Dual transformer parallel operation load adaptive regulation method

Abstract

The application discloses a double-transformer parallel operation load adaptive regulation method and relates to the technical field of double-transformer parallel operation, and comprises the following steps: in the case that two transformers are in the state that regulation actions have been executed but load difference values have not changed correspondingly, variation intensity features of regulation action trajectories and response lag features of load difference value evolution trajectories are extracted, the variation intensity features and the response lag features are subjected to unified scale mapping, and a regulation response deviation degree parameter is constructed; the regulation response deviation degree parameter is mapped to a continuous interval, regulation effectiveness determination is completed according to the position of the regulation response deviation degree parameter in the continuous interval, and whether secondary regulation or regulation strategy adjustment is performed is determined according to the regulation response deviation degree parameter. The application solves the problem that regulation effectiveness cannot be determined in the case that regulation actions have been executed but loads have not responded, and realizes dynamic determination and progressive regulation based on deviation degree.

Description

Technical Field

[0001] This invention relates to the field of parallel operation technology of dual transformers, and more specifically to a method for adaptive load adjustment of parallel operation of dual transformers. Background Technology

[0002] Dual-transformer parallel operation load adaptive regulation refers to a control method in which, in a system where two transformers are connected in parallel to supply power to the same electrical load, the load distribution between the two transformers is dynamically adjusted based on a preset control strategy and software algorithm, by real-time monitoring of the operating status of each transformer and the load side, allowing the load to be adaptively and reasonably shared according to changes in operating conditions. In existing technologies, this regulation typically relies on a complete closed-loop control system. First, sensors installed on the transformer and bus sides continuously collect parameters such as voltage, current, active power, reactive power, and load rate, which are then transmitted to a centralized or distributed control unit via a communication network. Second, the algorithm in the control unit... The module processes and identifies the status of the collected data, calculates the target allocation scheme based on preset load allocation rules (such as rated capacity ratio, impedance characteristics, or operating status weight), and then converts the calculation results into control commands, which are then adjusted by the actuator. For example, the output voltage is adjusted by the on-load tap changer, the operating mode of the control equipment is changed, or relevant electrical parameters are adjusted, thereby realizing the redistribution of the load between the two transformers. Subsequently, the system continuously monitors the adjusted operating status and inputs the feedback information into the control unit for strategy updates and dynamic optimization. Overall, it forms an adaptive adjustment process consisting of data acquisition, communication transmission, status analysis, decision calculation, execution control, and feedback updates.

[0003] The existing technology has the following shortcomings: In the adaptive load regulation process of dual transformers operating in parallel, after the system completes a load distribution adjustment, the regulation actions of the two transformers have been executed. However, due to the coupling relationship on the load side of the parallel system and the influence of dynamic changes in electrical parameters, the regulation effect cannot be effectively transmitted to the load distribution result in a short period of time, resulting in a state where the load difference does not change accordingly. Existing technology cannot determine whether the regulation is effective and whether a secondary regulation or adjustment strategy is needed based on the degree of deviation of the regulation response of the two transformers when the regulation action has been executed but the load difference has not changed accordingly. This will cause the system to misjudge the regulation process as effective regulation and continue to use the original control strategy, thus causing the ineffective regulation to accumulate continuously. Consequently, the load distribution result deviates from the expected target, affecting the accuracy and stability of the adaptive load regulation of dual transformers operating in parallel.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method for adaptive load adjustment of dual transformers operating in parallel, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for adaptive load adjustment of dual transformers operating in parallel, specifically including the following steps: S1. Obtain the regulation action trajectory formed during the load distribution adjustment process, and extract the load difference evolution trajectory of the two transformers within the corresponding time interval. Align the regulation action trajectory with the load difference evolution trajectory in time and match the intervals to form a coupled sequence relationship. S2. Calculate the consistency of change direction and the matching degree of change response between the regulation action trajectory and the load difference evolution trajectory based on the coupling sequence relationship. Determine whether there is a situation where the two transformers have performed the regulation action but the load difference has not changed accordingly based on the combination relationship between the consistency of change direction and the matching degree of change response. S3. In the case where two transformers have performed regulation actions but the load difference has not changed accordingly, extract the change intensity characteristics of the regulation action trajectory and the response lag characteristics of the load difference evolution trajectory, perform a unified scale mapping on the change intensity characteristics and the response lag characteristics, and construct a parameter for the degree of deviation of the regulation response. S4. Map the adjustment response deviation parameter to a continuous interval, determine the adjustment effectiveness based on the position of the adjustment response deviation parameter in the continuous interval, and determine whether to perform secondary adjustment or adjust the adjustment strategy based on the adjustment response deviation parameter. S5. Based on the determination of the effectiveness of regulation, update the regulation trajectory and combine it with the load difference evolution trajectory to form a new coupling sequence relationship, and perform progressive adjustment to complete the load distribution regulation.

[0007] Preferably, S1 is as follows: During the load distribution adjustment process, a continuous data sequence of adjustment commands over time is obtained, and the changes in adjustment commands are continuously connected in chronological order to construct the adjustment action trajectory. The load data sequences of two transformers are collected within the time interval covered by the corresponding adjustment trajectory. The load data of the two transformers are then calculated to obtain the load difference evolution trajectory by continuously connecting the difference changes in chronological order. Using time sequence as a unified reference, time alignment processing is performed on the regulation trajectory and the load difference evolution trajectory. The time interval is divided according to continuous time intervals. Within each time interval, the regulation trajectory and the load difference evolution trajectory are matched. A one-to-one mapping is established according to the time correspondence to form a coupled sequence relationship.

[0008] Preferably, S2 is as follows: Based on the coupling sequence relationship, the change direction of adjacent time points of the regulation trajectory and the change direction of adjacent time points of the load difference evolution trajectory are obtained in each time interval. The two types of change directions in each time interval are compared interval by interval. Time intervals with the same change direction are marked, and time intervals with opposite change directions or one of them not changing are marked. The marking results of all time intervals are statistically processed, and the proportion of the number of time intervals with the same change direction in the total number of time intervals is used as the consistency of the change direction between the regulation trajectory and the load difference evolution trajectory. After the consistency of the change direction is calculated, the change amplitude of the regulation trajectory and the change amplitude of the load difference evolution trajectory in each time interval are obtained. The corresponding difference is calculated for the change amplitude in each time interval to form an amplitude difference sequence. The amplitude difference sequence is accumulated in chronological order and combined with the cumulative value of the change amplitude of the regulation trajectory in all time intervals to normalize the amplitude difference sequence. The result after normalization is used as the change response matching degree between the regulation trajectory and the load difference evolution trajectory. The determination is based on the combination of the consistency of change direction and the matching degree of change response. When the consistency of change direction is greater than the preset consistency threshold and the matching degree of change response is less than the preset matching threshold, it is determined that there is a situation where the two transformers have performed the adjustment action but the load difference has not changed accordingly.

[0009] Preferably, S3 specifically includes the following steps: S301. When two transformers have performed adjustment actions but the load difference has not changed accordingly, obtain the change amplitude sequence of the adjustment action trajectory in each time interval, and perform interval cumulative processing on the change amplitude sequence. Use the cumulative value of the change amplitude in each time interval as the change intensity feature of the adjustment action trajectory. At the same time, obtain the time difference between the response start time of the load difference evolution trajectory and the change start time of the adjustment action trajectory in each time interval, and perform interval statistical processing on the time difference. Use the statistical result as the response lag feature of the load difference evolution trajectory. S302. Perform unified scale mapping processing on the change intensity feature and the response lag feature, divide the change intensity feature and the response lag feature into segments according to the preset numerical interval, and convert the segment mapping result into a standardized numerical sequence within the same numerical interval, and use the standardized numerical sequence as the unified scale representation of the change intensity feature and the response lag feature. S303. Based on the change intensity characteristics and response lag characteristics after unified scale representation, they are combined in the order of time intervals. The change intensity characteristics and response lag characteristics in each time interval are weighted and the weighted results are accumulated within the interval. The accumulated results are used as parameters to adjust the degree of response deviation.

[0010] Preferably, S303 is as follows: Based on the change intensity features and response lag features represented by a unified scale, they are arranged one-to-one according to the time interval order to form a time-order aligned combination sequence, and the change intensity features and response lag features in each time interval are paired and labeled. The intensity of change and the response lag characteristics within each time interval are weighted, with the intensity of change corresponding to the first weight value and the response lag characteristic corresponding to the second weight value. The intensity of change is multiplied by the first weight value and the response lag characteristic is multiplied by the second weight value. The product results within the same time interval are summed to form a weighted result sequence within each time interval. The weighted result sequence within each time interval is accumulated sequentially according to the time interval, the accumulated result is continuously updated, and the final accumulated value is used as the parameter for adjusting the deviation of the response.

[0011] Preferably, S4 specifically includes the following steps: S401. Divide the adjustment response deviation parameter into intervals according to a preset continuous value range, divide the preset continuous value range into multiple adjacent and non-overlapping continuous intervals, and map the adjustment response deviation parameter to the corresponding continuous interval, using the position of the continuous interval where the adjustment response deviation parameter is located as the interval positioning result. S402. Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and each continuous interval is classified according to the numerical size rule. A one-to-one correspondence is established between each continuous interval and the corresponding interval classification. The corresponding interval classification is matched according to the interval positioning result, and the matched interval classification is used as the adjustment effectiveness judgment result. S403. Based on the position of the adjustment response deviation parameter in the continuous interval and the adjustment effectiveness determination result, the continuous interval is divided into different adjustment processing intervals, and corresponding processing is performed according to the interval division result. Specifically, if the parameter is located in the first adjustment processing interval, the current adjustment strategy is maintained; if the parameter is located in the second adjustment processing interval, a secondary adjustment is performed; and if the parameter is located in the third adjustment processing interval, the adjustment strategy is adjusted.

[0012] Preferably, S402 is as follows: Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and the continuous intervals are sorted according to the numerical boundary order of the continuous intervals to form a continuous interval sequence arranged according to the numerical size. The continuous interval sequence is divided into multiple interval classification categories according to the numerical size rule. Each continuous interval is then associated with a corresponding interval classification category, so that each continuous interval corresponds to only one interval classification category. Based on the position of the interval location result in the continuous interval sequence, the corresponding continuous interval is matched, and the interval classification category is determined by the correspondence between the continuous interval and the interval classification category. The interval classification category is divided into effective regulation category, deviation regulation category and ineffective regulation category, and the category type corresponding to the matched interval classification category is used as the regulation effectiveness judgment result.

[0013] Preferably, S5 is as follows: Based on the results of the regulation effectiveness assessment, the regulation trajectory is updated. According to the category type corresponding to the regulation effectiveness assessment results, the change amplitude of the regulation trajectory in each time interval is corrected, and the change direction of the regulation trajectory is adjusted accordingly to form the updated regulation trajectory. The updated regulation trajectory and the load difference evolution trajectory are time-aligned, and the updated regulation trajectory and the load difference evolution trajectory are interval-matched according to the time interval order. A one-to-one correspondence between the regulation trajectory and the load difference evolution trajectory is established in each time interval, forming a new coupling sequence relationship. Based on the new coupling sequence relationship, the updated adjustment trajectory is progressively adjusted according to the time interval order. The change in the adjustment trajectory in each time interval is progressively corrected interval by interval. After the correction is completed in each time interval, the status of the adjustment trajectory is updated until the progressive adjustment of all time intervals is completed to complete the load distribution adjustment.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention constructs a coupled sequence relationship consisting of the regulation action trajectory and the load difference evolution trajectory, achieving a precise correspondence between regulation behavior and load response in the time dimension. Furthermore, it introduces a combined judgment of the consistency of change direction and the matching degree of change response, elevating the identification of the regulation action execution status from "whether it was executed" to "whether it was effectively executed." Based on this, by extracting change intensity features and response hysteresis features and performing unified scale mapping, a parameter for the degree of regulation response deviation is constructed. This transforms the originally difficult-to-quantify regulation failure phenomenon into a measurable continuous parameter. Then, through continuous interval mapping and interval grading, a structured judgment of regulation effectiveness is achieved. This allows for accurate identification of the specific degree of regulation response deviation even when the regulation action has been executed but the load difference has not changed accordingly, avoiding misjudging ineffective regulation as effective regulation, and giving the system higher state identification accuracy and judgment reliability.

[0015] 2. This invention dynamically updates the regulation trajectory based on the regulation effectiveness determination result, and reconstructs the coupling sequence relationship with the load difference evolution trajectory. It performs progressive adjustments within the time interval, enabling the regulation behavior to be gradually corrected as the degree of deviation changes, forming a continuous closed-loop adaptive regulation process. By transforming the regulation process from a single execution to a progressive correction mechanism, the problem of short-term failure of the regulation effect under coupling influence being continuously amplified can be avoided. Simultaneously, based on different deviation categories, it executes different strategies, such as maintaining the current regulation strategy, secondary regulation, or regulation strategy adjustment, giving the regulation decision a hierarchical response capability. This improves the precision and dynamic adaptability of the adaptive regulation of the load in parallel operation of dual transformers, enhancing the stability of the system and the accuracy of load distribution control under complex operating conditions. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0018] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0019] This invention provides, for exampleFigure 1 The adaptive load adjustment method for dual transformers operating in parallel, as shown, specifically includes the following steps: S1. Obtain the regulation action trajectory formed during the load distribution adjustment process, and extract the load difference evolution trajectory of the two transformers within the corresponding time interval. Align the regulation action trajectory with the load difference evolution trajectory in time and match the intervals to form a coupled sequence relationship. In this embodiment, S1 specifically refers to: During the load distribution adjustment process, a continuous data sequence of adjustment commands over time is obtained, and the changes in adjustment commands are continuously connected in chronological order to construct the adjustment action trajectory. In the adaptive load regulation process of dual transformers operating in parallel, load distribution adjustment is typically achieved by changing the load ratio borne by the two transformers. This adjustment is manifested in the form of regulation commands, such as changing the tap position, adjusting control quantities, or altering the power distribution ratio. During implementation, the value of the regulation command and its corresponding time marker can be recorded at each regulation moment. For example, control output quantities can be collected at fixed time intervals, and the regulation commands at each moment can be sorted chronologically to form a continuously changing data sequence. These discrete sampling points are then connected sequentially in chronological order to preserve the relationship between adjacent moments, thus forming a trajectory that reflects the continuous evolution trend of the regulation process. For instance, if the regulation command gradually increases from a small value to a large value over a certain period, continuous sampling and connection can yield a monotonically changing curve, which is the regulation trajectory, used to characterize the continuous change process of the regulation behavior over time. The reason for using this approach is that a single-moment regulation command cannot reflect the overall trend of the regulation behavior, while the trajectory formed by continuous connection can fully describe the direction, magnitude, and continuity of the regulation process, thus providing a foundation for subsequent analysis of the relationship between regulation behavior and load changes.

[0020] The load distribution adjustment process refers to the dynamic change of the load ratio of the two transformers in parallel operation, based on the control objective. The adjustment command refers to the control quantity used to drive the load distribution change during this process, which can be represented by control signals or numerical changes. The continuous data sequence formed over time refers to the data set arranged chronologically after collecting adjustment commands at multiple consecutive moments. The temporal order refers to arranging the data according to the order of acquisition, giving the data a clear temporal evolution relationship. Continuous connection refers to connecting data points from adjacent moments in chronological order, transforming discrete data into a continuously changing representation. The adjustment trajectory refers to the change path formed by the continuously connected data sequence, reflecting the dynamic trend of the adjustment command throughout the adjustment process. Through the synergistic construction of these elements, discrete adjustment behavior can be transformed into a trajectory expression with continuity and directionality, thus providing a unified data foundation for subsequent adjustment effect analysis.

[0021] The load data sequences of two transformers are collected within the time interval covered by the corresponding adjustment trajectory. The load data of the two transformers are then calculated to obtain the load difference evolution trajectory by continuously connecting the difference changes in chronological order. Within the time range corresponding to the regulating trajectory, the operating load of the two transformers can be synchronously collected. The collected data can include quantities reflecting the load status, such as current, active power, or load rate, and are arranged into a time-ordered load data sequence according to a uniform sampling interval. Based on this, the load data of the two transformers at the same time marker are mapped one-to-one, and the difference is calculated for the load data at the corresponding time point. This yields a difference sequence reflecting the load distribution differences between the two transformers. Subsequently, the difference sequences are continuously connected in chronological order, preserving the changing relationships between each time point, thus forming a load difference evolution trajectory with continuous variation characteristics. For example, when the load of one transformer gradually increases while the other remains relatively stable, by calculating and connecting the differences moment by moment, a continuously changing curve can be obtained. This curve can intuitively reflect the changing trend of the load difference between the two transformers. Through this implementation, discrete load measurement data can be transformed into a continuous changing path, allowing the load distribution relationship to be fully expressed in the time dimension.

[0022] The time interval covered by the regulation action trajectory refers to the time range from the start of the regulation behavior to the end of the regulation effect, used to define the time boundary for load data acquisition; the load data sequence of the two transformers refers to the set of load data collected in chronological order within this time range; difference calculation refers to calculating the difference between the load data of the two transformers under the same time marker to obtain the difference reflecting the load distribution relationship; difference change refers to the continuous change process of the difference in the time dimension; continuous connection refers to connecting the differences of adjacent time points in chronological order, transforming discrete data into a continuously changing representation; the load difference evolution trajectory refers to the path formed by the difference changes after continuous connection, used to describe the dynamic evolution of the load difference between the two transformers over time. Through the synergistic construction of these elements, a unified and continuous expression of difference change can be formed, providing a basic support for subsequent regulation response analysis.

[0023] Using time sequence as a unified reference, time alignment processing is performed on the regulation trajectory and the load difference evolution trajectory. The time interval is divided according to continuous time intervals. Within each time interval, the regulation trajectory and the load difference evolution trajectory are matched. A one-to-one mapping is established according to the time correspondence to form a coupled sequence relationship.

[0024] Using time sequence as a unified reference, the data in the regulation action trajectory and the load difference evolution trajectory can be mapped onto the same time axis. Specifically, the time markers in the two types of trajectories can be standardized, for example, by converting the sampling time to a unified time scale and reordering them according to chronological order. Then, time alignment is performed on the two types of trajectories to ensure a one-to-one correspondence between the regulation action changes and load difference changes at the same time scale. When time offsets exist, alignment can be achieved through time interpolation or selection of nearby time points. After time alignment, the entire time range is divided into several continuous time intervals, for example, according to fixed time intervals. The time interval is divided into windows or change phases, and within each time interval, a control action trajectory segment and a load difference evolution trajectory segment are extracted. Then, interval matching is performed on the two types of trajectory segments to establish a correlation between the control behavior and the corresponding load difference change in each time interval. For example, if the control action trajectory shows a continuous increase in a certain time interval, while the load difference evolution trajectory changes little, interval matching can identify the inconsistency between the control behavior and the load response. Finally, by sequentially connecting the matching results of all time intervals, a set of time-arranged coupled sequence relationships is formed to describe the overall correspondence between the control behavior and the load response.

[0025] Among these elements, unified reference refers to using the same time scale as a benchmark to ensure consistent time signatures for data from different sources; the regulation trajectory and load difference evolution trajectory represent the paths of regulation behavior change and load difference change, respectively, both exhibiting temporal continuity; time alignment processing involves mapping the two types of trajectories to the same time scale, ensuring a matching relationship at every time point; continuous time intervals refer to several adjacent and uninterrupted time segments within the overall time range; interval division refers to segmenting the overall time range according to chronological order; interval matching refers to correspondingly associating the two types of trajectory segments within the same time interval; one-to-one mapping refers to establishing a correspondence between changes in regulation and changes in load difference along the time dimension; and coupled sequence relationships refer to paired data sets arranged in chronological order, used to characterize the dynamic correlation structure between regulation behavior and load response. By combining these elements, the change processes from two different sources can be unified and expressed within the same time frame, thus providing a consistent data foundation for subsequent analysis.

[0026] S2. Calculate the consistency of change direction and the matching degree of change response between the regulation action trajectory and the load difference evolution trajectory based on the coupling sequence relationship. Determine whether there is a situation where the two transformers have performed the regulation action but the load difference has not changed accordingly based on the combination relationship between the consistency of change direction and the matching degree of change response. In this embodiment, S2 specifically refers to: Based on the coupling sequence relationship, the change direction of adjacent time points of the regulation trajectory and the change direction of adjacent time points of the load difference evolution trajectory are obtained in each time interval. The two types of change directions in each time interval are compared interval by interval. Time intervals with the same change direction are marked, and time intervals with opposite change directions or one of them not changing are marked. The marking results of all time intervals are statistically processed, and the proportion of the number of time intervals with the same change direction in the total number of time intervals is used as the consistency of the change direction between the regulation trajectory and the load difference evolution trajectory. The consistency of change direction can be calculated by analyzing each time interval in the coupled sequence relationship segment by segment. Specifically, in each time interval, the trend of the regulation trajectory between adjacent time points is extracted. For example, it is determined whether the values ​​of two adjacent time points are rising, falling, or remaining unchanged. At the same time, the trend of the load difference evolution trajectory between adjacent time points is extracted in the same time interval, and the two types of trends are recorded accordingly. Then, the two types of change directions are compared one by one in each time interval. When both types of change directions are rising or both are falling, the time interval is marked as consistent in direction. When one type is rising and the other is falling, or one type changes and the other remains unchanged, the time interval is marked as inconsistent in direction. After marking all time intervals, the number of time intervals with consistent direction is counted and normalized according to the total number of time intervals, so that the proportion of consistent direction reflects the consistency of the overall change trend. For example, in multiple consecutive time intervals, if the two types of trajectories show synchronous rise or synchronous fall in most intervals, the consistency is high. However, if the trajectories show inconsistent directions or no change in one direction in most intervals, the consistency is low. This process can transform discrete directional change information into an overall consistency measure, thereby providing a basis for subsequent judgment on the relationship between adjustment behavior and load response.

[0027] The direction of change of adjacent time points of the regulation trajectory and the direction of change of adjacent time points of the load difference evolution trajectory within each time interval refer to the numerical change trend between two adjacent time points within the same time interval, used to characterize the directionality of change; the two types of change directions within each time interval refer to the set of change trends corresponding to the regulation trajectory and the load difference evolution trajectory respectively within the same time interval; interval-by-interval comparison refers to a one-to-one comparison of the two types of change directions within each time interval; time intervals with the same change direction refer to time intervals where both types of change trends are in the same direction; time intervals with opposite change directions or where one does not change refer to time intervals where the two types of change trends have directional differences or only one type changes; the proportion of time intervals with the same change direction to the total number of time intervals refers to the ratio of the number of intervals with consistent direction to the total number of intervals, used to reflect the overall consistency; the consistency of change direction between the regulation trajectory and the load difference evolution trajectory refers to a metric based on the above proportions, used to describe the level of directional consistency between the two types of trajectories in the time dimension.

[0028] After the consistency of the change direction is calculated, the change amplitude of the regulation trajectory and the change amplitude of the load difference evolution trajectory in each time interval are obtained. The corresponding difference is calculated for the change amplitude in each time interval to form an amplitude difference sequence. The amplitude difference sequence is accumulated in chronological order and combined with the cumulative value of the change amplitude of the regulation trajectory in all time intervals to normalize the amplitude difference sequence. The result after normalization is used as the change response matching degree between the regulation trajectory and the load difference evolution trajectory. In practical implementation, the degree of matching of change response can be calculated by performing interval-by-interval quantitative analysis of the change amplitudes of the two types of trajectories within each time interval. Specifically, the numerical change amplitude between adjacent time points of the regulation trajectory is extracted within each time interval, and the numerical change amplitude of the load difference evolution trajectory within the corresponding time interval is also extracted. The corresponding differences of the two types of change amplitudes are calculated within the same time interval, thus forming a set of amplitude difference sequences arranged by time. Subsequently, the amplitude difference sequences are accumulated interval by interval according to the time order, so that the local differences are gradually summarized into the overall difference. Furthermore, the accumulated amplitude difference is proportionalized by combining the cumulative value of the change amplitude of the regulation trajectory within all time intervals, so that regulation processes of different scales are comparable. For example, if the regulating action trajectory changes significantly over a certain period, while the load difference evolution trajectory changes relatively little, calculating and accumulating the difference across intervals will result in a large cumulative difference. After proportionalizing this difference with the overall change amplitude of the regulating action trajectory, a low degree of matching can be obtained, indicating that the regulating behavior has not been effectively transmitted to the load change. Conversely, when the change amplitudes of the two types of trajectories are similar across intervals, the cumulative difference is small, and after proportionalization, a higher degree of matching is obtained, indicating a more consistent regulating response. This implementation transforms local amplitude differences into an overall degree of matching, providing a quantitative basis for subsequent judgments of regulating effectiveness.

[0029] The variation amplitudes of the regulation trajectory and the load difference evolution trajectory within each time interval refer to the magnitude of the numerical change between adjacent time points within each time interval, used to characterize the intensity of the change. Corresponding difference calculation refers to obtaining the difference between the two types of variation amplitudes within the same time interval, reflecting the degree of deviation between regulation change and load response. The amplitude difference sequence refers to the set of differences between each time interval arranged in chronological order, used to describe the distribution of deviation over time. Accumulation processing refers to accumulating the amplitude differences interval by interval in chronological order to form the overall deviation. The cumulative value of the variation amplitude of the regulation trajectory across all time intervals refers to the sum of the variation intensity of the regulation trajectory over the entire time range, used as a normalization reference. Normalization processing refers to proportionalizing the cumulative deviation with the overall variation intensity to give the results a uniform scale. The matching degree of change response between the regulation trajectory and the load difference evolution trajectory refers to the quantitative result formed through normalization processing, used to describe the degree of matching between changes in regulation behavior and changes in load response. Through the coordinated construction of these elements, a unified quantitative expression from local differences to overall matching relationships can be achieved.

[0030] The determination is made based on the combination of the consistency of change direction and the matching degree of change response. When the consistency of change direction is greater than the preset consistency threshold and the matching degree of change response is less than the preset matching threshold, it is determined that there is a situation where the two transformers have performed the adjustment action but the load difference has not changed accordingly.

[0031] A comprehensive judgment can be achieved by using the consistency of change direction and the matching degree of change response as two interrelated criteria. Specifically, the consistency of change direction is first compared with a pre-set consistency threshold to determine whether the adjustment trajectory and the load difference evolution trajectory maintain the same overall trend. Then, the matching degree of change response is compared with a pre-set matching threshold to determine whether the adjustment action effectively transmits the change in load difference in terms of amplitude. Based on this, the two judgment results are combined for analysis. Only when the consistency of change direction exceeds the consistency threshold and the matching degree of change response is lower than the matching threshold is the adjustment action considered to be consistent with the load change in direction but without a corresponding response in amplitude. For example, if the adjustment trajectory continuously rises over a certain period, and the load difference evolution trajectory also shows an upward trend, the consistency of direction is high. However, if the amplitude of the load difference change remains small, the matching degree is low. When both conditions are met simultaneously, it can be determined that the adjustment action has not effectively affected load distribution. This dual-indicator joint judgment avoids misjudgments caused by relying on only a single indicator, making the judgment results more stable and reliable.

[0032] The preset consistency threshold is a benchmark value used to define the degree of consistency in the direction of change. It can be set based on historical operating data or experience and is used to distinguish the boundary between consistent and inconsistent directions. The preset matching threshold is a benchmark value used to define the degree of change response and is used to distinguish between sufficient and insufficient response. The consistency of change direction reflects the degree of consistency between the change direction of the regulation action trajectory and the change difference evolution trajectory in the time dimension. The matching degree of change response reflects the degree of correspondence between the two types of trajectories in terms of change amplitude. The combined relationship refers to the logical relationship of jointly judging the consistency of change direction and the matching degree of change response. It identifies specific operating states by simultaneously satisfying the judgment conditions of consistency of direction and insufficient response. Through the synergistic effect of these quantities, a fine-grained basis for judging the relationship between regulation behavior and load response can be formed, thereby supporting subsequent regulation decisions.

[0033] S3. In the case where two transformers have performed regulation actions but the load difference has not changed accordingly, extract the change intensity characteristics of the regulation action trajectory and the response lag characteristics of the load difference evolution trajectory, perform a unified scale mapping on the change intensity characteristics and the response lag characteristics, and construct a parameter for the degree of deviation of the regulation response. In this embodiment, S3 specifically includes the following steps: S301. When two transformers have performed adjustment actions but the load difference has not changed accordingly, obtain the change amplitude sequence of the adjustment action trajectory in each time interval, and perform interval cumulative processing on the change amplitude sequence. Use the cumulative value of the change amplitude in each time interval as the change intensity feature of the adjustment action trajectory. At the same time, obtain the time difference between the response start time of the load difference evolution trajectory and the change start time of the adjustment action trajectory in each time interval, and perform interval statistical processing on the time difference. Use the statistical result as the response lag feature of the load difference evolution trajectory. The intensity of change can be extracted by segmenting and quantifying the change process of the regulation trajectory over time, while the response lag can be characterized by combining the response time difference of the load difference evolution trajectory. Specifically, firstly, according to the divided time intervals, the numerical changes of the regulation trajectory between adjacent time points within each time interval are calculated, forming a sequence of change amplitudes within each time interval. Then, the change amplitudes within the same time interval are accumulated, transforming discrete changes into overall changes within the interval. This accumulated value represents the intensity of the regulation behavior within that time interval. Simultaneously, the time point when the regulation trajectory begins to change significantly and the time point when the load difference evolution trajectory begins to show response changes are identified within each time interval. The time difference between these two points is calculated and statistically summarized across multiple time intervals, for example, by accumulating or averaging in chronological order, thus forming a quantitative result reflecting the degree of response delay. For example, if the regulating effect trajectory begins to rise continuously within a certain time interval, while the load difference evolution trajectory only begins to change after several time units, this time difference can reflect the response lag. By performing statistical analysis over multiple time intervals, the overall degree of lag can be obtained. This approach allows for the simultaneous characterization of both the intensity of the regulating effect and the delay in the response, thus providing a two-dimensional quantitative basis for constructing subsequent deviation assessments.

[0034] The sequence of changes in the amplitude of the regulating action trajectory within each time interval refers to the set of numerical changes between adjacent time points within each time interval, used to reflect the intensity of local changes; interval cumulative processing refers to the continuous accumulation of the amplitude of changes within the same time interval, transforming multiple discrete change values ​​into an overall change; the cumulative value of the amplitude of changes within each time interval refers to the result of accumulating all amplitudes of changes within that time interval, used to characterize the overall intensity of the regulating behavior within that interval; the intensity characteristic of the regulating action trajectory refers to the intensity characterization composed of the cumulative changes in each time interval, used to describe the strength distribution of the regulating behavior in the time dimension; the time difference between the response start time of the load difference evolution trajectory and the change start time of the regulating action trajectory within each time interval refers to the time interval between the start of the regulating change and the start of the response change within the same time interval, used to reflect the degree of lag in the response; interval statistical processing refers to the unified processing of time differences within multiple time intervals, so that discrete time differences form an overall distribution or cumulative result; the response lag characteristic of the load difference evolution trajectory refers to the quantitative index formed by the statistically processed time difference results, used to describe the degree of delay in the load response relative to the regulating behavior. By combining these elements, a structured expression of the relationship between regulation intensity and response hysteresis can be achieved.

[0035] S302. Perform unified scale mapping processing on the change intensity feature and the response lag feature, divide the change intensity feature and the response lag feature into segments according to the preset numerical interval, and convert the segment mapping result into a standardized numerical sequence within the same numerical interval, and use the standardized numerical sequence as the unified scale representation of the change intensity feature and the response lag feature. Unified scaling can be achieved by uniformly transforming the intensity of change and the response lag at the numerical level. Specifically, firstly, the ranges of both the intensity and response lag characteristics across all time intervals are statistically analyzed. Then, based on the distribution of these two types of characteristics, several predefined numerical intervals are defined, for example, dividing them into multiple continuous segments according to their magnitude. Next, the intensity and response lag characteristics are mapped to corresponding interval levels; for example, smaller changes are mapped to lower levels, and larger changes to higher levels. The same interval division and level mapping are applied to the response lag characteristics. Subsequently, the mapped level values ​​are further transformed to a unified numerical range, allowing both types of characteristics to be represented on the same numerical scale. For example, different interval levels are converted into continuous values ​​within a unified range, thus forming a standardized numerical sequence. For instance, if the intensity of change is large and the response lag is long within a certain time interval, after segmented mapping, both correspond to higher levels, and after unified range transformation, they can be directly compared and combined. The reason for adopting this implementation is that the change intensity feature and the response hysteresis feature have different sources and different dimensions, and direct comparison will produce bias. By using a unified scale mapping, the influence of different dimensions can be eliminated, making the two types of features comparable and fusionable.

[0036] Unified scaling refers to the process of converting feature values ​​from different sources and with different dimensions into values ​​that can be compared within the same numerical range. Pre-defined numerical intervals refer to multiple continuous numerical segments pre-divided according to the feature value range, used to stratify the original data. Segmented mapping refers to mapping the original feature values ​​to corresponding interval levels or interval identifiers according to their respective numerical intervals. The same numerical interval range refers to uniformly converting different features to the same value range, giving all features a consistent scale standard. Standardized numerical sequences refer to the set of numerical values ​​arranged in chronological order after interval mapping and range conversion. Unified scaling representation of change intensity features and response lag features means that after the above processing, the two types of features are expressed at the same scale, allowing for direct combination and calculation in subsequent processes. Through these processes, multidimensional features that were originally not directly comparable can be transformed into a unified expression, providing a foundation for subsequently constructing parameters that regulate the degree of response deviation.

[0037] S303. Based on the change intensity characteristics and response lag characteristics after unified scale representation, they are combined in the order of time intervals. The change intensity characteristics and response lag characteristics in each time interval are weighted and the weighted results are accumulated within the interval. The accumulated results are used as parameters to adjust the degree of response deviation.

[0038] In this embodiment, S303 specifically refers to: Based on the change intensity features and response lag features represented by a unified scale, they are arranged one-to-one according to the time interval order to form a time-order aligned combination sequence, and the change intensity features and response lag features in each time interval are paired and labeled. Given that the intensity of change and the hysteresis of response, after being represented by a unified scale, already have the same numerical range, the two types of feature sequences are synchronously sorted based on the time interval order. Within each time interval, the corresponding intensity of change and hysteresis feature values ​​are extracted. The two types of features within the same time interval are paired one-to-one, forming a set of data pairs containing information on intensity of change and hysteresis for each time interval. Subsequently, the data pairs formed in each time interval are arranged sequentially according to time chronological order to construct a continuous combined sequence. Each data pair is assigned a corresponding time interval identifier to maintain the structural consistency of the sequence. For example, if the intensity of change shows a gradually increasing trend while the hysteresis of response fluctuates across multiple consecutive time intervals, a time-series sequence can be formed by arranging them in a one-to-one correspondence. Sequentially aligned data pairs ensure that the two types of features within the same time interval always maintain a corresponding relationship. This allows for joint analysis of change intensity and response lag based on the same time reference during subsequent processing, avoiding feature association bias caused by time misalignment. Here, the change intensity feature and response lag feature after unified scale representation refer to two types of feature sequences that have undergone unified numerical range processing. One-to-one correspondence means matching the two types of features interval by interval within the same time interval. The time-order aligned combination sequence refers to a continuous sequence structure composed of paired data in each time interval in chronological order. Pairing identifier means assigning a unique time interval index to each pair of paired data to clarify its positional relationship in the sequence. Through the above processing, strict alignment and structured organization of the two types of features in the time dimension are achieved.

[0039] The intensity of change and the response lag characteristics within each time interval are weighted, with the intensity of change corresponding to the first weight value and the response lag characteristic corresponding to the second weight value. The intensity of change is multiplied by the first weight value and the response lag characteristic is multiplied by the second weight value. The product results within the same time interval are summed to form a weighted result sequence within each time interval. After completing the unified scale representation and forming the combined sequence corresponding to each time interval, based on the change intensity feature value and response lag feature value within each time interval, corresponding weight parameters are assigned to the two types of features respectively. The change intensity feature corresponds to the first weight value, and the response lag feature corresponds to the second weight value. Then, within the same time interval, the change intensity feature value is numerically multiplied with the first weight value, and the response lag feature value is numerically multiplied with the second weight value, so that the two types of features reflect their respective contributions under the same weight system. Subsequently, the two product results within the same time interval are summed to form the comprehensive quantification result for that time interval. The comprehensive quantification results of all time intervals are arranged in time interval order to form a continuous weighted result sequence. For example, if the change intensity feature is large and the response lag feature is small within a certain time interval, by assigning a higher weight to the change intensity feature and performing a product, its... The results show that the summation process is dominant, while in another time interval, the response lag feature is larger. After corresponding weighting, it will make a higher contribution to the summation result. This processing method enables features from different sources to be fused and expressed within a unified framework. The weighting process refers to the process of assigning different degrees of influence to different features by introducing weight parameters. The first weight value and the second weight value are used to adjust the influence ratio of the change intensity feature and the response lag feature in the fusion process, respectively. The product processing refers to the numerical combination of feature value and corresponding weight to reflect the weighting effect. The summation processing refers to the merging of weighted results in the same time interval to form a single quantitative result. The weighted result sequence in each time interval is a comprehensive result set arranged in chronological order, used to reflect the changes of the two types of features after fusion in different time intervals. This processing realizes the transformation of multi-dimensional features into a single sequence, providing a continuous data foundation for subsequent cumulative processing.

[0040] The weighted result sequence within each time interval is accumulated sequentially according to the time interval, the accumulated result is continuously updated, and the final accumulated value is used as the parameter for adjusting the deviation of the response.

[0041] After obtaining the weighted result sequence for each time interval, the sequence is accumulated using the time interval order as the basis. Specifically, starting from the first time interval, the weighted result of the current time interval is used as the initial accumulated value. Then, the weighted results of subsequent time intervals are sequentially superimposed in chronological order. After each superposition, the current accumulated value is updated, allowing the accumulated value to continuously change over time until the accumulation process for all time intervals is completed. This ultimately forms a cumulative result reflecting the overall deviation across the entire time range. For example, in multiple consecutive time intervals, if the weighted result is small in the early stages and gradually increases in the later stages, the cumulative value will gradually increase after interval-by-interval accumulation, thus reflecting the cumulative effect of deviation over time. Conversely, if the weighted results for each interval are large, the final accumulated value will be... The value will increase significantly, reflecting a high overall deviation. Through this progressively cumulative processing method, local deviation information scattered across various time intervals can be integrated into an overall deviation, allowing for a centralized expression of the deviation relationship between the regulatory action and the load response. The progressively cumulative processing refers to the continuous superposition of the weighted results of each interval according to the time interval order, emphasizing the progressiveness and continuity over time. The final cumulative value is the overall numerical result formed after the accumulation of all time intervals, used to characterize the global deviation. The regulation response deviation parameter is an index constructed using this final cumulative value as the quantification result, used to describe the magnitude of the deviation between the regulatory behavior and the load response. This processing can uniformly map deviation information from multiple time intervals into a single quantitative index, thereby providing a basis for subsequent regulation decisions.

[0042] S4. Map the adjustment response deviation parameter to a continuous interval, determine the adjustment effectiveness based on the position of the adjustment response deviation parameter in the continuous interval, and determine whether to perform secondary adjustment or adjust the adjustment strategy based on the adjustment response deviation parameter. In this embodiment, S4 specifically includes the following steps: S401. Divide the adjustment response deviation parameter into intervals according to a preset continuous value range, divide the preset continuous value range into multiple adjacent and non-overlapping continuous intervals, and map the adjustment response deviation parameter to the corresponding continuous interval, using the position of the continuous interval where the adjustment response deviation parameter is located as the interval positioning result. After obtaining the parameter of the deviation degree of the adjustment response, a preset continuous numerical range is first determined based on its value range. This range can be defined by the minimum and maximum values ​​obtained from historical operating data statistics, or by empirically setting a range that covers all possible values. Then, within this continuous numerical range, intervals are divided according to the numerical magnitude, dividing the overall range into multiple adjacent and non-overlapping continuous intervals. Each interval corresponds to a specific numerical interval boundary; for example, smaller values ​​are divided into low intervals, medium values ​​into middle intervals, and larger values ​​into high intervals. After completing the interval division, the parameter of the deviation degree of the adjustment response is compared one by one with the boundary range of each continuous interval to determine its interval position, thus completing the interval mapping. For example, when the parameter of the deviation degree of the adjustment response falls within the middle interval range, it is mapped to that interval, and the position of that interval in the overall interval sequence is used as the positioning basis. The reason for adopting this processing method is that by converting continuous values ​​into discrete intervals, complex continuous changes can be transformed into a hierarchical structure, providing a clear segmentation basis for subsequent determination of adjustment effectiveness, while avoiding the instability that occurs when directly using continuous values ​​for judgment.

[0043] The preset continuous numerical range refers to the boundary of a continuous numerical interval determined based on the possible value range of the adjustment response deviation parameter, used to limit the overall range of interval division. Interval division refers to the process of dividing this continuous numerical range into multiple sub-intervals according to the numerical magnitude rule. Multiple adjacent and non-overlapping continuous intervals mean that each sub-interval is numerically connected end-to-end and has no intersection, so that any value belongs to only one interval. The continuous interval position of the adjustment response deviation parameter refers to the specific interval to which the parameter belongs among all intervals and its ranking position in the interval sequence. The interval positioning result refers to the interval assignment information and its position identifier obtained after interval mapping, used as the basis for subsequent interval classification and adjustment effectiveness determination. Through the coordinated construction of these elements, the transformation from continuous numerical values ​​to interval structure can be realized, providing a stable data foundation for subsequent decision-making.

[0044] S402. Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and each continuous interval is classified according to the numerical size rule. A one-to-one correspondence is established between each continuous interval and the corresponding interval classification. The corresponding interval classification is matched according to the interval positioning result, and the matched interval classification is used as the adjustment effectiveness judgment result. S403. Based on the position of the adjustment response deviation parameter in the continuous interval and the adjustment effectiveness determination result, the continuous interval is divided into different adjustment processing intervals, and corresponding processing is performed according to the interval division result. Specifically, if the parameter is located in the first adjustment processing interval, the current adjustment strategy is maintained; if the parameter is located in the second adjustment processing interval, a secondary adjustment is performed; and if the parameter is located in the third adjustment processing interval, the adjustment strategy is adjusted.

[0045] After determining the interval location of the regulation response deviation parameter and the regulation effectiveness judgment result, the continuous interval can be further functionally divided based on its position in the continuous interval and the corresponding regulation effectiveness judgment result. This maps the continuous interval to different regulation processing intervals, and binds corresponding processing logic to each type of interval. Specifically, based on the correspondence between the interval classification and the regulation effectiveness judgment result, the continuous interval is first divided into three types of regulation processing intervals. Then, during operation, the regulation response deviation parameter is located in real time. When it falls into a certain continuous interval, its corresponding range can be determined. The adjustment range triggers corresponding processing operations. For example, when the deviation parameter of the adjustment response is in a low range, corresponding to the first adjustment range, the current adjustment strategy is maintained unchanged; when it is in the middle range, corresponding to the second adjustment range, further adjustment actions are performed to correct the deviation; when it is in a high range, corresponding to the third adjustment range, the original adjustment strategy is adjusted to change the adjustment direction. Through this range-based processing mechanism, continuously changing deviations can be transformed into discrete and executable adjustment actions, thus giving the adjustment process a clear decision path and avoiding decision instability caused by continuous numerical fluctuations.

[0046] The different adjustment processing intervals refer to several categories of intervals divided according to the distribution of the adjustment response deviation parameter in a continuous interval. Each category of interval corresponds to a type of adjustment processing logic. The first adjustment processing interval corresponds to the interval with a relatively low deviation, and its corresponding processing is to maintain the current adjustment strategy, that is, to keep the original adjustment behavior unchanged. The current adjustment strategy refers to the load distribution adjustment method being executed in the current operating state. The second adjustment processing interval corresponds to the interval with a deviation in the middle range, and its corresponding processing is secondary adjustment, that is, to continue to make supplementary adjustments on the basis of the original adjustment to correct the deviation. Secondary adjustment refers to making further adjustments on the basis of the existing adjustment actions to change the load distribution state. The third adjustment processing interval corresponds to the interval with a relatively high deviation, and its corresponding processing is adjustment strategy adjustment, that is, to change the original adjustment method or adjustment direction to re-establish the adjustment path. Adjustment strategy adjustment refers to modifying or replacing the original adjustment logic to adapt the adjustment behavior to the new operating state. Through the correspondence between these interval divisions and processing logics, the conversion of adjustment decisions from continuous values ​​to discrete control actions can be realized.

[0047] In this embodiment, S402 specifically refers to: Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and the continuous intervals are sorted according to the numerical boundary order of the continuous intervals to form a continuous interval sequence arranged according to the numerical size. After mapping the regulation response deviation parameter to continuous intervals, the interval positioning results are mapped to specific continuous intervals based on the parameter's position within each interval. All continuous intervals are then arranged in ascending order of their numerical boundaries, forming a sequence of continuous intervals with a clear size pattern. Specifically, the upper and lower boundary values ​​of each continuous interval can be extracted first, and then sorted according to the lower boundary or center value of the interval, arranging all intervals sequentially according to their numerical size. For example, intervals with smaller numerical ranges are placed at the beginning of the sequence, and intervals with larger numerical ranges are placed at the end. Subsequently, the interval containing the regulation response deviation parameter is mapped to this ordered interval sequence, thus clarifying the parameter's relative position within the entire interval system. For example, when a parameter falls into the middle interval, its position in the continuous interval sequence is also in the middle position. This process can further transform the originally discrete interval assignment into a positional expression with an ordered relationship, so that the subsequent interval classification and judgment process has a unified sorting basis. Its significance lies in the fact that by establishing the order of interval size, the relative relationship between different intervals can be clarified, thereby avoiding the judgment confusion caused by the disorder of intervals. Among them, the continuous interval sequence refers to the ordered set of intervals formed by sorting all continuous intervals according to the numerical size. Each interval in this sequence has a clear sequential relationship, which is used to reflect the relative positional distribution of the parameter of the degree of deviation of the adjustment response in the overall numerical range. Through this sequence, the structured expression of the interval position can be realized.

[0048] The continuous interval sequence is divided into multiple interval classification categories according to the numerical size rule. Each continuous interval is then associated with a corresponding interval classification category, so that each continuous interval corresponds to only one interval classification category. After forming a continuous interval sequence, the continuous intervals are divided into hierarchical levels based on their numerical values ​​within the sequence. Specifically, the sequence can be divided into segments according to the position of the continuous intervals. For example, the first few continuous intervals can be classified as low-level intervals, the middle part as mid-level intervals, and the last part as high-level intervals. Each continuous interval is assigned a corresponding hierarchical category identifier, and a one-to-one correspondence is established between intervals and hierarchical categories. This ensures that each continuous interval belongs to only one hierarchical category within the entire interval system. For instance, when a continuous interval sequence contains multiple intervals, the first few intervals can be uniformly classified into the first category, the middle few into the second category, and the last few into the third category. This classification method can transform continuously changing numerical intervals into a hierarchical classification system, thus providing a clear basis for subsequent judgments based on interval categories. Its significance lies in expressing the differences between intervals of different numerical ranges in a structured way through hierarchical processing, avoiding the problem of unclear boundaries that arises when directly using continuous intervals for judgment. Among them, interval classification according to the numerical size rule means dividing the continuous intervals into levels according to the numerical size rule. Multiple interval classification categories refer to several sets of hierarchical categories formed during the classification process. Each category corresponds to a set of continuous intervals, used to represent the hierarchical attributes of different numerical ranges. By establishing a unique correspondence between each continuous interval and its corresponding interval classification category, it can be ensured that the affiliation of any continuous interval in the classification system is unique and definite, thus providing a stable classification basis for the subsequent judgment process.

[0049] Based on the position of the interval location result in the continuous interval sequence, the corresponding continuous interval is matched, and the interval classification category is determined by the correspondence between the continuous interval and the interval classification category. The interval classification category is divided into effective regulation category, deviation regulation category and ineffective regulation category, and the category type corresponding to the matched interval classification category is used as the regulation effectiveness judgment result.

[0050] After constructing and classifying the continuous interval sequence, based on the specific location of the interval positioning results within the continuous interval sequence, the continuous interval corresponding to the deviation parameter of the regulation response is first determined. Then, using the established one-to-one correspondence between the continuous interval and the interval classification category, the interval classification category to which the continuous interval belongs is found. Based on this, the interval classification categories are further divided into effective regulation, deviation regulation, and ineffective regulation categories according to their numerical values. Interval classification categories with smaller values ​​are assigned to the effective regulation category, indicating a low degree of deviation between the regulation behavior and the load response. Interval classification categories in the middle range are assigned to the deviation regulation category, indicating a certain degree of deviation between the regulation behavior and the load response. Interval classification categories with larger values ​​are assigned to the ineffective regulation category, indicating that the regulation behavior has not been effectively transmitted to the load change. For example, when... When the deviation parameter of the adjustment response falls into a lower value range, it can be determined to belong to the effective adjustment category through interval matching. When it falls into the middle or high value range, it corresponds to the deviation adjustment category or the failure adjustment category, respectively. Through this hierarchical and mapping relationship, the continuously changing deviation can be transformed into a clear category judgment result, thus forming the adjustment effectiveness judgment result. Among them, the effective adjustment category indicates that the adjustment behavior has good consistency with the load change, the deviation adjustment category indicates that there is a certain degree of deviation but there is still room for adjustment, and the failure adjustment category indicates that the adjustment behavior has not produced an effective response. The adjustment effectiveness judgment result refers to the judgment output determined according to the category type of the interval hierarchical category, which is used to determine whether to perform secondary adjustment and whether to adjust the adjustment strategy. Through this processing, a structured transformation from numerical deviation to category judgment can be achieved, so that the adjustment decision has a clear basis.

[0051] S5. Based on the determination of the effectiveness of regulation, update the regulation trajectory and combine it with the load difference evolution trajectory to form a new coupling sequence relationship, and perform progressive adjustment to complete the load distribution regulation.

[0052] In this embodiment, S5 specifically refers to: Based on the results of the regulation effectiveness assessment, the regulation trajectory is updated. According to the category type corresponding to the regulation effectiveness assessment results, the change amplitude of the regulation trajectory in each time interval is corrected, and the change direction of the regulation trajectory is adjusted accordingly to form the updated regulation trajectory. After obtaining the results of the regulation effectiveness assessment, the changes in the regulation trajectory over time can be updated in a targeted manner according to different category types. Specifically, the change amplitude of the regulation trajectory in each time interval is first extracted in the order of time intervals, and the correction strategy is determined according to the category type corresponding to the regulation effectiveness assessment result. For example, when it is determined to be a deviation from the regulation category, the change amplitude in each time interval is increased or decreased according to a preset ratio to change the intensity of the regulation. When it is determined to be a failed regulation category, not only is the change amplitude corrected, but the change direction of the regulation trajectory is also adjusted, such as changing the original upward trend to a downward trend or changing the change rhythm, thereby reconstructing the regulation path. After completing the change amplitude correction and change direction adjustment, the updated change results in each time interval are reconnected to form a continuous updated trajectory. For example, if the change amplitude of the original regulation trajectory is small and does not cause load changes in a certain period of time, the change amplitude of that time interval is enhanced and the change direction is adjusted to make subsequent regulation more targeted. This processing enables the regulation trajectory to be dynamically corrected according to the assessment result, thereby improving the adaptability of subsequent regulation.

[0053] The adjustment effectiveness judgment result refers to the category type obtained based on the deviation parameter of the adjustment response, used to distinguish different adjustment states; the adjustment action trajectory refers to the continuous change path formed by the adjustment command over time during the load distribution adjustment process, used to describe the evolution process of adjustment behavior in the time dimension; the change amplitude within each time interval refers to the magnitude of the change in the adjustment action trajectory value within each time interval, used to characterize the adjustment intensity; interval correction refers to adjusting the change amplitude within each time interval, causing it to increase or decrease; the change direction refers to the change trend of the adjustment action trajectory in the time dimension, used to distinguish between rising or falling states; corresponding adjustment refers to changing or maintaining the change direction based on the adjustment effectiveness judgment result; the updated adjustment action trajectory refers to the continuous change path re-formed after completing the change amplitude correction and change direction adjustment, used as the basis for subsequent coupling and progressive adjustments. Through the coordinated processing of these elements, the adjustment behavior can be transformed from static execution to dynamic correction.

[0054] The updated regulation trajectory and the load difference evolution trajectory are time-aligned, and the updated regulation trajectory and the load difference evolution trajectory are interval-matched according to the time interval order. A one-to-one correspondence between the regulation trajectory and the load difference evolution trajectory is established in each time interval, forming a new coupling sequence relationship. After updating the regulation trajectory, to ensure a unified time reference between the regulation behavior and the load response, time alignment processing is required between the updated regulation trajectory and the load difference evolution trajectory. Specifically, this can be achieved by first uniformly calibrating the timestamps of both types of trajectories. For example, data from different sampling time points can be resampled or interpolated according to a unified time reference, ensuring that the two types of trajectories have corresponding data at the same time node. Then, the two types of trajectories are divided into time intervals, and the corresponding data of the updated regulation trajectory and the load difference evolution trajectory are extracted within each time interval. Interval matching processing establishes a correspondence between two types of data within the same time interval. For example, within a certain time interval, the adjustment trajectory shows increased change, while the load difference evolution trajectory shows delayed change. Interval matching clarifies the correspondence between the two types of changes within that time interval. After completing the matching of all time intervals, the corresponding data within each time interval are arranged in chronological order to form a sequence structure containing the correspondence between adjustment behavior and load response, thereby constructing a new coupled sequence relationship. This processing can eliminate the impact of time misalignment, enabling subsequent progressive adjustments to be based on a consistent time reference.

[0055] The updated regulation action trajectory refers to the continuous change path formed after the change amplitude correction and change direction adjustment, used to reflect the new regulation behavior; the load difference evolution trajectory refers to the continuous change sequence formed by the change of the load difference between two transformers over time, used to reflect the load response; time alignment processing refers to the process of establishing a correspondence between the two types of trajectories at the same time node by unifying the time base; time interval order refers to the sequential structure of sorting data according to time sequence; interval matching refers to the corresponding pairing of data of the two types of trajectories within the same time interval; one-to-one correspondence means that there is only one set of regulation action trajectory data and one set of load difference evolution trajectory data in each time interval; the new coupling sequence relationship refers to the corresponding data sequence arranged in chronological order after completing time alignment and interval matching, used to describe the coupling relationship between regulation behavior and load response. Through the construction of these elements, a structured association expression between regulation and response can be realized.

[0056] Based on the new coupling sequence relationship, the updated adjustment trajectory is progressively adjusted according to the time interval order. The change in the adjustment trajectory in each time interval is progressively corrected interval by interval. After the correction is completed in each time interval, the status of the adjustment trajectory is updated until the progressive adjustment of all time intervals is completed to complete the load distribution adjustment.

[0057] After forming a new coupling sequence, the updated adjustment trajectory is progressively adjusted using the time interval sequence as the main processing line. Specifically, the change in the adjustment trajectory is first extracted in the first time interval, and the change is corrected by combining it with the corresponding change in the load difference evolution trajectory within the same time interval. For example, if the adjustment change deviates from the load change, the change is strengthened or weakened. After completing the correction for the first time interval, the corrected result is used as the initial state for subsequent time intervals, and the same process is repeated in the next time interval: extracting the change, correcting it based on the corresponding load change, and updating the trajectory state. This process is advanced interval by interval in chronological order, forming a gradually accumulating adjustment process. For example, if insufficient adjustment in the early stage leads to a lag in load change in multiple consecutive time intervals, the adjustment change is gradually strengthened in subsequent time intervals, so that the adjustment effect gradually approaches the target. This progressive processing avoids the instability caused by a large adjustment at once, making the adjustment process smoother and more continuous.

[0058] The new coupling sequence relationship refers to the corresponding sequence between the adjustment trajectory and the load difference evolution trajectory formed on the basis of time alignment and interval matching, which serves as the basis for progressive adjustment. "According to time interval order" means processing intervals sequentially based on their chronological order. "Progressive adjustment" means continuously correcting the results of the previous time interval within each time interval, allowing the adjustment process to progress gradually. "Change in adjustment trajectory" refers to the degree of change in the adjustment trajectory value within a single time interval, reflecting the local adjustment intensity. "Interval-by-interval progressive correction" means correcting the change in each time interval and passing it to subsequent intervals. "Update adjustment trajectory state" means forming a new trajectory state after completing the correction of each time interval, serving as the input basis for the next time interval. "Completing progressive adjustment across all time intervals to complete load distribution adjustment" means forming the final adjustment result after continuous correction across all time intervals, used to achieve the load distribution adjustment target. Through the coordinated processing of these elements, continuous iteration and gradual convergence of the adjustment process can be achieved.

[0059] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means (e.g., infrared, wireless, microwave, etc.). A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0060] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0061] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0062] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0063] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0064] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0065] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for adaptive load adjustment of dual transformers operating in parallel, characterized in that, Specifically, the following steps are included: S1. Obtain the regulation action trajectory formed during the load distribution adjustment process, and extract the load difference evolution trajectory of the two transformers within the corresponding time interval. Align the regulation action trajectory with the load difference evolution trajectory in time and match the intervals to form a coupled sequence relationship. S2. Calculate the consistency of change direction and the matching degree of change response between the regulation action trajectory and the load difference evolution trajectory based on the coupling sequence relationship. Determine whether there is a situation where the two transformers have performed the regulation action but the load difference has not changed accordingly based on the combination relationship between the consistency of change direction and the matching degree of change response. S3. In the case where two transformers have performed regulation actions but the load difference has not changed accordingly, extract the change intensity characteristics of the regulation action trajectory and the response lag characteristics of the load difference evolution trajectory, perform a unified scale mapping on the change intensity characteristics and the response lag characteristics, and construct a parameter for the degree of deviation of the regulation response. S4. Map the adjustment response deviation parameter to a continuous interval, determine the adjustment effectiveness based on the position of the adjustment response deviation parameter in the continuous interval, and determine whether to perform secondary adjustment or adjust the adjustment strategy based on the adjustment response deviation parameter. S5. Based on the determination of the effectiveness of regulation, update the regulation trajectory and combine it with the load difference evolution trajectory to form a new coupling sequence relationship, and perform progressive adjustment to complete the load distribution regulation.

2. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 1, characterized in that, S1 specifically refers to: During the load distribution adjustment process, a continuous data sequence of adjustment commands over time is obtained, and the changes in adjustment commands are continuously connected in chronological order to construct the adjustment action trajectory. The load data sequences of two transformers are collected within the time interval covered by the corresponding adjustment trajectory. The load data of the two transformers are then calculated to obtain the load difference evolution trajectory by continuously connecting the difference changes in chronological order. Using time sequence as a unified reference, time alignment processing is performed on the regulation trajectory and the load difference evolution trajectory. The time interval is divided according to continuous time intervals. Within each time interval, the regulation trajectory and the load difference evolution trajectory are matched. A one-to-one mapping is established according to the time correspondence to form a coupled sequence relationship.

3. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 1, characterized in that, S2 specifically refers to: Based on the coupling sequence relationship, the change direction of adjacent time points of the regulation trajectory and the change direction of adjacent time points of the load difference evolution trajectory are obtained in each time interval. The two types of change directions in each time interval are compared interval by interval. Time intervals with the same change direction are marked, and time intervals with opposite change directions or one of them not changing are marked. The marking results of all time intervals are statistically processed, and the proportion of the number of time intervals with the same change direction in the total number of time intervals is used as the consistency of the change direction between the regulation trajectory and the load difference evolution trajectory. After the consistency of the change direction is calculated, the change amplitude of the regulation trajectory and the change amplitude of the load difference evolution trajectory in each time interval are obtained. The corresponding difference is calculated for the change amplitude in each time interval to form an amplitude difference sequence. The amplitude difference sequence is accumulated in chronological order and combined with the cumulative value of the change amplitude of the regulation trajectory in all time intervals to normalize the amplitude difference sequence. The result after normalization is used as the change response matching degree between the regulation trajectory and the load difference evolution trajectory. The determination is based on the combination of the consistency of change direction and the matching degree of change response. When the consistency of change direction is greater than the preset consistency threshold and the matching degree of change response is less than the preset matching threshold, it is determined that there is a situation where the two transformers have performed the adjustment action but the load difference has not changed accordingly.

4. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 1, characterized in that, S3 specifically includes the following steps: S301. When two transformers have performed adjustment actions but the load difference has not changed accordingly, obtain the change amplitude sequence of the adjustment action trajectory in each time interval, and perform interval cumulative processing on the change amplitude sequence. Use the cumulative value of the change amplitude in each time interval as the change intensity feature of the adjustment action trajectory. At the same time, obtain the time difference between the response start time of the load difference evolution trajectory and the change start time of the adjustment action trajectory in each time interval, and perform interval statistical processing on the time difference. Use the statistical result as the response lag feature of the load difference evolution trajectory. S302. Perform unified scale mapping processing on the change intensity feature and the response lag feature, divide the change intensity feature and the response lag feature into segments according to the preset numerical interval, and convert the segment mapping result into a standardized numerical sequence within the same numerical interval, and use the standardized numerical sequence as the unified scale representation of the change intensity feature and the response lag feature. S303. Based on the change intensity characteristics and response lag characteristics after unified scale representation, they are combined in the order of time intervals. The change intensity characteristics and response lag characteristics in each time interval are weighted and the weighted results are accumulated within the interval. The accumulated results are used as parameters to adjust the degree of response deviation.

5. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 4, characterized in that, S303 specifically refers to: Based on the change intensity features and response lag features represented by a unified scale, they are arranged one-to-one according to the time interval order to form a time-order aligned combination sequence, and the change intensity features and response lag features in each time interval are paired and labeled. The intensity of change and the response lag characteristics within each time interval are weighted, with the intensity of change corresponding to the first weight value and the response lag characteristic corresponding to the second weight value. The intensity of change is multiplied by the first weight value and the response lag characteristic is multiplied by the second weight value. The product results within the same time interval are summed to form a weighted result sequence within each time interval. The weighted result sequence within each time interval is accumulated sequentially according to the time interval, the accumulated result is continuously updated, and the final accumulated value is used as the parameter for adjusting the deviation of the response.

6. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 1, characterized in that, S4 specifically includes the following steps: S401. Divide the adjustment response deviation parameter into intervals according to a preset continuous value range, divide the preset continuous value range into multiple adjacent and non-overlapping continuous intervals, and map the adjustment response deviation parameter to the corresponding continuous interval, using the position of the continuous interval where the adjustment response deviation parameter is located as the interval positioning result. S402. Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and each continuous interval is classified according to the numerical size rule. A one-to-one correspondence is established between each continuous interval and the corresponding interval classification. The corresponding interval classification is matched according to the interval positioning result, and the matched interval classification is used as the adjustment effectiveness judgment result. S403. Based on the position of the adjustment response deviation parameter in the continuous interval and the adjustment effectiveness determination result, the continuous interval is divided into different adjustment processing intervals, and corresponding processing is performed according to the interval division result. Specifically, if the parameter is located in the first adjustment processing interval, the current adjustment strategy is maintained; if the parameter is located in the second adjustment processing interval, a secondary adjustment is performed; and if the parameter is located in the third adjustment processing interval, the adjustment strategy is adjusted.

7. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 6, characterized in that, S402 specifically refers to: Based on the position of the adjustment response deviation parameter in the continuous interval, the interval positioning result is mapped to the corresponding continuous interval, and the continuous intervals are sorted according to the numerical boundary order of the continuous intervals to form a continuous interval sequence arranged according to the numerical size. The continuous interval sequence is divided into multiple interval classification categories according to the numerical size rule. Each continuous interval is then associated with a corresponding interval classification category, so that each continuous interval corresponds to only one interval classification category. Based on the position of the interval location result in the continuous interval sequence, the corresponding continuous interval is matched, and the interval classification category is determined by the correspondence between the continuous interval and the interval classification category. The interval classification category is divided into effective regulation category, deviation regulation category and ineffective regulation category, and the category type corresponding to the matched interval classification category is used as the regulation effectiveness judgment result.

8. The method for adaptive load adjustment of dual transformers operating in parallel according to claim 1, characterized in that, S5 specifically refers to: Based on the results of the regulation effectiveness assessment, the regulation trajectory is updated. According to the category type corresponding to the regulation effectiveness assessment results, the change amplitude of the regulation trajectory in each time interval is corrected, and the change direction of the regulation trajectory is adjusted accordingly to form the updated regulation trajectory. The updated regulation trajectory and the load difference evolution trajectory are time-aligned, and the updated regulation trajectory and the load difference evolution trajectory are interval-matched according to the time interval order. A one-to-one correspondence between the regulation trajectory and the load difference evolution trajectory is established in each time interval, forming a new coupling sequence relationship. Based on the new coupling sequence relationship, the updated adjustment trajectory is progressively adjusted according to the time interval order. The change in the adjustment trajectory in each time interval is progressively corrected interval by interval. After the correction is completed in each time interval, the status of the adjustment trajectory is updated until the progressive adjustment of all time intervals is completed to complete the load distribution adjustment.

Images