A power load demand tracking analysis method and system

Abstract

This invention provides a method and system for tracking and analyzing power load demand, belonging to the field of power system technology. The method acquires actual user power load data and target load instructions within the current dispatch cycle, calculates the load error sequence and identifies deviation intervals to obtain current user electricity consumption behavior segments. These segments are then compared one by one with a historical database of electricity consumption behavior segments from the same period, identifying local deviation intervals and determining morphological shift characteristic patterns. A load feature reference template is obtained through multiple feature filtering. Based on the load feature reference template and the target load instructions, the target load trajectory is reconstructed. This allows for tracking and analysis of user power load in the next dispatch cycle, obtaining load demand tracking results. The power load demand tracking and analysis method provided by this invention can adapt to dynamic changes in user power load characteristics, effectively improving the accuracy and real-time performance of power load demand tracking, thereby enhancing the reliability of power grid dispatch operation.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, specifically relating to a method and system for tracking and analyzing power load demand. Background Technology

[0002] With the increasing penetration of renewable energy and the growing complexity of load-side response demands, power system dispatching and operation place higher demands on the real-time performance and accuracy of load demand tracking. Existing power load demand tracking methods typically rely on static matching of historical load data with current dispatch instructions. They use users' historical electricity consumption behavior as a fixed reference template, directly comparing the deviation between the current actual load and the preset target load within the dispatching cycle, and generating control instructions accordingly. However, such methods have significant shortcomings in practical applications: because they do not consider the dynamic changes in user load characteristics due to seasonality, production plans, or external environmental factors, they continue to use outdated load patterns for tracking and analysis even after user electricity consumption behavior has deviated. This leads to distorted tracking results, delayed control instructions, and even misjudgments, making it difficult to accurately respond to real load demand and limiting the real-time performance and accuracy of power grid dispatching. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method and system for tracking and analyzing power load demand, thereby solving the aforementioned problems. This method can adapt to the dynamic changes in user power load characteristics, effectively improving the accuracy and real-time performance of power load demand tracking, and thus enhancing the reliability of power grid dispatching and operation.

[0004] To address the aforementioned technical problems, this invention provides a method for power load demand tracking and analysis, comprising the following steps: Within the current scheduling cycle, obtain the user's current actual power load data and target load command, and obtain the load error sequence based on the user's current actual power load data and target load command; Based on the load error sequence and the preset load deviation threshold, deviation intervals are identified to obtain current user electricity consumption behavior segments; Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library is identified for morphological deviation to obtain a set of local morphological deviation intervals. Among them, one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. For each historical electricity consumption behavior segment in the preset historical electricity consumption behavior segment library, the corresponding morphological deviation feature pattern is determined based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and the preset morphological deviation rules. Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, multiple feature filtering is performed to obtain the load feature reference template. Based on the load characteristic reference template and the target load command, the trajectory is reconstructed to obtain the target load trajectory; Based on the target load trajectory, the actual power load of users in the next scheduling cycle is tracked and analyzed to obtain the load demand tracking results.

[0005] In the above scheme, a load error sequence is constructed by acquiring the current actual power load data and the target load command. The current user's electricity consumption behavior segment is obtained through deviation interval identification. The historical electricity consumption behavior segments and the current user's electricity consumption behavior segments are then compared one by one to identify morphological deviations, forming a set of local morphological deviation intervals. The local morphological deviation intervals corresponding to each historical electricity consumption behavior segment are defined, laying the data foundation for determining the morphological deviation feature pattern. Based on each local morphological deviation interval and preset morphological deviation rules, the corresponding morphological deviation feature pattern is determined. After multiple feature filtering, a load feature reference template is obtained, the target load trajectory is reconstructed, and the power load demand tracking analysis for the next scheduling cycle is completed. This scheme abandons the traditional static matching mode, adapts to the dynamic changes in user load characteristics, effectively solves the problems of tracking distortion and control lag in existing technologies, improves the accuracy and real-time performance of power load demand tracking, and provides a basis for dynamic control strategies and load allocation optimization, thereby improving the reliability of power grid dispatching operations.

[0006] Further, the step of identifying deviation intervals based on the load error sequence and a preset load deviation threshold to obtain a segment of the current user's electricity consumption behavior includes: Based on the preset load deviation threshold, each load error time point in the load error sequence is determined to deviate one by one to obtain an initial deviation mark sequence; Based on the initial deviation marker sequence and the preset first duration period, continuous time points are merged to obtain candidate deviation time intervals; Based on the candidate deviation time interval, the user's current actual power load data is segmented to obtain the current user's electricity consumption behavior segment.

[0007] In the above scheme, each load error time point in the load error sequence is individually judged for deviation by a preset load deviation threshold to form an initial deviation mark sequence. Then, based on the initial deviation mark sequence and a preset first duration period, continuous time points are merged to obtain candidate deviation time intervals. Finally, based on the candidate deviation time intervals, the user's current actual power load data is segmented to obtain the current user's electricity consumption behavior segment. The above scheme can accurately identify the effective deviation interval, ensuring that the obtained current user electricity consumption behavior segment is authentic and representative, providing a reliable data foundation for subsequent local morphological deviation identification and morphological shift feature pattern determination, and improving the accuracy of subsequent multi-feature screening and trajectory reconstruction.

[0008] Furthermore, the step of identifying morphological deviations in each historical electricity consumption behavior segment in a preset historical data library based on the current user's electricity consumption behavior segment, to obtain a set of local morphological deviation intervals, includes: Based on the current user's electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library is aligned with the boundary to obtain a set of electricity consumption behavior contour pairs. In the set of electricity consumption behavior contour pairs, any electricity consumption behavior contour pair includes the current user's electricity consumption behavior segment and an aligned historical electricity consumption behavior segment. Based on each set of electricity consumption behavior profile pairs in the set of electricity consumption behavior profile pairs, the load value difference is calculated point by point in time based on the set of electricity consumption behavior profile pairs to obtain the corresponding load value difference sequence. Based on a preset deviation threshold and a preset same-direction deviation rule, the absolute values ​​of load value differences in the load value difference sequence are compared point-by-point to obtain the corresponding difference marker sequence. Based on the difference marker sequence and the preset deviation interval rule, continuous time points are merged to obtain the corresponding preliminary morphological deviation interval set; Based on the preliminary set of morphological deviation intervals and the preset second duration period, effective intervals are filtered to obtain several corresponding local morphological deviation intervals. The set of local shape deviation intervals is obtained based on several local shape deviation intervals corresponding to all electricity consumption behavior profiles.

[0009] In the above scheme, the current user's electricity consumption behavior segment is aligned with each historical electricity consumption behavior segment in a preset historical database of the same period to obtain a set of electricity consumption behavior profile pairs. For each set of electricity consumption behavior profile pairs, the load value difference is calculated point-by-point to obtain a load value difference sequence. Based on a preset deviation threshold and a preset same-direction deviation rule, the difference marker sequence is obtained by comparing point-by-point according to preset deviation interval rules. Combined with preset deviation interval rules, continuous time points are merged to obtain a preliminary set of morphological deviation intervals. Then, a preset second duration period is used to filter the effective intervals to obtain several local morphological deviation intervals and form a set of local morphological deviation intervals. The above scheme can accurately obtain the local morphological deviation intervals corresponding to each historical electricity consumption behavior segment, laying an accurate data foundation for subsequently determining the corresponding morphological deviation feature patterns and ensuring the effectiveness and reliability of morphological deviation identification.

[0010] Further, for each historical electricity consumption behavior segment in the preset historical concurrent electricity consumption behavior segment library, determining the corresponding morphological deviation feature pattern based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological deviation rules includes: For each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, time coverage information is extracted based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment to obtain the corresponding local deviation coverage parameter set. Based on the alignment time range and the corresponding local deviation coverage parameter set in the historical electricity consumption behavior segment, obtain the corresponding interval coverage ratio value set; Based on the interval coverage ratio value set and the preset ratio threshold, the offset coverage type is determined to obtain the corresponding coverage type determination result; Based on the preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result, the corresponding morphological offset feature pattern is determined.

[0011] In the above scheme, for each historical electricity consumption behavior segment in a pre-set historical contemporaneous electricity consumption behavior segment library, time coverage information is extracted based on several corresponding local morphological deviation intervals to obtain a local deviation coverage parameter set. This is combined with the aligned time range to obtain an interval coverage ratio value set. Coverage type is determined according to a pre-set ratio threshold. Then, based on pre-set morphological deviation rules, the local deviation coverage parameter set, and the coverage type determination results, the corresponding morphological deviation feature pattern is determined. This scheme can quantify and standardize local morphological deviation features, forming standardized morphological deviation feature patterns. This provides a standardized and reliable feature basis for subsequent multi-feature screening, improves the accuracy of feature determination, and lays a solid data foundation for obtaining load feature reference templates.

[0012] Furthermore, the morphological offset feature mode includes a no-offset mode, a full-offset mode, and a partial-offset mode. The determination of the corresponding morphological offset feature mode based on preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result includes: The existence of the deviation interval is determined based on the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the result of the deviation interval existence determination is obtained. When the result of the deviation interval existence determination is that there is no deviation interval, the morphological offset feature mode is determined as the no-offset mode; When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment covers the alignment time range, and all deviation directions in the local deviation coverage parameter set are consistent, the morphological offset feature mode is determined as the comprehensive offset mode. When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment does not cover the alignment time range, and the deviation directions of all deviations in the local deviation coverage parameter set are inconsistent, the morphological offset feature mode is determined as the local offset mode.

[0013] In the above scheme, the existence of deviation intervals is determined by analyzing the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment. Based on the determination result, the coverage relationship between the time range of the local morphological deviation interval and the alignment time range, and the consistency of the deviation direction, the morphological offset feature mode is determined as a no-offset mode, a full-offset mode, or a local offset mode, respectively. This scheme can accurately classify different morphological offset feature modes, clearly distinguish the offset type between historical electricity consumption behavior segments and current user electricity consumption behavior segments, provide a clear and standardized classification basis for subsequent multi-feature screening, improve the targeting and accuracy of feature screening, and provide reliable support for obtaining load feature reference templates.

[0014] Furthermore, based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, multiple feature filtering is performed to obtain a load feature reference template, including: Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the morphological offset feature pattern corresponding to each historical electricity consumption behavior segment, historical electricity consumption behavior segments with morphological offset feature pattern of full offset mode or local offset mode are selected to obtain a candidate morphological matching segment set. For each candidate morphology matching segment in the candidate morphology matching segment set, calculate the corresponding time overlap value based on the candidate morphology matching segment and the current user electricity consumption behavior segment; Based on each candidate morphological matching segment in the candidate morphological matching segment set and its corresponding time overlap value, overlap degree screening is performed, and the candidate morphological matching segment with the highest time overlap value is determined as the first target electricity consumption behavior segment. Based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment, the contour fitting degree is compared to obtain the contour fitting degree judgment result. When the contour fit determination result is contour fit, the first target power consumption behavior segment is determined as the load characteristic reference template; When the contour fit determination result is that the contour does not fit, the first target power consumption behavior segment is removed from the candidate shape matching segment set to obtain an updated candidate shape matching segment set. Then, the process of calculating the time overlap value based on the updated candidate shape matching segment set and selecting the first target power consumption behavior segment is returned until a historical power consumption behavior segment with contour fit is selected as the load feature reference template.

[0015] In the above scheme, a candidate morphological matching segment set is obtained by filtering historical electricity consumption behavior segments with morphological offset feature patterns of either full offset or partial offset. The time overlap value between each candidate segment and the current user's electricity consumption behavior segment is calculated to determine the first target electricity consumption behavior segment. After contour fitting comparison, if the segments fit, they are determined as load feature reference templates; otherwise, they are discarded and re-filtered. This scheme, through multi-feature progressive filtering, can accurately obtain historical electricity consumption behavior segments that match the current user's electricity consumption behavior segments as load feature reference templates, ensuring the accuracy and reliability of reference template selection, avoiding matching errors caused by single-dimensional filtering, and providing stable and effective data support for subsequent target load trajectory reconstruction.

[0016] Further, the step of comparing the contour fit based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment to obtain the contour fit determination result includes: Based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment, the corresponding first historical load fluctuation profile and current load fluctuation profile are extracted respectively. Based on several local morphological deviation intervals corresponding to the current user's electricity consumption behavior segment, the deviation intervals of the current user's electricity consumption behavior segment are eliminated to obtain the non-deviation intervals; Based on the non-deviation interval, the first historical load fluctuation profile and the current load fluctuation profile are truncated to obtain the corresponding first sub-profile and the current sub-profile. The load value difference is calculated point-by-point based on the first sub-contour and the current sub-contour to obtain the sub-contour difference sequence; Based on the sub-contour difference sequence and the preset fitting threshold, deviation is determined point by point to obtain a set of fitting marks at each time point. Based on the set of bonding marks at the time points, bonding consistency is determined to obtain the contour bonding degree determination result.

[0017] In the above scheme, the first historical load fluctuation profile and the current load fluctuation profile corresponding to the first target electricity consumption behavior segment and the current user electricity consumption behavior segment are extracted respectively. Based on the local morphological deviation intervals, deviation intervals are eliminated to obtain the non-deviation intervals. The first sub-profile and the current sub-profile are extracted, and the load value difference is calculated point-by-point to obtain the sub-profile difference sequence. A deviation judgment is performed using a preset fitting threshold to obtain the time-point fitting mark set. The fitting consistency judgment result is then obtained. This scheme can avoid the interference of local morphological deviation intervals on profile comparison, and only performs profile matching judgment within the non-deviation intervals, ensuring the objectivity and accuracy of the profile fitting judgment result. This provides a rigorous and reliable judgment basis for the selection of load feature reference templates, improving the accuracy of reference template selection.

[0018] Further, the step of reconstructing the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory includes: Based on the load feature reference template and the target load instruction, change features are extracted point-by-point to obtain a time pair set of instruction templates; Based on any instruction template time pair in the instruction template time pair set, the load change direction of the load feature reference template and the instruction change direction of the target load instruction are extracted based on the instruction template time pair, and the direction consistency judgment is performed based on the load change direction and the instruction change direction to obtain the direction consistency judgment result. Based on the direction consistency judgment result, several instruction template time pairs in the instruction template time pair set are subjected to continuous time merging processing to obtain direction consistent time periods and direction inconsistent time periods. Based on the load characteristic reference template and the target load instruction, the directional consistent time period and the directional inconsistent time period are reconstructed respectively to obtain the corresponding directional consistent reconstructed load instruction sequence and the directional inconsistent reconstructed load instruction sequence. Based on the reconstructed load command sequence with consistent direction, the reconstructed load command sequence with inconsistent direction, and the preset user equipment physical response capability threshold, an abnormal jump point set is identified. After smoothing and correction based on the abnormal jump point set, the reconstructed load command sequence with consistent direction, and the reconstructed load command sequence with inconsistent direction, the two sets are merged to obtain the target load trajectory.

[0019] In the above scheme, a set of instruction template time pairs is obtained by extracting change features based on load feature reference templates and target load instructions at each time point. The consistency between the load change direction and the instruction change direction of each instruction template time pair is determined, and these are merged to obtain time periods with consistent and inconsistent directions. The corresponding reconstructed load instruction sequences are then reconstructed separately. Abnormal jump point sets are identified based on a preset user equipment physical response capability threshold, and after smoothing correction, they are merged to obtain the target load trajectory. This scheme can reconstruct the trajectory separately for different directional time periods and perform smoothing correction based on the physical response capability threshold, ensuring the target load trajectory is reasonable and feasible.

[0020] Further, the reconstruction processing of the directionally consistent time period and the directionally inconsistent time period based on the load characteristic reference template and the target load command, respectively, to obtain the corresponding directionally consistent reconstructed load command sequence and the directionally inconsistent reconstructed load command sequence, includes: A preliminary adaptation value is determined based on the directional consistency time period and the target load command, and the directional consistency reconfiguration load command sequence is obtained based on the preliminary adaptation value and the directional consistency time period; Based on the aforementioned inconsistent time period and load characteristic reference template, the user's actual load adjustment capability boundary is extracted and the user's acceptable load limit range is determined. Based on the user's actual load adjustment capability boundary, the user's acceptable load limit range, and the target load command, a local command correction signal is generated. The alternative command value is determined based on the local command correction signal and the load characteristic reference template, and the direction inconsistency reconstructed load command sequence is obtained based on the alternative command value.

[0021] In the above scheme, a preliminary adaptation value is determined based on the directional consistency time period and the target load command, resulting in a directional consistency reconfiguration load command sequence. Then, based on the directional inconsistency time period and load characteristic reference template, the user's actual load adjustment capacity boundary is extracted, and the user's acceptable load limit range is determined. A local command correction signal is generated in conjunction with the target load command. Finally, an alternative command value is determined based on the local command correction signal and the load characteristic reference template, resulting in a directional inconsistency reconfiguration load command sequence. This scheme ensures that the reconfigured load command sequence adapts to the target load command requirements and the user's actual load adjustment capacity, guaranteeing the rationality and executability of the command sequence.

[0022] The present invention also provides an electricity load demand tracking and analysis system, comprising: The data acquisition module is used to acquire the user's current actual power load data and target load instruction within the current scheduling cycle, and to acquire the load error sequence based on the user's current actual power load data and target load instruction; The electricity consumption behavior identification module is used to identify deviation intervals based on the load error sequence and the preset load deviation threshold, and obtain the current user's electricity consumption behavior segment. The morphological deviation recognition module is used to perform morphological deviation recognition on each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library based on the current user electricity consumption behavior segment, and obtain a set of local morphological deviation intervals, wherein one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. The offset mode determination module is used to determine the corresponding morphological offset feature mode for each historical electricity consumption behavior segment in the preset historical electricity consumption behavior segment library, based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological offset rules. The multi-feature filtering module is used to perform multi-feature filtering based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library and the corresponding morphological offset feature patterns, to obtain a load feature reference template. The trajectory reconstruction module is used to reconstruct the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory. The tracking and analysis module is used to track and analyze the actual power load of users in the next scheduling cycle based on the target load trajectory, and obtain the load demand tracking results.

[0023] In the above scheme, the data acquisition module obtains the user's current actual power load data and target load command, and obtains the load error sequence. The electricity consumption behavior identification module identifies the current user's electricity consumption behavior segments. The morphological deviation identification module obtains the local morphological deviation interval set. The offset pattern determination module determines the morphological deviation feature pattern. The multi-feature filtering module filters to obtain the load feature reference template. The trajectory reconstruction module reconstructs the target load trajectory. The tracking analysis module completes the load tracking analysis for the next scheduling cycle and obtains the load demand tracking results. These modules work together to form a complete analysis process, adapting to dynamic changes in user load characteristics, improving the accuracy and real-time performance of power load demand tracking, avoiding tracking distortion and control lag, and providing reliable data support for power grid dispatching and operation. Attached Figure Description

[0024] Figure 1 This is a schematic flowchart of a power load demand tracking and analysis method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of an electrical load demand tracking and analysis system architecture provided in an embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see Figure 1 This embodiment provides a method for electricity load demand tracking and analysis, including the following steps: Step S1: Within the current scheduling cycle, obtain the user's current actual power load data and target load instruction, and obtain the load error sequence based on the user's current actual power load data and target load instruction; Step S2: Based on the load error sequence and the preset load deviation threshold, identify the deviation interval to obtain the current user's electricity consumption behavior segment; Step S3: Based on the current user electricity consumption behavior segment, perform morphological deviation identification on each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library to obtain a set of local morphological deviation intervals, wherein one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. Step S4: For each historical electricity consumption behavior segment in the preset historical synchronous electricity consumption behavior segment library, determine the corresponding morphological deviation feature pattern based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and the preset morphological deviation rules. Step S5: Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, perform multiple feature filtering to obtain a load feature reference template; Step S6: Based on the load characteristic reference template and the target load command, reconstruct the trajectory to obtain the target load trajectory; Step S7: Based on the target load trajectory, perform tracking analysis on the actual power load of users in the next scheduling cycle to obtain the load demand tracking results.

[0027] In this embodiment, a load error sequence is constructed by acquiring the current actual power load data and the target load command. The current user's electricity consumption behavior segment is obtained through deviation interval identification. The historical electricity consumption behavior segments and the current user's electricity consumption behavior segment are then compared one by one to identify morphological deviations, forming a set of local morphological deviation intervals. The local morphological deviation intervals corresponding to each historical electricity consumption behavior segment are defined, laying the data foundation for determining the morphological deviation feature pattern. Based on each local morphological deviation interval and preset morphological deviation rules, the corresponding morphological deviation feature pattern is determined. A load feature reference template is obtained through multiple feature filtering, reconstructing the target load trajectory and completing the power load demand tracking analysis for the next scheduling cycle. This embodiment abandons the traditional static matching mode, adapts to the dynamic changes in user load characteristics, effectively solves the problems of tracking distortion and control lag in existing technologies, improves the accuracy and real-time performance of power load demand tracking, and provides a basis for dynamic control strategies and load allocation optimization, thereby improving the reliability of power grid dispatching operations.

[0028] It should be noted that the current scheduling cycle is a fixed-duration scheduling interval preset by the power system, which can be set to 15 minutes, 30 minutes, 1 hour, etc. according to the grid scheduling needs, and is consistent with the duration of the next scheduling cycle; the load error sequence is the set of differences between the actual power load data and the target load instruction at each time point within the current scheduling cycle. A positive difference indicates that the actual load exceeds the target instruction, a negative difference indicates that the actual load is lower than the target instruction, and a difference of zero indicates that the two are completely matched.

[0029] Further, the step of identifying deviation intervals based on the load error sequence and a preset load deviation threshold to obtain a segment of the current user's electricity consumption behavior includes: Based on the preset load deviation threshold, each load error time point in the load error sequence is determined to deviate one by one to obtain an initial deviation mark sequence; Based on the initial deviation marker sequence and the preset first duration period, continuous time points are merged to obtain candidate deviation time intervals; Based on the candidate deviation time interval, the user's current actual power load data is segmented to obtain the current user's electricity consumption behavior segment.

[0030] In this embodiment, each load error time point in the load error sequence is individually determined by a preset load deviation threshold to form an initial deviation marker sequence. Then, based on the initial deviation marker sequence and a preset first duration period, continuous time points are merged to obtain candidate deviation time intervals. Finally, based on the candidate deviation time intervals, fragments of the user's current actual power load data are extracted to obtain current user electricity consumption behavior segments. This embodiment can accurately identify effective deviation intervals, ensuring the authenticity and representativeness of the obtained current user electricity consumption behavior segments. This provides a reliable data foundation for subsequent local morphological deviation identification and morphological shift feature pattern determination, improving the accuracy of subsequent multi-feature screening and trajectory reconstruction.

[0031] It should be noted that the preset load deviation threshold can be divided into positive threshold and negative threshold, which correspond to the maximum allowable deviation of the actual load exceeding or falling below the target load command, respectively, and are jointly determined by the grid dispatch accuracy, user electricity consumption characteristics, and equipment safe operating range; the preset first duration period is the minimum continuous time length for determining the continuous deviation of the load error, so as to avoid instantaneous fluctuations being misjudged as valid deviation intervals.

[0032] Furthermore, the step of identifying morphological deviations in each historical electricity consumption behavior segment in a preset historical data library based on the current user's electricity consumption behavior segment, to obtain a set of local morphological deviation intervals, includes: Based on the current user's electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library is aligned with the boundary to obtain a set of electricity consumption behavior contour pairs. In the set of electricity consumption behavior contour pairs, any electricity consumption behavior contour pair includes the current user's electricity consumption behavior segment and an aligned historical electricity consumption behavior segment. Based on each set of electricity consumption behavior profile pairs in the set of electricity consumption behavior profile pairs, the load value difference is calculated point by point in time based on the set of electricity consumption behavior profile pairs to obtain the corresponding load value difference sequence. Based on a preset deviation threshold and a preset same-direction deviation rule, the absolute values ​​of load value differences in the load value difference sequence are compared point-by-point to obtain the corresponding difference marker sequence. Based on the difference marker sequence and the preset deviation interval rule, continuous time points are merged to obtain the corresponding preliminary morphological deviation interval set; Based on the preliminary set of morphological deviation intervals and the preset second duration period, effective intervals are filtered to obtain several corresponding local morphological deviation intervals. The set of local shape deviation intervals is obtained based on several local shape deviation intervals corresponding to all electricity consumption behavior profiles.

[0033] In this embodiment, a set of electricity consumption behavior profiles is obtained by aligning the current user's electricity consumption behavior segment with each historical electricity consumption behavior segment in a preset historical database of the same period. For each electricity consumption behavior profile pair, the load value difference is calculated point-by-point to obtain a load value difference sequence. This sequence is then compared point-by-point according to a preset deviation threshold and a preset same-direction deviation rule to obtain a difference marker sequence. Combined with a preset deviation interval rule, consecutive time points are merged to obtain a preliminary set of morphological deviation intervals. Then, a preset second duration period is used to filter effective intervals, resulting in several local morphological deviation intervals, forming a set of local morphological deviation intervals. This embodiment can accurately obtain the local morphological deviation intervals corresponding to each historical electricity consumption behavior segment, laying a precise data foundation for subsequently determining the corresponding morphological offset feature pattern and ensuring the effectiveness and reliability of morphological deviation identification.

[0034] It should be noted that the boundary alignment is to match the start and end time boundaries of historical electricity consumption behavior segments with the current user electricity consumption behavior segments. When the durations are inconsistent, they are supplemented or truncated according to the original load fluctuation trend to ensure complete alignment in the time dimension. The preset same-direction deviation rule can be that the load value differences at multiple consecutive time points are all positive or all negative. The preset second duration is the minimum continuous duration for determining the validity of local morphological deviation, which is consistent with the value rule of the preset first duration.

[0035] Further, for each historical electricity consumption behavior segment in the preset historical concurrent electricity consumption behavior segment library, determining the corresponding morphological deviation feature pattern based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological deviation rules includes: For each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, time coverage information is extracted based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment to obtain the corresponding local deviation coverage parameter set. Based on the alignment time range and the corresponding local deviation coverage parameter set in the historical electricity consumption behavior segment, obtain the corresponding interval coverage ratio value set; Based on the interval coverage ratio value set and the preset ratio threshold, the offset coverage type is determined to obtain the corresponding coverage type determination result; Based on the preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result, the corresponding morphological offset feature pattern is determined.

[0036] In this embodiment, for each historical electricity consumption behavior segment in a preset historical contemporaneous electricity consumption behavior segment library, time coverage information is extracted based on several corresponding local morphological deviation intervals to obtain a local deviation coverage parameter set. This is combined with the aligned time range to obtain an interval coverage ratio value set. Coverage type determination is completed according to a preset ratio threshold. Then, based on preset morphological deviation rules, the local deviation coverage parameter set, and the coverage type determination result, the corresponding morphological deviation feature pattern is determined. This embodiment can quantify and standardize local morphological deviation features to form a standardized morphological deviation feature pattern, providing a standardized and reliable feature basis for subsequent multi-feature screening, improving the accuracy of feature determination, and laying a solid data foundation for obtaining load feature reference templates.

[0037] It should be noted that the local deviation coverage parameter set may include the start and end time, deviation direction and coverage duration of the local morphological deviation interval; the preset ratio threshold is a quantitative standard for determining the degree of local deviation coverage, used to distinguish between full coverage and local coverage, and to ensure the objectivity of morphological deviation pattern determination.

[0038] Furthermore, the morphological offset feature mode includes a no-offset mode, a full-offset mode, and a partial-offset mode. The determination of the corresponding morphological offset feature mode based on preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result includes: The existence of the deviation interval is determined based on the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the result of the deviation interval existence determination is obtained. When the result of the deviation interval existence determination is that there is no deviation interval, the morphological offset feature mode is determined as the no-offset mode; When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment covers the alignment time range, and all deviation directions in the local deviation coverage parameter set are consistent, the morphological offset feature mode is determined as the comprehensive offset mode. When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment does not cover the alignment time range, and the deviation directions of all deviations in the local deviation coverage parameter set are inconsistent, the morphological offset feature mode is determined as the local offset mode.

[0039] In this embodiment, by determining the existence of deviation intervals in the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and based on the determination result, the coverage relationship between the time range of the local morphological deviation interval and the alignment time range, and the consistency of the deviation direction, the morphological offset feature mode is determined as a no-offset mode, a full-offset mode, or a local offset mode, respectively. This embodiment can accurately classify different morphological offset feature modes, clearly distinguish the offset type between historical electricity consumption behavior segments and current user electricity consumption behavior segments, provide a clear and standardized classification basis for subsequent multi-feature screening, improve the targeting and accuracy of feature screening, and provide reliable support for obtaining load feature reference templates.

[0040] It should be noted that the consistent deviation direction means that the load value differences in all local morphological deviation intervals are either positive or negative; the alignment time range is the unified time interval between the historical electricity consumption behavior segment and the current user electricity consumption behavior segment after boundary alignment, which is the core time reference for morphological deviation mode determination.

[0041] Furthermore, based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, multiple feature filtering is performed to obtain a load feature reference template, including: Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the morphological offset feature pattern corresponding to each historical electricity consumption behavior segment, historical electricity consumption behavior segments with morphological offset feature pattern of full offset mode or local offset mode are selected to obtain a candidate morphological matching segment set. For each candidate morphology matching segment in the candidate morphology matching segment set, calculate the corresponding time overlap value based on the candidate morphology matching segment and the current user electricity consumption behavior segment; Based on each candidate morphological matching segment in the candidate morphological matching segment set and its corresponding time overlap value, overlap degree screening is performed, and the candidate morphological matching segment with the highest time overlap value is determined as the first target electricity consumption behavior segment. Based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment, the contour fitting degree is compared to obtain the contour fitting degree judgment result. When the contour fit determination result is contour fit, the first target power consumption behavior segment is determined as the load characteristic reference template; When the contour fit determination result is that the contour does not fit, the first target power consumption behavior segment is removed from the candidate shape matching segment set to obtain an updated candidate shape matching segment set. Then, the process of calculating the time overlap value based on the updated candidate shape matching segment set and selecting the first target power consumption behavior segment is returned until a historical power consumption behavior segment with contour fit is selected as the load feature reference template.

[0042] In this embodiment, a candidate morphological matching segment set is obtained by filtering historical electricity consumption behavior segments with morphological offset feature patterns of either full offset or partial offset. The time overlap value between each candidate segment and the current user's electricity consumption behavior segment is calculated to determine the first target electricity consumption behavior segment. After contour fitting comparison, if the segments fit, they are determined as load feature reference templates; otherwise, they are discarded and re-filtered. This embodiment, through multi-feature progressive filtering, can accurately obtain historical electricity consumption behavior segments that match the current user's electricity consumption behavior segments as load feature reference templates, ensuring the accuracy and reliability of reference template selection, avoiding matching errors caused by single-dimensional filtering, and providing stable and effective data support for subsequent target load trajectory reconstruction.

[0043] In one embodiment, the load analysis agent performs multi-feature filtering based on morphological offset feature patterns. Specifically, based on the current user's electricity consumption behavior segment, each historical electricity consumption behavior segment in a preset historical same-period electricity consumption behavior segment library, and the morphological offset feature patterns corresponding to each historical electricity consumption behavior segment, historical electricity consumption behavior segments with morphological offset feature patterns of full offset or partial offset are filtered out to obtain a candidate morphological matching segment set. For each candidate morphological matching segment in the candidate morphological matching segment set, the corresponding time overlap value is calculated based on the candidate morphological matching segment and the current user's electricity consumption behavior segment. Based on each candidate morphological matching segment in the candidate morphological matching segment set and its corresponding time overlap value, overlap degree filtering is performed, and the candidate morphological matching segment with the highest time overlap value is determined as the first candidate morphological matching segment. A target electricity consumption behavior segment is identified. The first target electricity consumption behavior segment and the current user electricity consumption behavior segment are compared in terms of contour fit to obtain a contour fit determination result. When the contour fit determination result is contour fit, the first target electricity consumption behavior segment is determined as a load feature reference template, i.e., a second target electricity consumption behavior segment with consistent shape. When the contour fit determination result is contour misfit, the first target electricity consumption behavior segment is removed from the candidate shape matching segment set to obtain an updated candidate shape matching segment set. The process then returns to the previous steps of calculating the time overlap value based on the updated candidate shape matching segment set and selecting the first target electricity consumption behavior segment, until a historical electricity consumption behavior segment with contour fit is selected as a load feature reference template, i.e., a second target electricity consumption behavior segment with consistent shape is determined.

[0044] In this embodiment, during the above iterative screening process, each time a new iteration is performed, candidate segments are selected strictly in descending order of the similarity of the morphological deviation patterns. That is, the candidate segment in the previous iteration is the m-th in terms of similarity, and the segment with the (m + 1)-th similarity is selected in the next iteration, ensuring the orderliness of the iterative screening and avoiding repeated screening. In addition, to balance the screening efficiency and real-time performance, a preset maximum number of iterations Mmax is set, and its value is dynamically set according to the scale of the historical同期用电行为片段库 (historical power consumption behavior segment library). When the total number of segments N in the historical同期用电行为片段库 ≤ 20, Mmax is set to N. Since the scale of the segment library is small, full-scale iteration does not affect the processing efficiency and can ensure the comprehensiveness of the screening. When 20 < N ≤ 50, Mmax is set to 20, taking into account both comprehensiveness and real-time performance. Screening the first 20 segments with high similarity can cover more than 90% of the matching possibilities. When N > 50, Mmax is set to 30, effectively controlling the iteration duration to meet the millisecond- or second-level real-time requirements of power grid dispatching. The first 30 segments with high similarity can cover the vast majority of matching scenarios.

[0045] It should be noted that the time coincidence degree value is the proportion of the overlapping duration of the candidate morphological matching segment and the local morphological deviation interval of the current user's power consumption behavior segment to the total duration of the current segment's deviation interval, and its value range is 0 to 100%. The iterative screening process sets a termination condition. When no matching segment is screened out after the iteration reaches the preset maximum number of iterations Mmax, the segment with the best profile fitting degree during the iteration is selected as the load feature reference template to avoid infinite iteration.

[0046] It should be noted that for large industrial users and key power consumption users, due to their complex load characteristics and high requirements for dispatching accuracy, Mmax can be increased by 50% on the above basis to improve the comprehensiveness of the screening. For residential or small commercial users, due to their simple load characteristics and small fluctuations, Mmax can be decreased by 30% to further improve the real-time performance of dispatching. Through the above multi-feature progressive screening and iterative control mechanism, the historical power consumption behavior segment matching the current user's power consumption behavior segment can be accurately obtained as the load feature reference template, ensuring the accuracy and reliability of the selection of the reference template, avoiding the matching errors caused by single-dimensional screening, and providing stable and effective data support for the subsequent reconstruction of the target load trajectory.

[0047] Furthermore, the contour fitting degree comparison based on the first target power consumption behavior segment and the current user's power consumption behavior segment to obtain the contour fitting degree determination result includes: Respectively extract the corresponding first historical load fluctuation profile and the current load fluctuation profile based on the first target power consumption behavior segment and the current user's power consumption behavior segment; Based on several local morphological deviation intervals corresponding to the current user's electricity consumption behavior segment, the deviation intervals of the current user's electricity consumption behavior segment are eliminated to obtain the non-deviation intervals; Based on the non-deviation interval, the first historical load fluctuation profile and the current load fluctuation profile are truncated to obtain the corresponding first sub-profile and the current sub-profile. The load value difference is calculated point-by-point based on the first sub-contour and the current sub-contour to obtain the sub-contour difference sequence; Based on the sub-contour difference sequence and the preset fitting threshold, deviation is determined point by point to obtain a set of fitting marks at each time point. Based on the set of bonding marks at the time points, bonding consistency is determined to obtain the contour bonding degree determination result.

[0048] In this embodiment, the first historical load fluctuation profile and the current load fluctuation profile corresponding to the first target electricity consumption behavior segment and the current user electricity consumption behavior segment are extracted respectively. Based on the local morphological deviation intervals, deviation intervals are eliminated to obtain non-deviation intervals. The first sub-profile and the current sub-profile are extracted, and the load value difference is calculated point-by-point to obtain the sub-profile difference sequence. A deviation judgment is performed using a preset fitting threshold to obtain a time-point fitting mark set. The fitting consistency judgment result is then obtained. This embodiment can avoid the interference of local morphological deviation intervals on profile comparison, and only performs profile matching judgment within the non-deviation intervals, ensuring the objectivity and accuracy of the profile fitting judgment result. This provides a rigorous and reliable judgment basis for the selection of load feature reference templates, improving the accuracy of reference template selection.

[0049] It should be noted that the preset fitting threshold and the preset deviation threshold are taken to ensure that the load deviation judgment standard is consistent; the fitting consistency judgment is that the difference of the load value at all time points within the non-deviation interval does not exceed the preset fitting threshold, that is, it is judged as contour fitting, otherwise it is contour non-fitting.

[0050] Further, the step of reconstructing the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory includes: Based on the load feature reference template and the target load instruction, change features are extracted point-by-point to obtain a time pair set of instruction templates; Based on any instruction template time pair in the instruction template time pair set, the load change direction of the load feature reference template and the instruction change direction of the target load instruction are extracted based on the instruction template time pair, and the direction consistency judgment is performed based on the load change direction and the instruction change direction to obtain the direction consistency judgment result. Based on the direction consistency judgment result, several instruction template time pairs in the instruction template time pair set are subjected to continuous time merging processing to obtain direction consistent time periods and direction inconsistent time periods. Based on the load characteristic reference template and the target load instruction, the directional consistent time period and the directional inconsistent time period are reconstructed respectively to obtain the corresponding directional consistent reconstructed load instruction sequence and the directional inconsistent reconstructed load instruction sequence. Based on the reconstructed load command sequence with consistent direction, the reconstructed load command sequence with inconsistent direction, and the preset user equipment physical response capability threshold, an abnormal jump point set is identified. After smoothing and correction based on the abnormal jump point set, the reconstructed load command sequence with consistent direction, and the reconstructed load command sequence with inconsistent direction, the two sets are merged to obtain the target load trajectory.

[0051] In this embodiment, a set of instruction template time pairs is obtained by extracting change features based on a load feature reference template and the target load instruction at each time point. The consistency between the load change direction and the instruction change direction of each instruction template time pair is determined, and these are merged to obtain time periods with consistent and inconsistent directions. The corresponding reconstructed load instruction sequences are then reconstructed. Abnormal jump point sets are identified based on a preset user equipment physical response capability threshold, and after smoothing correction, these are merged to obtain the target load trajectory. This embodiment can complete trajectory reconstruction for different directional time periods separately, and perform smoothing correction in conjunction with the physical response capability threshold, ensuring that the target load trajectory is reasonable and feasible.

[0052] It should be noted that the load change direction and the instruction change direction are divided into three categories: increasing, decreasing, and stable. Consistent direction means that the two change types are exactly the same, while inconsistent direction means that the two change types conflict. The instruction template time pair set is a set of pairs of load characteristic reference template load values ​​and target load instruction values ​​at the same time point, covering the entire scheduling cycle without omission.

[0053] In one embodiment, before reconstructing the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory, a step of determining changes in user load characteristics is included. Specifically, the load analysis agent extracts the user load operation mode from historical user electricity consumption behavior segments corresponding to the load characteristic reference template, and the actual electricity consumption response characteristics reflected in the current user electricity consumption behavior segment. The user load operation mode refers to the core patterns of user electricity consumption in the same historical period corresponding to the load characteristic reference template, including fixed characteristics such as the peak load occurrence period, peak load value, load change rate, and duration of continuous operation; the actual electricity consumption response characteristics refer to the user's actual response capability to grid dispatch commands reflected in the current user electricity consumption behavior segment, including dynamic characteristics such as load adjustment sensitivity and deviation control stability. The load analysis agent compares the core and dynamic characteristics in the user load operation mode with the actual electricity consumption response characteristics to determine the results of changes in user load characteristics and clarify whether the current user electricity consumption behavior has undergone a structural change relative to the historical operation mode.

[0054] It should be noted that the aforementioned load change rate is a quantitative, objective indicator, referring to the unit change amplitude of the load value within a segment of electricity consumption behavior over a data collection time interval. Its calculation formula is the ratio of the load difference between adjacent data collection points to the data collection time interval, directly reflecting how quickly the load changes over time and representing a core characteristic of the user's load operation mode. Load adjustment sensitivity, on the other hand, is a comprehensive representation of the actual electricity consumption response characteristics. Its level is determined by multiple quantitative indicators, used to characterize the speed and adjustment capability of the user's electricity load in response to grid dispatch instructions and changes in its own electricity demand. There is a clear correspondence between the load change rate and load adjustment sensitivity: the greater the load change rate, the greater the adjustment amplitude of the user's load per unit time, the faster the response to dispatch instructions, and the higher the load adjustment sensitivity; conversely, the lower the rate of change, the lower the sensitivity. Meanwhile, the peak load occurrence period and peak load value serve as characteristic node indicators of load changes, used to help determine the rhythm and limit capacity of load adjustment. The peak occurrence period reflects the time rhythm of user load adjustment, and the peak value reflects the upper limit of user load adjustment. When the peak occurrence period matches the load adjustment period of the dispatching instruction, the effectiveness of actual adjustment sensitivity is improved. However, when the peak value deviates significantly from the target load instruction, it will restrict the actual performance of load adjustment sensitivity.

[0055] As a specific example, suppose the user load operation mode of the second target electricity consumption behavior segment corresponding to the load characteristic reference template is that the load peak occurs at the 25th minute, the peak value is 59 kW, and the load change rate increases by 1 kW every 5 minutes. The actual electricity consumption response characteristics reflected by the current user electricity consumption behavior segment are that the load peak occurs at the 25th minute, the peak value is 60 kW, the load change rate increases by 1 kW every 5 minutes, and the deviation control stability is fixed. In this scenario, since the core quantitative indicator, i.e., the load change rate, is exactly the same, and the auxiliary characteristic, i.e., the peak occurrence time, is consistent, only the peak value fluctuates normally by 1 kW, which does not affect the speed and rhythm of load adjustment. Therefore, it is determined that the current user's load adjustment sensitivity is consistent with the historical period, the load adjustment sensitivity in the actual electricity consumption response characteristics has not undergone structural changes, and the user load characteristic change result is no structural change. Through the above feature extraction and comparison, this embodiment can accurately grasp the degree of matching between the user's load operation mode and the actual response characteristics before trajectory reconstruction. This provides a reference for subsequent directional consistency judgment, segmented reconstruction and smooth correction that conforms to the user's actual response capability, ensuring that the target load trajectory follows historical patterns and adapts to current dynamic changes, further improving the rationality and feasibility of trajectory reconstruction.

[0056] It should be noted that the abnormal jump point is the time point where the load jump amplitude between adjacent time points exceeds the preset threshold of the physical response capability of user equipment, and it is the core object of trajectory smoothing correction.

[0057] In another embodiment, for identified abnormal transition points, this embodiment uses piecewise linear interpolation to generate a continuous load adjustment path for smooth correction. For ease of description, a single abnormal transition point is denoted as P, and the corresponding collection point number is... The initial load reconfiguration command value is Its preceding adjacent collection point is denoted as Serial Number Instruction value The adjacent collection point is recorded as... Serial Number Instruction value If the abnormal transition point is the first or last sampling point in the sequence, only adjacent points on one side are taken, and the path is generated by linear extension on one side. User equipment physical response capability threshold. This is the maximum load variation that a user's electrical equipment can stably withstand within a single data collection time interval Δt. Exceeding this threshold will cause the equipment to fail to respond in a timely manner, experience start-up or shutdown failures, or lose control of load regulation. Therefore, the load variation of adjacent data collection points along the adjustment path must be less than or equal to the maximum load variation of the user's electrical equipment within a single data collection time interval Δt. This is a hard constraint for path generation. The objective of generating continuous load adjustment paths is: in to Within the time interval, a set of smoothly transitioning load command value sequences is generated. This ensures that it satisfies boundary constraints, rate constraints, and shape constraints, with the boundary constraints requiring... , The rate constraint requires that the load variation range of any adjacent acquisition points be... The morphological constraints require maintaining the original trend of the initial reconfiguration load command sequence and not changing the overall load change direction.

[0058] As a basic scenario, when the abnormal jump point is located between two valid adjacent points, the following steps are used to generate a continuous load adjustment path.

[0059] First, calculate the total load change at the abnormal jump point. This value reflects the total load adjustment required to complete a smooth transition. >0 indicates that the path shows an overall increasing trend. <0 indicates a decreasing trend. =0 indicates a stable trend.

[0060] Then calculate the maximum adjustable range in a single operation: Where sign() is the sign function, which guarantees The direction is consistent with the total change.

[0061] Next, calculate the required number of transitional collection points. ,in The function is used to round up, ensuring that the variation of each segment after splitting does not exceed [a certain value]. .

[0062] Finally, a load value sequence for the continuous load adjustment path is generated according to the principle of uniform division. Begin recursively calculating point by point: The piecewise linear curve equation corresponding to this path can be expressed as: when hour, ; when hour, .

[0063] The above equations satisfy boundary constraints and rate constraints, and are linear, which highly matches the linear response characteristics of user electrical equipment.

[0064] As a concrete example, let's say the previous neighbor node... The corresponding sampling point number is 4, time is 9:15, and load value is 57kW. The abnormal jump point P corresponds to sampling point number 5, time is 9:20, and initial reconstructed load value is 60kW. The subsequent adjacent points... Corresponding to data collection point number 6, time 9:25, load value 60kW, user equipment physical response capability threshold. =2kW / acquisition segment. The jump amplitude in the initial reconstruction sequence is 3kW, which is greater than 2kW, triggering anomaly correction. The total load change was calculated. =60-57=3kW, showing an increasing trend; maximum adjustable range per operation. =+1×min(3,2)=2kW;Number of transition acquisition points required The total change is divided into a first segment of 2kW and a second segment of 1kW; the generated path load value sequence is: 57kW when x=4, 57+2=59kW when x=5, and 59+1=60kW when x=6, with adjacent change ranges of 2kW and 1kW respectively, neither exceeding [the specified value]. To achieve a smooth transition.

[0065] For extreme and special scenarios, this embodiment extends the path generation rules. When the abnormal jump point is the first sampling point in the sequence and has no preceding adjacent points, only the following adjacent points are taken. As the boundary, starting from P, according to amplitude towards The direction gradually approaches until the load value equals When the abnormal transition point is the last sampling point in the sequence and has no subsequent adjacent points, only the preceding adjacent points are taken. As the boundary, from Start pressing The amplitude gradually approaches P until the load value equals When multiple adjacent abnormal jump points exist, the consecutive abnormal jump points are treated as a single large abnormal segment. The preceding adjacent point of the segment's beginning and the following adjacent point of the segment's end are taken as the overall boundary. The total load change, the maximum adjustable range per instance, and the required number of transition collection points are calculated according to the basic scenario algorithm. The entire large abnormal segment is then divided into segments using linear interpolation to ensure that the change range of each segment after division does not exceed the limit. Through the above smoothing correction process, this embodiment can complete trajectory reconstruction for different directions and time periods, and complete the smooth transition of abnormal jump points by combining the user equipment physical response capability threshold, generating a target load trajectory that conforms to the equipment response characteristics, thus ensuring the rationality and feasibility of trajectory reconstruction.

[0066] Further, the reconstruction processing of the directionally consistent time period and the directionally inconsistent time period based on the load characteristic reference template and the target load command, respectively, to obtain the corresponding directionally consistent reconstructed load command sequence and the directionally inconsistent reconstructed load command sequence, includes: A preliminary adaptation value is determined based on the directional consistency time period and the target load command, and the directional consistency reconfiguration load command sequence is obtained based on the preliminary adaptation value and the directional consistency time period; Based on the aforementioned inconsistent time period and load characteristic reference template, the user's actual load adjustment capability boundary is extracted and the user's acceptable load limit range is determined. Based on the user's actual load adjustment capability boundary, the user's acceptable load limit range, and the target load command, a local command correction signal is generated. The alternative command value is determined based on the local command correction signal and the load characteristic reference template, and the direction inconsistency reconstructed load command sequence is obtained based on the alternative command value.

[0067] In this embodiment, a preliminary adaptation value is determined based on the directional consistency time period and the target load command to obtain a directional consistency reconfigured load command sequence. Based on the directional inconsistency time period and the load characteristic reference template, the user's actual load adjustment capability boundary is extracted, and the user's acceptable load limit range is determined. A local command correction signal is generated in conjunction with the target load command. An alternative command value is determined based on the local command correction signal and the load characteristic reference template to obtain the directional inconsistency reconfigured load command sequence. This embodiment ensures that the reconfigured load command sequence adapts to the target load command requirements and the user's actual load adjustment capability, guaranteeing the rationality and executability of the command sequence.

[0068] It should be noted that the user's actual load adjustment capacity boundary is obtained from the historical load data statistics of the load characteristic reference template, including the upper and lower limits of load adjustment; the user's acceptable load limit range is a closed interval, and the replacement instruction value must fall within this interval to ensure that the load instruction conforms to the user's equipment carrying capacity.

[0069] Furthermore, in the power load demand tracking and analysis method provided in this embodiment, the step of performing tracking and analysis based on the target time node to obtain the load demand tracking result specifically includes: Extract the start and end times and deviation directions of all load tracking continuous deviation intervals. The deviation direction refers to the positive or negative attribute of the instantaneous load deviation within the continuous deviation interval, i.e., positive continuous deviation or negative continuous deviation. Positive continuous deviation indicates that all deviations are positive and the actual load continuously exceeds the expected load value, while negative continuous deviation indicates that all deviations are negative and the actual load continuously falls below the expected load value. Only one deviation direction exists within the same continuous deviation interval.

[0070] Furthermore, based on dispatch control requirements, the ability of the user's actual load to follow the target load trajectory in the next dispatch cycle is comprehensively assessed. The dispatch control requirements refer to indicators set by the power grid dispatch center, such as the maximum permissible deviation, average deviation, and duration of continuous deviation between the user's actual load and the target load trajectory. The judgment logic is as follows: if there is no continuous deviation interval in load tracking, and the absolute value of the instantaneous load deviation at all time anchor points is less than or equal to the preset deviation judgment threshold, then the load demand tracking result is "following ability meets dispatch control requirements"; if there is a continuous deviation interval in load tracking, or the absolute value of the instantaneous load deviation exceeds the preset deviation judgment threshold, and the duration exceeds the maximum permissible duration required by dispatch control, then the load demand tracking result is "following ability does not meet dispatch control requirements".

[0071] To quantify the above judgment logic, this embodiment introduces three core indicators: maximum permissible deviation, average deviation, and duration of continuous deviation. The setting and judgment criteria for each indicator are as follows.

[0072] The maximum permissible deviation is set as a reasonable fluctuation range for the equipment's physical response threshold. A symmetrical interval is directly selected to account for both scenarios where the actual load is too high or too low. in, The maximum limit by which the actual load is allowed to exceed the expected load. To ensure that the actual load is allowed to be lower than the expected load by a minimum limit, the instantaneous deviation of all time anchor points must meet the following requirements. This ensures that no single anchor point deviation exceeds the limit.

[0073] average deviation The quantitative calculation is performed by taking the arithmetic mean of the absolute values ​​to avoid the cancellation of positive and negative deviations. The calculation formula is as follows: The average deviation reflects the overall average level of load deviation in the next scheduling cycle. The smaller the value, the higher the fit of load tracking.

[0074] According to the general adaptation standard for power grid dispatch, the average deviation must be less than or equal to 50% of the equipment's physical response threshold, i.e., the average deviation ≤ 0.5×ΔLmax, in order to balance tracking accuracy and equipment fault tolerance.

[0075] Continuous deviation in duration Based on the statistics of continuous deviation intervals, first identify the continuous deviation intervals of load tracking, that is, consecutive time periods with the same sign deviation that exceed the maximum allowable deviation, and then accumulate the duration of each interval. The calculation formula is as follows: Where m is the number of consecutive deviation intervals. and These represent the start and end times of the k-th deviation interval, respectively. This indicator reflects the total duration for which the actual load continuously exceeds the maximum allowable deviation within the next dispatch cycle. A smaller value indicates higher load tracking stability. According to general stability requirements for power grid dispatch, the total duration of continuous deviation must be less than or equal to 10% of the total duration of the next dispatch cycle, i.e. ≤0.1×T period T period The total duration of the next dispatch cycle is set by the power grid dispatch cycle. If there is no continuous deviation interval, this standard is directly met. The above three indicators must simultaneously meet the corresponding judgment criteria to determine that "following capability meets dispatch control requirements". If any indicator is not met, it is judged as "not met", forming a clear and unambiguous one-vote pass or veto mechanism.

[0076] As a specific example, this embodiment assumes a total number of time anchor points n=6, and an instantaneous deviation Δ The physical response thresholds for the equipment are 0.2kW, -0.2kW, -0.2kW, 0.1kW, -0.1kW, and -0.2kW, respectively. =2kW, sampling interval T samp =5min, total duration T of the next scheduling cycle period =30min, and no sustained deviation from the interval ( =0). Therefore, the maximum permissible deviation range is [-2kW, +2kW], and all Δ All values ​​are within the specified range, meeting the requirements; the average deviation = (0.2 + 0.2 + 0.2 + 0.1 + 0.1 + 0.2) / 6 = 1.0 / 6 ≈ 0.167 kW, which is less than or equal to 0.5 × 2 = 1 kW, meeting the requirements; the duration of continuous deviation... =0, less than or equal to 0.1 × 30 = 3 minutes, which meets the requirements. All three indicators are met, therefore, it is determined that the user's actual load's ability to follow the target load trajectory in the next scheduling cycle meets the scheduling control requirements. Through the quantification and comprehensive judgment of the above indicators, this embodiment can accurately assess the user's load tracking ability, providing a reliable basis for dynamic control strategy formulation and load allocation optimization, further improving the stability and reliability of power grid dispatching operation.

[0077] It should be noted that the time anchor point is the load data comparison time point set within the allowable tracking time band, which corresponds completely to the time point of the target load trajectory; the preset deviation judgment threshold and the preset fit threshold are consistent to ensure that the deviation judgment standard is uniform throughout the entire process.

[0078] Please see Figure 2 This embodiment also provides an electricity load demand tracking and analysis system, including: The data acquisition module is used to acquire the user's current actual power load data and target load instruction within the current scheduling cycle, and to acquire the load error sequence based on the user's current actual power load data and target load instruction; The electricity consumption behavior identification module is used to identify deviation intervals based on the load error sequence and the preset load deviation threshold, and obtain the current user's electricity consumption behavior segment. The morphological deviation recognition module is used to perform morphological deviation recognition on each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library based on the current user electricity consumption behavior segment, and obtain a set of local morphological deviation intervals, wherein one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. The offset mode determination module is used to determine the corresponding morphological offset feature mode for each historical electricity consumption behavior segment in the preset historical electricity consumption behavior segment library, based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological offset rules. The multi-feature filtering module is used to perform multi-feature filtering based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library and the corresponding morphological offset feature patterns, to obtain a load feature reference template. The trajectory reconstruction module is used to reconstruct the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory. The tracking and analysis module is used to track and analyze the actual power load of users in the next scheduling cycle based on the target load trajectory, and obtain the load demand tracking results.

[0079] In this embodiment, the data acquisition module acquires the user's current actual power load data and target load command, and obtains a load error sequence. The electricity consumption behavior identification module identifies segments of the current user's electricity consumption behavior. The morphological deviation identification module obtains a set of local morphological deviation intervals. The offset pattern determination module determines the morphological deviation feature pattern. The multi-feature filtering module filters to obtain a load feature reference template. The trajectory reconstruction module reconstructs the target load trajectory. The tracking analysis module completes the load tracking analysis for the next scheduling cycle and obtains the load demand tracking results. These modules work together to form a complete analysis process, adapting to dynamic changes in user load characteristics, improving the accuracy and real-time performance of power load demand tracking, avoiding tracking distortion and control lag, and providing reliable data support for power grid dispatching and operation.

[0080] It should be noted that the system can be integrated into the power dispatch center server, edge computing terminal, or load analysis intelligent agent to adapt to the hardware deployment requirements of different power grid dispatch scenarios.

[0081] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for electricity load demand tracking analysis, characterized in that, Includes the following steps: Within the current scheduling cycle, obtain the user's current actual power load data and target load command, and obtain the load error sequence based on the user's current actual power load data and target load command; Based on the load error sequence and the preset load deviation threshold, deviation intervals are identified to obtain current user electricity consumption behavior segments; Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library is identified for morphological deviation to obtain a set of local morphological deviation intervals. Among them, one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. For each historical electricity consumption behavior segment in the preset historical electricity consumption behavior segment library, the corresponding morphological deviation feature pattern is determined based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and the preset morphological deviation rules. Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, multiple feature filtering is performed to obtain the load feature reference template. Based on the load characteristic reference template and the target load command, the trajectory is reconstructed to obtain the target load trajectory; Based on the target load trajectory, the actual power load of users in the next scheduling cycle is tracked and analyzed to obtain the load demand tracking results.

2. The power load demand tracking and analysis method according to claim 1, characterized in that, The step of identifying deviation intervals based on the load error sequence and a preset load deviation threshold to obtain a segment of the current user's electricity consumption behavior includes: Based on the preset load deviation threshold, each load error time point in the load error sequence is determined to deviate one by one to obtain an initial deviation mark sequence; Based on the initial deviation marker sequence and the preset first duration period, continuous time points are merged to obtain candidate deviation time intervals; Based on the candidate deviation time interval, the user's current actual power load data is segmented to obtain the current user's electricity consumption behavior segment.

3. The power load demand tracking and analysis method according to claim 1, characterized in that, The process involves identifying morphological deviations in each historical electricity consumption behavior segment from a pre-defined historical data library based on the current user's electricity consumption behavior segment, resulting in a set of local morphological deviation intervals, including: Based on the current user's electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library is aligned with the boundary to obtain a set of electricity consumption behavior contour pairs. In the set of electricity consumption behavior contour pairs, any electricity consumption behavior contour pair includes the current user's electricity consumption behavior segment and an aligned historical electricity consumption behavior segment. Based on each set of electricity consumption behavior profile pairs in the set of electricity consumption behavior profile pairs, the load value difference is calculated point by point in time based on the set of electricity consumption behavior profile pairs to obtain the corresponding load value difference sequence. Based on a preset deviation threshold and a preset same-direction deviation rule, the absolute values ​​of load value differences in the load value difference sequence are compared point-by-point to obtain the corresponding difference marker sequence. Based on the difference marker sequence and the preset deviation interval rule, continuous time points are merged to obtain the corresponding preliminary morphological deviation interval set; Based on the preliminary set of morphological deviation intervals and the preset second duration period, effective intervals are filtered to obtain several corresponding local morphological deviation intervals. The set of local shape deviation intervals is obtained based on several local shape deviation intervals corresponding to all electricity consumption behavior profiles.

4. The power load demand tracking and analysis method according to claim 1, characterized in that, For each historical electricity consumption behavior segment in the preset historical concurrent electricity consumption behavior segment library, the corresponding morphological deviation feature pattern is determined based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological deviation rules, including: For each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, time coverage information is extracted based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment to obtain the corresponding local deviation coverage parameter set. Based on the alignment time range and the corresponding local deviation coverage parameter set in the historical electricity consumption behavior segment, obtain the corresponding interval coverage ratio value set; Based on the interval coverage ratio value set and the preset ratio threshold, the offset coverage type is determined to obtain the corresponding coverage type determination result; Based on the preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result, the corresponding morphological offset feature pattern is determined.

5. The power load demand tracking and analysis method according to claim 4, characterized in that, The morphological offset feature patterns include no offset mode, full offset mode, and partial offset mode. The determination of the corresponding morphological offset feature pattern based on preset morphological offset rules, the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the coverage type determination result includes: The existence of the deviation interval is determined based on the local deviation coverage parameter set corresponding to the historical electricity consumption behavior segment, and the result of the deviation interval existence determination is obtained. When the result of the deviation interval existence determination is that there is no deviation interval, the morphological offset feature mode is determined as the no-offset mode; When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment covers the alignment time range, and all deviation directions in the local deviation coverage parameter set are consistent, the morphological offset feature mode is determined as the comprehensive offset mode. When the existence determination result of the deviation interval is that a deviation interval exists, and the time range of several local morphological deviation intervals in the historical electricity consumption behavior segment does not cover the alignment time range, and the deviation directions of all deviations in the local deviation coverage parameter set are inconsistent, the morphological offset feature mode is determined as the local offset mode.

6. The power load demand tracking analysis method according to claim 5, characterized in that, Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library, and the corresponding morphological offset feature patterns, multiple feature filtering is performed to obtain a load feature reference template, including: Based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library, and the morphological offset feature pattern corresponding to each historical electricity consumption behavior segment, historical electricity consumption behavior segments with morphological offset feature pattern of full offset mode or local offset mode are selected to obtain a candidate morphological matching segment set. For each candidate morphology matching segment in the candidate morphology matching segment set, calculate the corresponding time overlap value based on the candidate morphology matching segment and the current user electricity consumption behavior segment; Based on each candidate morphological matching segment in the candidate morphological matching segment set and its corresponding time overlap value, overlap degree screening is performed, and the candidate morphological matching segment with the highest time overlap value is determined as the first target electricity consumption behavior segment. Based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment, the contour fitting degree is compared to obtain the contour fitting degree judgment result. When the contour fit determination result is contour fit, the first target power consumption behavior segment is determined as the load characteristic reference template; When the contour fit determination result is that the contour does not fit, the first target power consumption behavior segment is removed from the candidate shape matching segment set to obtain an updated candidate shape matching segment set. Then, the process of calculating the time overlap value based on the updated candidate shape matching segment set and selecting the first target power consumption behavior segment is returned until a historical power consumption behavior segment with contour fit is selected as the load feature reference template.

7. The power load demand tracking analysis method according to claim 6, characterized in that, The step of comparing the contour fit based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment to obtain the contour fit determination result includes: Based on the first target electricity consumption behavior segment and the current user electricity consumption behavior segment, the corresponding first historical load fluctuation profile and current load fluctuation profile are extracted respectively. Based on several local morphological deviation intervals corresponding to the current user's electricity consumption behavior segment, the deviation intervals of the current user's electricity consumption behavior segment are eliminated to obtain the non-deviation intervals; Based on the non-deviation interval, the first historical load fluctuation profile and the current load fluctuation profile are truncated to obtain the corresponding first sub-profile and the current sub-profile. The load value difference is calculated point-by-point based on the first sub-contour and the current sub-contour to obtain the sub-contour difference sequence; Based on the sub-contour difference sequence and the preset fitting threshold, deviation is determined point by point to obtain a set of fitting marks at each time point. Based on the set of bonding marks at the time points, bonding consistency is determined to obtain the contour bonding degree determination result.

8. The power load demand tracking and analysis method according to claim 1, characterized in that, The process of reconstructing the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory includes: Based on the load feature reference template and the target load instruction, change features are extracted point-by-point to obtain a time pair set of instruction templates; Based on any instruction template time pair in the instruction template time pair set, the load change direction of the load feature reference template and the instruction change direction of the target load instruction are extracted based on the instruction template time pair, and the direction consistency judgment is performed based on the load change direction and the instruction change direction to obtain the direction consistency judgment result. Based on the direction consistency judgment result, several instruction template time pairs in the instruction template time pair set are subjected to continuous time merging processing to obtain direction consistent time periods and direction inconsistent time periods. Based on the load characteristic reference template and the target load instruction, the directional consistent time period and the directional inconsistent time period are reconstructed respectively to obtain the corresponding directional consistent reconstructed load instruction sequence and the directional inconsistent reconstructed load instruction sequence. Based on the reconstructed load command sequence with consistent direction, the reconstructed load command sequence with inconsistent direction, and the preset user equipment physical response capability threshold, an abnormal jump point set is identified. After smoothing and correction based on the abnormal jump point set, the reconstructed load command sequence with consistent direction, and the reconstructed load command sequence with inconsistent direction, the two sets are merged to obtain the target load trajectory.

9. The power load demand tracking analysis method according to claim 8, characterized in that, The process of reconstructing the directionally consistent time period and the directionally inconsistent time period based on the load characteristic reference template and the target load command, respectively, to obtain the corresponding directionally consistent reconstructed load command sequence and the directionally inconsistent reconstructed load command sequence, includes: A preliminary adaptation value is determined based on the directional consistency time period and the target load command, and the directional consistency reconfiguration load command sequence is obtained based on the preliminary adaptation value and the directional consistency time period; Based on the aforementioned inconsistent time period and load characteristic reference template, the user's actual load adjustment capability boundary is extracted and the user's acceptable load limit range is determined. Based on the user's actual load adjustment capability boundary, the user's acceptable load limit range, and the target load command, a local command correction signal is generated. The alternative command value is determined based on the local command correction signal and the load characteristic reference template, and the direction inconsistency reconstructed load command sequence is obtained based on the alternative command value.

10. A power load demand tracking and analysis system, characterized in that, include: The data acquisition module is used to acquire the user's current actual power load data and target load instruction within the current scheduling cycle, and to acquire the load error sequence based on the user's current actual power load data and target load instruction; The electricity consumption behavior identification module is used to identify deviation intervals based on the load error sequence and the preset load deviation threshold, and obtain the current user's electricity consumption behavior segment. The morphological deviation recognition module is used to perform morphological deviation recognition on each historical electricity consumption behavior segment in the preset historical same-period electricity consumption behavior segment library based on the current user electricity consumption behavior segment, and obtain a set of local morphological deviation intervals, wherein one historical electricity consumption behavior segment corresponds to several local morphological deviation intervals in the set of local morphological deviation intervals. The offset mode determination module is used to determine the corresponding morphological offset feature mode for each historical electricity consumption behavior segment in the preset historical electricity consumption behavior segment library, based on several local morphological deviation intervals corresponding to the historical electricity consumption behavior segment and preset morphological offset rules. The multi-feature filtering module is used to perform multi-feature filtering based on the current user electricity consumption behavior segment, each historical electricity consumption behavior segment in the preset historical same period electricity consumption behavior segment library and the corresponding morphological offset feature patterns, to obtain a load feature reference template. The trajectory reconstruction module is used to reconstruct the trajectory based on the load characteristic reference template and the target load command to obtain the target load trajectory. The tracking and analysis module is used to track and analyze the actual power load of users in the next scheduling cycle based on the target load trajectory, and obtain the load demand tracking results.

Images