism, which is used for generating the modeling of the regulation target, and the unified regulation target modeling formula of the regulation target is:

[0027]

[0028] Among them, indicates the unified regulation target modeling, indicates the weight of the distribution load safety constraint target, indicates the normalized distribution load safety constraint target function, indicates the weight of the three-phase load balance target, indicates the normalized three-phase load balance target function, indicates the weight of the user charging demand satisfaction degree target, indicates the normalized user charging demand satisfaction degree target function, indicates the weight of the power regulation smoothness target, indicates the normalized power regulation smoothness target function, indicates the weight of the device operation energy efficiency target, indicates the dynamic coupling matrix between targets, indicates the weight vector, indicates the target coupling item.

[0029] As a preferred scheme of the time-sharing and segmenting regulation method linked with the charging load of the distribution system, in the application, the optimization of the priority information and the adjustment of the weight of the priority information comprise interacting the unified regulation target modeling with the dynamic coupling matrix between targets, evaluating the conflicts between the regulation targets under the current weight distribution, generating a regulation instruction based on the coupling items of the regulation targets under the current weight distribution, and combining the load prediction value of the current time period and the operation state of the distribution, to complete the adaptive adjustment in the time-sharing and segmenting regulation process.

[0030] In a second aspect, an embodiment of the application provides a time-sharing and segmenting regulation system linked with the charging load of a distribution system, which comprises:

[0031] S310: a dynamic segmentation and time division module, which collects multi-source data of charging load, establishes fused features of the extracted multi-source data, analyzes a time series load prediction mechanism of the fused features, dynamically segments a time of charging load linkage of the distribution transformer system throughout the day on the basis of the time series load prediction mechanism, and provides a time division basis for time division regulation;

[0032] S320: a closed-loop linkage regulation module between the distribution transformer system and the charging load, which monitors the distribution transformer system on the basis of dynamic segmentation and time division, acquires priority information of the charging process, dynamically adjusts output power and access order of the charging process according to a load prediction value of a current time period and the priority information, and completes closed-loop linkage regulation between the distribution transformer system and the charging load;

[0033] S330: an adaptive adjustment module in the time division segmentation regulation process, which models a regulation target, optimizes priority information and adjusts a weight of the priority information, generates a regulation decision adapted to a current operating environment, and completes adaptive adjustment in the time division segmentation regulation process.

[0034] In a third aspect, an embodiment of the present application provides a computer device, including a memory and a processor, and the memory stores a computer program, wherein the processor implements any step of the time division segmentation regulation method linked with charging load of the distribution transformer system when executing the computer program.

[0035] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement any step of the time division segmentation regulation method linked with charging load of the distribution transformer system.

[0036] The present application has the following beneficial effects: the present application realizes accurate prediction of charging load of the distribution transformer system and fine division in the time dimension by constructing multi-source data fusion features and a dynamic time division segmentation mechanism, improves the foresight and adaptability of load regulation, and based on a closed-loop regulation strategy of real-time monitoring and priority information, can dynamically adjust output power and access order of the charging process, effectively balances the load of the distribution transformer system, suppresses three-phase imbalance, and smoothes voltage fluctuation, improves equipment operation safety and user charging satisfaction, through establishing a multi-objective optimization model and introducing a weight adaptive adjustment mechanism, makes the regulation decision dynamically evolve according to the actual operating environment, enhances the robustness and adaptive ability of the system under complex working conditions, and the overall scheme realizes intelligent linkage and fine management of the charging load and the distribution transformer system, which is helpful to improve the operation efficiency of the distribution network and the utilization efficiency of electric energy. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments description will be briefly introduced as follows. Obviously, the drawings in the following description only represent some of the embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings. Among them:

[0038] Figure 1 The method flow chart of the time-sharing and section-regulating method linked with the charging load of the distribution system provided by one embodiment of the present application.

[0039] Figure 2 The system flow chart of the time-sharing and section-regulating method linked with the charging load of the distribution system provided by one embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the above objectives, features and advantages of the present application more apparent, the specific embodiments of the present application will be described in detail below with the drawings in the specification. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without any creative effort should fall within the protection scope of the present application.

[0041] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced without the specific details, other than those described in the specification, and it is understood that the present application is not limited to the embodiments described herein. In some instances, well-known structures and functions have not been described in detail in order to avoid obscuring the application.

[0042] Secondly, the "one embodiment" or "embodiment" referred to herein can include specific features, structures or characteristics contained in at least one implementation of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments.

[0043] The present application is described in detail in combination with the schematic drawings, and in the detailed description of the embodiments of the present application, the sectional view of the device structure will be partially enlarged without the general proportion for the convenience of description, and the schematic drawings are only examples, which should not limit the protection scope of the present application herein. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in the actual manufacture.

[0044] Meanwhile, in the description of the present application, it should be noted that the terms "upper, lower, inner and outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first, second or third" are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0045] Unless otherwise expressly specified and limited, the terms "mounting, connecting, connecting" in the present application should be understood broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0046] Embodiment 1

[0047] Reference Figure 1 And Figure 2 The first embodiment of the present application provides a time-sharing and segmented control method linked with the charging load of the distribution system, comprising:

[0048] S1: By collecting multi-source data of charging load, establishing extracted multi-source data fusion features, analyzing the time series load prediction mechanism of multi-source data fusion features, and on the basis of the time series load prediction mechanism, dynamically segmenting the time of the whole day of the distribution system charging load linkage, providing the time division basis of time-sharing control.

[0049] Among them, collecting multi-source data of charging load includes constructing a multi-source data collection framework, deploying an edge collection cloud data platform in the collection framework, synchronously collecting multi-source data of charging load in the distribution system, and completing time sequence alignment through a unified clock synchronization mechanism;

[0050] Establishing the extracted multi-source data fusion features includes processing the collected multi-source data of charging load in the distribution system, generating a data stream of the multi-source data of charging load, constructing a feature encoder for different types of multi-source data of charging load in the distribution system, vectorizing preprocessing different types of multi-source data of charging load in the distribution system, and extracting local time sequence features of the data stream through a trainable linear transformation layer;

[0051] The feature encoder construction includes inputting a multi-scale attention fusion network under the feature encoder constructed by the multi-source data of the charging load in different types of distribution system, the multi-scale attention fusion network is composed of several parallel attention heads, each attention head focuses on the interaction of the multi-source data of the charging load in different types of distribution system, and generates a representation of the multi-source data fusion feature.

[0052] S1.1: The time series load prediction mechanism of the multi-source data fusion feature includes that based on the completion of the representation of the multi-source data fusion feature, a time series modeling method is used to model the multi-source data fusion feature in time sequence;

[0053] The time series modeling of the multi-source data fusion feature includes:

[0054] The multi-source data fusion feature and the vector of the preprocessed multi-source data of the charging load in the distribution system are arranged in time stamp order to form a high-dimensional time sequence feature sequence, which is input into a hybrid time sequence modeling network;

[0055] The dynamic segmentation of the charging load linkage time of the distribution system throughout the day includes that after the time series modeling of the multi-source data fusion feature, a predicted load curve under the time scale throughout the day is generated, the predicted load curve includes the predicted load value of the time point, and based on the predicted load curve, a dynamic density clustering algorithm is used to segment the curve;

[0056] The dynamic density clustering algorithm calculates the load change rate between adjacent time points through a sliding window, and dynamically identifies the inflection point of the load change by combining the local density and the distance threshold, so as to divide the whole day into several time periods with load change characteristics.

[0057] S1.2: The time division of time division control includes introducing a dynamic adjustment mechanism, the dynamic adjustment mechanism refers to taking the state between the current predicted load curve and the historical load mode as input to train a Q network model, the Q network model is used to online adjust the parameters of the distance threshold and the local density in the dynamic density clustering algorithm, so that the time period division result is adaptively optimized according to the real-time load change trend, and the dynamic segmentation of the charging load linkage time of the distribution system is completed.

[0058] S2: Based on the dynamic segmentation and time division, the distribution system is monitored to obtain priority information of the charging process, and according to the load prediction value and the priority information of the current time period, the output power and the access sequence of the charging process are dynamically adjusted to complete the closed-loop linkage control between the distribution system and the charging load.

[0059] Among them, monitoring the distribution transformer system includes deploying edge computing units based on time series load forecasting and dynamic segmentation based on multi-source data fusion features; real-time monitoring of the operating status of the distribution transformer system;

[0060] The system collects operating parameters including the load factor, three-phase current imbalance, voltage fluctuation, ambient temperature, and historical load trends of distribution transformers. Timestamps are aligned with time segment labels through a unified clock mechanism to form a distribution transformer operating status data stream with time segment labels.

[0061] The user interaction platform connected through the communication interface obtains priority information of the charging process. The priority information includes the current state of charge of the battery, the user's scheduled charging time, the user's historical credit score, and the charging type. The priority information is processed by the feature encoder to generate a priority feature vector. This vector is then integrated with the load forecast value of the current time segment and used as the input for the control decision.

[0062] The fused load forecast value for the current time segment is input into the dynamic control decision engine of the hybrid strategy, which consists of two subsystems running in parallel:

[0063] Prioritization subsystem;

[0064] Power distribution and access control subsystem.

[0065] S3: During the execution of the closed-loop linkage control strategy, the control objectives are modeled, priority information is optimized and the weight of the priority information is adjusted to generate control decisions that adapt to the current operating environment, completing adaptive adjustments during the time-sharing and segmented control process.

[0066] The modeling of control objectives involves determining a set of interrelated control objectives during the execution of the closed-loop linkage control strategy based on the operating characteristics of the distribution transformer system and the charging load management requirements. The control objectives include the distribution transformer load safety constraint objective, the three-phase load balance objective, the user charging demand satisfaction objective, the power regulation smoothness objective, and the equipment operation energy efficiency objective.

[0067] According to the unified control target modeling formula of the control target, through weighted normalization and multi-target coupling mechanism, an optimized objective function is constructed to generate the model of the control target. The unified control target modeling formula of the control target is:

[0068]

[0069] in, represents unified control target modeling, represents the weight of the distribution transformer load safety constraint target, represents the normalized distribution transformer load safety constraint objective function, a weight representing a three-phase load balance target, a normalized three-phase load balance target function, a weight representing a user charging demand satisfaction target, a normalized user charging demand satisfaction target function, a weight representing a power regulation smoothness target, a normalized power regulation smoothness target function, a weight representing a device operation energy efficiency target, a dynamic coupling matrix between targets, a weight vector, a target coupling term.

[0070] S3.1: The optimization of priority information and the adjustment of the weight of the priority information include the interaction of the unified regulation target modeling and the dynamic coupling matrix between targets, the evaluation of the conflict between the regulation targets under the current weight distribution, the generation of the regulation instruction based on the evaluation of the coupling term of the regulation targets under the current weight distribution, and the combination of the load prediction value and the distribution transformer operation state of the current time period, to complete the adaptive adjustment in the time-of-use and time-of-day regulation process.

[0071] In a preferred embodiment, a time-of-use and time-of-day regulation system linked with the charging load of the distribution transformer system, the system includes a dynamic segmentation and time division module that establishes the extracted multi-source data fusion features by collecting multi-source data of the charging load, analyzes the time series load prediction mechanism of the multi-source data fusion features, dynamically segments the time of the charging load linkage of the distribution transformer system throughout the day based on the time series load prediction mechanism, and provides the time division basis for time-of-use regulation; a closed-loop linkage regulation module between the distribution transformer system and the charging load, which is based on the dynamic segmentation and time division, monitors the distribution transformer system, obtains the priority information of the charging process, dynamically adjusts the output power and access sequence of the charging process according to the load prediction value of the current time period and the priority information, and completes the closed-loop linkage regulation between the distribution transformer system and the charging load; an adaptive adjustment module in the time-of-use and time-of-day regulation process, which models the regulation target during the execution of the closed-loop linkage regulation strategy, optimizes the priority information and adjusts the weight of the priority information, generates regulation decisions that adapt to the current operating environment, and completes the adaptive adjustment in the time-of-use and time-of-day regulation process.

[0072] The above-mentioned each unit module can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to call and execute the operation corresponding to each module by the processor.

[0073] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used for wired or wireless communication with external terminals. The wireless mode can be achieved by WIFI, operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device. It can also be an external keyboard, touchpad or mouse, etc.

[0074] In summary, by constructing multi-source data fusion features and dynamic time division and segmentation mechanism, the application realizes accurate prediction of charging load of the distribution transformer system and fine division in time dimension, improves the foresight and adaptability of load regulation, and based on the closed-loop control strategy of real-time monitoring and priority information, dynamically adjusts the output power and access sequence of the charging process, effectively balances the load of the distribution transformer system, suppresses three-phase imbalance and smoothes voltage fluctuation, improves the safety of equipment operation and user charging satisfaction, establishes a multi-objective optimization model and introduces a weight self-adaptive adjustment mechanism, so that the control decision can dynamically evolve according to the actual operation environment, and the robustness and adaptive ability of the system under complex working conditions are enhanced. The overall scheme realizes intelligent linkage and fine management of charging load and distribution transformer system, which helps to improve the operation efficiency of distribution network and the utilization benefit of electric energy.

[0075] Embodiment 2

[0076] Reference Figure 1 and Figure 2 As the second embodiment of the application, the embodiment provides a time division and segmentation control method linked with the charging load of the distribution transformer system. In order to verify the beneficial effects of the application, scientific demonstration is carried out through simulation experiment.

[0077] In the practical application of the power distribution station in the residential area of the coastal city, the technical scheme of the application is deployed in the charging load management system of a power distribution transformer with a rated capacity of 800 kVA. There are 60 electric vehicle charging piles in the area, and the daily charging load accounts for 42% of the peak load of the distribution transformer. The system has been running since March 2024, and through the edge collection terminal, it synchronously obtains charging pile operation data, user reservation information, battery SOC, weather temperature, and real-time load of the distribution transformer. The sampling period is 1 minute. After multi-source data fusion and hybrid time series modeling, the root mean square error of load prediction is controlled within 3.8% for 24 hours. Based on the dynamic density clustering and Q network parameter adaptive mechanism, the system divides the whole day into 6-8 dynamic periods, with an average period length of 1.5-3.5 hours, which is significantly different from the traditional fixed peak-valley segmentation.

[0078] In the regulation execution, the priority sorting subsystem dynamically adjusts the task order in combination with user credit and reservation time. The power distribution module adjusts the output power according to real-time load and three-phase current imbalance degree, from an average of 18.7% to 9.2%. For example, during the load peak period of 18:00-20:30 on a high-temperature working day in July 2024, the system automatically reduces the power of 15 non-emergency charging tasks from 11kW to 6kW, and delays the access of 3 low-priority vehicles, so that the maximum load rate of the distribution transformer is reduced from 102% to 93%, avoiding tripping accidents, while the charging completion rate of high-priority users remains at 98.6%. The six-month continuous operation data shows that the daily load rate fluctuation range of the distribution transformer is reduced by 31.5%, and the voltage deviation is converged to within ±4%, verifying the effectiveness and stability of the scheme in real complex scenarios. The comparison between the application and the prior art is shown in Table 1 below:

[0079] Table 1 Comparison table of the application and the prior art

[0080] Comparison item Prior art Technical solution of the present application Load prediction basis Single historical load data or simple multi-source data Deep time series modeling based on multi-source heterogeneous data fusion features, with improved prediction accuracy (RMSE ≤ 3.8%) Time segmentation method Fixed peak-valley flat period division Dynamic density clustering based on load change inflection point, realizing adaptive dynamic segmentation Regulation and decision mechanism Static priority or rule control Closed-loop linkage regulation and control integrating real-time load prediction, user priority and system state Optimization target processing Single target or weighted fixed multi-target Multi-target coupling modeling and online adaptive weight adjustment, realizing dynamic balance System response capability Open-loop control, lacking feedback adjustment Adaptive closed-loop regulation and control capability with online learning and feedback iteration

[0081] Table 1 shows that compared with the prior art, the load prediction accuracy, time segmentation flexibility, regulation decision intelligence, multi-objective optimization ability, and system response closed-loop nature are all significantly improved. Through dynamic segmentation, closed-loop linkage, and adaptive optimization mechanism, the complexity and real-time problem of charging load and distribution system collaborative regulation are effectively solved.

[0082] It should be noted that the above embodiments are only used to illustrate the technical solutions of the application and not to limit it. Although the application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the application, which should be covered in the scope of the claims of the application.

Claims

1. A time-sharing and segmented control method linked to the charging load of a distribution transformer system, characterized by: include, By collecting multi-source data on charging loads, establishing the extracted multi-source data fusion features, and analyzing the time series load forecasting mechanism of the multi-source data fusion features, the time of the distribution transformer system charging load linkage throughout the day is dynamically segmented based on the time division basis for time-sharing control. Based on dynamic segmentation and time division, the distribution transformer system is monitored to obtain priority information of the charging process. According to the load forecast value and priority information of the current time period, the output power and access sequence of the charging process are dynamically adjusted to complete the closed-loop linkage control between the distribution transformer system and the charging load; During the execution of the closed-loop linkage control strategy, the control objectives are modeled, priority information is optimized and the weight of the priority information is adjusted to generate control decisions that adapt to the current operating environment, and complete adaptive adjustments in the time-sharing and segmented control process.

2. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 1 is characterized in that: The multi-source data collection of charging loads includes building a multi-source data collection framework, deploying an edge data collection cloud platform in the collection framework, synchronously collecting multi-source data of the distribution transformer system and the charging load, and completing timing alignment through a unified clock synchronization mechanism; The establishment of the extracted multi-source data fusion features includes processing the collected multi-source data related to the charging load in the distribution transformer system to generate a data stream of the multi-source data of the charging load, constructing a feature encoder for the collected multi-source data related to the charging load in different types of distribution transformer systems, performing vectorization preprocessing on the multi-source data related to the charging load in different types of distribution transformer systems, and extracting local time series features of the data stream through a trainable linear transformation layer; The constructing of the feature encoder includes inputting a multi-scale attention fusion network into a feature encoder constructed with multi-source data of charging loads in different types of distribution transformer systems. The multi-scale attention fusion network is composed of a plurality of parallel attention heads, each of which focuses on interacting with multi-source data of charging loads in different types of distribution transformer systems to generate a representation of the multi-source data fusion feature.

3. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 2 is characterized in that: The time series load forecasting mechanism of analyzing multi-source data fusion characteristics includes performing time series modeling on the multi-source data fusion characteristics by using a time series modeling method based on the completion of the representation of the multi-source data fusion characteristics; The time series modeling of multi-source data fusion features includes: The multi-source data fusion features and the pre-processed vectors of the multi-source data of the distribution transformer system and the charging load are arranged in the order of timestamps to form a high-dimensional time series feature sequence, which is input into the hybrid time series modeling network; The dynamic segmentation of the time of the charging load linkage of the distribution transformer system throughout the day includes generating a predicted load curve on the time scale of the whole day after performing time series modeling on the multi-source data fusion characteristics. The predicted load curve includes the predicted load value at the time point. Based on the predicted load curve, the curve is segmented using a dynamic density clustering algorithm; The dynamic density clustering algorithm calculates the load change rate between adjacent time points through a sliding window, and dynamically identifies the inflection points of load change by combining local density and distance threshold, dividing the whole day into several time periods with load change characteristics.

4. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 3 is characterized in that: The time division of the time-sharing regulation includes the introduction of a dynamic adjustment mechanism. The dynamic adjustment mechanism refers to training a Q network model using the state between the current predicted load curve and the historical load pattern as input. The Q network model is used to online adjust the distance threshold and local density parameters in the dynamic density clustering algorithm, so that the time period division result is adaptively optimized according to the real-time load change trend, completing the dynamic segmentation of the charging load linkage time of the distribution transformer system.

5. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 4 is characterized in that: The monitoring of the distribution transformer system includes deploying edge computing units based on time series load forecasting and dynamic segmentation based on multi-source data fusion features; and performing real-time monitoring of the operating status of the distribution transformer system. The system collects operating parameters including the load factor, three-phase current imbalance, voltage fluctuation, ambient temperature, and historical load trends of distribution transformers. Timestamps are aligned with time segment labels through a unified clock mechanism to form a distribution transformer operating status data stream with time segment labels. The user interaction platform connected via the communication interface obtains priority information for the charging process. This priority information includes the battery's current state of charge, the user's scheduled charging time, the user's historical credit score, and the charging type. This priority information is processed by a feature encoder to generate a priority feature vector, which is then integrated with the load forecast value for the current time segment as input for the control decision. The fused load forecast value for the current time segment is input into the dynamic control decision engine of the hybrid strategy, which consists of two subsystems running in parallel: Prioritization subsystem; Power distribution and access control subsystem.

6. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 5 is characterized in that: The modeling of the control objectives includes determining a set of interrelated control objectives during the execution of the closed-loop linkage control strategy based on the operating characteristics of the distribution transformer system and the charging load management requirements, wherein the control objectives include a distribution transformer load safety constraint objective, a three-phase load balance objective, a user charging demand satisfaction objective, a power regulation smoothness objective, and an equipment operation energy efficiency objective; According to the unified control target modeling formula of the control target, an optimized objective function is constructed through weighted normalization and multi-objective coupling mechanism to generate a model for the control target. The unified control target modeling formula of the control target is: ； in, represents unified control target modeling, represents the weight of the distribution transformer load safety constraint target, represents the normalized distribution transformer load safety constraint objective function, represents the weight of the three-phase load balancing objective, represents the normalized three-phase load balancing objective function, represents the weight of the user's charging demand satisfaction target, represents the normalized user charging demand satisfaction objective function, represents the weight of the power regulation smoothness objective, represents the normalized power regulation smoothness objective function, Indicates the weight of the equipment operation energy efficiency target, represents the dynamic coupling matrix between targets, represents the weight vector, Represents the target coupling term.

7. The time-sharing and segmented control method linked to the charging load of the distribution transformer system according to claim 6 is characterized in that: The optimization of priority information and adjustment of the weight of priority information include interacting the unified control target modeling with the dynamic coupling matrix between targets, evaluating the conflicts between the control targets under the current weight distribution, and generating control instructions based on the evaluation of the coupling items of the control targets under the current weight distribution, combined with the load forecast value and the distribution transformer operating status of the current time period, to complete the adaptive adjustment in the time-sharing and section-by-section control process.

8. A time-sharing and segmented control system linked to the charging load of a distribution transformer system, based on the time-sharing and segmented control method linked to the charging load of a distribution transformer system according to any one of claims 1 to 7, characterized in that: include, The dynamic segmentation and time division module collects multi-source data on charging loads, establishes the extracted multi-source data fusion features, and analyzes the time series load forecasting mechanism of the multi-source data fusion features. Based on the time series load forecasting mechanism, it dynamically segments the time of the charging load linkage of the distribution transformer system throughout the day, providing a time division basis for time-sharing control; The closed-loop linkage control module between the distribution transformer system and the charging load monitors the distribution transformer system based on dynamic segmentation and time division, obtains priority information of the charging process, and dynamically adjusts the output power and access sequence of the charging process according to the load forecast value and priority information of the current time period, thus completing the closed-loop linkage control between the distribution transformer system and the charging load; The adaptive adjustment module in the time-sharing and segmented control process models the control objectives, optimizes the priority information and adjusts the weight of the priority information during the execution of the closed-loop linkage control strategy, generates control decisions that adapt to the current operating environment, and completes the adaptive adjustment in the time-sharing and segmented control process.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the time-sharing and section-by-section control method linked to the charging load of the distribution transformer system according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the time-sharing and section-by-section control method linked to the charging load of the distribution transformer system according to any one of claims 1 to 7 are implemented.

Images