High-pile photovoltaic disaster-resistant prediction method and system fused with big data

Abstract

The invention relates to the technical field of photovoltaic prediction, in particular to a high-pile photovoltaic disaster-resistant prediction method and system fused with big data, and the method comprises the following steps: obtaining a structure number and angle change, screening an offset structure, extracting a time point and a number sequence, recognizing a response chain, tracking an interference direction, judging a staggered region, and extracting a number consistent in direction. And obtaining a disturbance trend prediction sequence. According to the method, component attitude association is established through structure numbers and arrangement sequences, angle changes, stress tracks and incident angle offset paths under disturbance are collected, a structure sequence of reverse offset and torsion synchronization features is extracted, delay chains are screened in combination with a response time point staggering relation, and disturbance paths are tracked based on the number connection trend. Direction change positions and arrangement interruption areas are identified, numbered sections consistent in sequence direction are extracted to construct a trend prediction sequence, structural linkage, response order and path coherence characteristics are fused, and the identification range and extension ability of the disturbance trend are enhanced.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic prediction technology, and in particular to a method and system for predicting disaster resistance of high-pile photovoltaic systems that integrates big data. Background Technology

[0002] The field of photovoltaic (PV) forecasting technology falls under the category of new energy information fusion and analysis. Its main research focus is on PV power output prediction and risk assessment based on multi-source data, including meteorological information, power plant operation data, and topographic data. Core aspects of this technology include irradiance simulation, PV module response analysis, power curve fitting, multi-timescale power prediction, data-driven model construction, and analysis of factors influencing severe meteorological events. It broadly involves several sub-directions such as solar radiation modeling, environmental coupling analysis, and big data analysis. Systematic research in this field includes the construction of short-term and ultra-short-term power output prediction models, historical data sample clustering analysis, and reliability and risk modeling of PV power plants under extreme weather interference. Traditional PV disaster prediction refers to a technical solution that uses a single historical measurement to predict disaster impact based on existing monitoring. Traditional methods typically establish a polynomial fitting model based on static meteorological variables such as solar irradiance, wind speed, and humidity. Critical values ​​are determined based on the average power fluctuation rate or fluctuation coefficient in the short term, combined with empirical formulas for PV array stress for structural early warning analysis, thereby outputting warning information before a disaster occurs.

[0003] Existing technologies rely on the single-point state of components for computation and primarily depend on power fluctuations to construct critical early warning judgments. They lack the means to identify the response sequence and spatial connection relationship between structural numbers. In scenarios where there is attitude linkage offset between structures or change in the direction of disturbance, it is difficult to capture the complete chain of disturbance transmission path. This leads to chaotic response sequence when numbering crossover or synchronous structural changes occur within the array, limiting prediction accuracy. In typical scenarios, such as offset delay of components at the array edge, traditional models cannot separate this from the overall trend, resulting in false alarms or response lag. In the process of continuous disturbance propagation, key structural segments are easily missed, affecting the completeness and accuracy of trend inference. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a high-pile photovoltaic disaster prediction method integrating big data, comprising the following steps:

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a high-pile photovoltaic disaster prevention prediction method integrating big data, comprising the following steps:

[0006] S1: Obtain the structural number and arrangement order of the high-pile photovoltaic modules, collect the angle change, pile top force and incident angle path under disturbance, analyze the direction of adjacent attitude change, screen the continuous offset structure, and obtain the structural response offset group.

[0007] S2: Based on the number set of the structural response offset group, extract the time point from the disturbance to the attitude change, determine whether adjacent numbered responses are interleaved, identify the continuous time response region according to the numbering order, and obtain the temporal chain of response behavior.

[0008] S3: Based on the arrangement structure of consecutive numbers in the temporal chain of the response behavior, analyze the connection direction of the numbers, identify the position of directional change in the path, track the direction of interference according to the number order, and obtain the response path structure linked list;

[0009] S4: Based on the position of the end node in the response path structure chain list, extract the arrangement trend and direction change path of adjacent regions, determine whether the end structure is misaligned or deflected, divide the arrangement interruption region, and obtain the list of evolution end boundary regions.

[0010] S5: Based on the structure number of the boundary region list at the end of the evolution, compare the number arrangement with the interference path, filter regions with the same order and direction, extract the structural response trend, and obtain the high pile structure disturbance trend prediction sequence.

[0011] As a further embodiment of the present invention, the structural response offset group includes structural number, arrangement relationship, reverse offset component, and torsional synchronous structural segment; the response behavior temporal chain includes response time sequence, reaction order, structural number correspondence, and temporal continuous region; the response path structure chain includes start number, end number, connection direction, and response direction change node; the evolution end boundary region list includes end concentration node, arrangement trend, structural misalignment region, and arrangement trend interruption edge; and the high pile structure disturbance trend prediction sequence includes numbering order, disturbance direction consistency region, and continuous numbering path.

[0012] As a further aspect of the present invention, the incident angle path refers to the angle change between the incident light direction and the component surface obtained by the illumination angle measuring device and the angle change trajectory formed in time sequence.

[0013] The attitude change direction refers to the directional change of the panel tilt angle over time during wind disturbances.

[0014] As a further aspect of the present invention, the tracking of interference direction refers to analyzing the spatial connection direction of the structure numbers one by one in the component sequence with a determined response time to identify the disturbance propagation path;

[0015] The so-called arrangement interruption region refers to the region of connection break and direction change caused by the interruption of the continuity of structural numbering and spatial arrangement during the extension of the disturbance path.

[0016] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0017] S101: Obtain the structural number and arrangement order of each column in the high-pile photovoltaic module array, collect the angle change, pile top directional force and incident angle path under wind disturbance, compare the arrangement continuity according to the numbering order, filter the module column area with consistent numbering spacing, and obtain the structural arrangement mapping segment.

[0018] S102: Based on the structure arrangement mapping segment, determine the attitude change direction of the components corresponding to adjacent numbers, compare the trend of angle change before and after wind direction disturbance, filter out number segments with alternating and continuous arrangement of directions, and obtain the direction offset alternating region.

[0019] S103: Based on the alternating directional offset region, extract the correspondence between the numbering order and the direction of torsion angle change, determine whether the consistent direction continues during the interference period, filter out numbering segments with consecutive numbers and the angle change maintaining order, and obtain the structural response offset group.

[0020] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0021] S201: Based on the number set of the structural response offset group, extract the time point when the structural angle offset first occurs within the disturbance start stage, identify the number of the attitude response that occurs within the disturbance period, filter the continuous number segment according to the number order, and obtain the attitude response number range.

[0022] S202: Based on the attitude response number range, extract the response time sequence corresponding to adjacent numbers, determine whether there is cross-response in time change, identify number pairs whose time response and number arrangement are not synchronized, extract the interleaved number segment, and obtain a set of time sequence abnormal numbers;

[0023] S203: Based on the set of time-sequence abnormal numbers, filter structural segments with adjacent response times and consecutive numbers, determine whether the numbers have no jumps in the arrangement direction and the time intervals are adjacent, extract the part with consecutive numbers and adjacent time, and obtain the response behavior time sequence chain.

[0024] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0025] S301: Based on the arrangement structure of consecutive numbers in the temporal chain of the response behavior, extract the arrangement direction of adjacent numbers, compare the spatial continuity relationship, filter the structure sequence with consistent direction and consecutive numbers, and obtain the consecutive number segment of arrangement direction.

[0026] S302: Based on the continuous numbering segment of the arrangement direction, extract the trend of the arrangement direction between the first and last numbers, filter the positions where the direction deviates, extract the numbers whose connection direction changes according to the changes in the arrangement before and after, and obtain the set of direction offset trigger numbers.

[0027] S303: Based on the set of direction offset trigger numbers, compare the correspondence between the numbers and the direction of the disturbance path, extract the consecutive numbered paths with the same direction trend, and obtain the response path structure linked list.

[0028] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0029] S401: Based on the position of the end node in the response path structure linked list, collect the coordinate sequence of the node in the component array diagram, extract the arrangement direction and direction change of adjacent numbers, judge the continuity of the direction trend according to the connection state of the node arrangement in the array structure, and obtain the end arrangement direction distribution.

[0030] S402: Based on the distribution of the end arrangement direction, extract the directional connection status of the component numbers in adjacent regions, identify whether there is a directional offset and arrangement distortion position between the numbers, and determine whether the arrangement direction is continuously misaligned by the trend of the number direction change to obtain the deflection and misalignment feature segment.

[0031] S403: Based on the deflection staggered feature segment, analyze the continuity of the numbering arrangement, extract the edge position of the deflection structure in space, filter the areas where the numbering is interrupted and the structure direction changes, and obtain a list of evolutionary end boundary regions.

[0032] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0033] S501: Based on the structure number in the list of boundary regions at the end of the evolution, extract the attitude change sequence of the corresponding position during the disturbance influence stage, determine whether the number arrangement order has the same direction as the response sequence, filter out continuous segments with the same direction of number change, and obtain the number group corresponding to the attitude response.

[0034] S502: Based on the attitude response corresponding number group, extract the correspondence between the number arrangement direction and the wind direction interference path, determine whether the number continues to expand along the wind direction in space, filter out the number positions with inconsistent expansion direction and path offset, and obtain a set of numbers with consistent direction.

[0035] S503: Call the set of consistent direction numbers, analyze whether the numbering has direction interruption and sequence jump during the response phase, filter the structural segments with continuous numbering and continuous response direction, and obtain the high pile structure disturbance trend prediction sequence.

[0036] A high-pile photovoltaic disaster prediction system integrating big data includes:

[0037] The structural offset identification module obtains the structural number, arrangement order, angle change trend, force trajectory and incident angle offset path, compares the attitude change direction according to the number, filters out the reverse offset and torsional synchronous segments, and obtains the structural response offset group.

[0038] The temporal behavior extraction module retrieves disturbance state records based on the structural response offset group, extracts the time point from the disturbance to the attitude change, determines whether the numbers overlap, and delineates the continuous proximity region in sequence to obtain the temporal chain of response behavior.

[0039] The response path construction module analyzes the numbering connection relationship and direction based on the temporal chain of response behavior, determines whether there are points of directional change, and traces the direction of disturbance according to the connection order to obtain the response path structure linked list.

[0040] The terminal boundary determination module is based on the response path structure linked list. It retrieves the terminal position and adjacent trends, judges the trend deflection and misalignment, delineates the trend interruption area, and obtains a list of evolution terminal boundary areas.

[0041] The disturbance trend prediction module is based on the list of boundary regions at the end of the evolution. By comparing the attitude change sequence, it judges the consistency between the numbering sequence and the response direction, extracts continuous and consistent regions, and obtains the disturbance trend prediction sequence of the high pile structure.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0043] In this invention, component attitude association is established by structural numbering and arrangement order. Angle changes, force trajectories and incident angle offset paths under disturbance are collected. Structural sequences with synchronous features of reverse offset and torsion are extracted. Delay chains are screened by combining the interlacing relationship of response time points. The disturbance path is tracked based on the connection direction of numbering. The location of direction change and arrangement interruption area are identified. Numbered segments with consistent order and direction are extracted to construct trend prediction sequences. The characteristics of structural linkage, response order and path coherence are integrated to enhance the identification range and extension capability of disturbance trend. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a schematic diagram of the steps of the present invention;

[0046] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0047] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0048] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0049] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0050] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0051] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0052] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0053] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0054] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0055] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0056] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0057] Please see Figure 1 This invention provides a high-pile photovoltaic disaster prediction method integrating big data, comprising the following steps:

[0058] S1: Obtain the structural number and arrangement order of each column of high-pile photovoltaic modules, collect the angle change trend, pile top force direction change trajectory and incident angle offset path under wind disturbance conditions, and compare the attitude change direction of adjacent modules before and after wind disturbance based on the structural number and arrangement relationship. Select the structural segments that continuously show reverse offset and synchronous torsion to obtain the structural response offset group.

[0059] S2: Based on the number set of structural response offset groups, retrieve the structural state change records when the disturbance is triggered, extract the time points between the occurrence of the disturbance and the attitude change of the structure, analyze whether there is a repetition or overlap between adjacent numbers in the response sequence, delineate the response regions with continuity and temporal proximity according to the order of the numbering, and obtain the response behavior time sequence chain.

[0060] S3: Based on the spatial arrangement structure corresponding to the consecutive numbers in the response behavior time sequence chain, analyze the connection direction between the start and end according to the connection relationship between the numbers, determine whether there is a position in the path where the response direction changes, and track the interference direction according to the connection order of the structure numbers to obtain the response path structure linked list.

[0061] S4: Based on the positional relationship of the terminal cluster nodes in the response path structure chain list in the component array diagram, retrieve the arrangement trend and structural direction transfer path of the adjacent areas of the nodes, determine whether the terminal arrangement has a direction deviation and structural misalignment with the adjacent components, and delineate the edge area of ​​the arrangement trend interruption according to the consistency state of the position arrangement to obtain the list of evolution terminal boundary areas.

[0062] S5: Based on the structure numbers in the list of boundary regions at the end of the evolution, compare the order of attitude changes of the structures in the numbered regions under the influence of disturbance, determine whether the order in which the structure numbers appear in the path is consistent with the response direction, compare the numbering order with the numbering trend in the wind direction interference path, extract continuous regions with consistent order and direction, and obtain the high pile structure disturbance trend prediction sequence.

[0063] The structural response offset group includes structural number, arrangement relationship, reverse offset component, and torsional synchronous structural segment; the response behavior time sequence chain includes response time sequence, reaction order, structural number correspondence, and temporal continuous region; the response path structure chain list includes start number, end number, connection direction, and response direction change node; the evolution end boundary region list includes end concentrated node, arrangement trend, structural misalignment region, and arrangement trend interruption edge; the high pile structure disturbance trend prediction sequence includes number order, disturbance direction consistency region, and continuous number path.

[0064] Please see Figure 2 The specific steps of S1 are as follows:

[0065] S101: Obtain the structural number and arrangement order of each column in the high-pile photovoltaic module array, collect the angle change, pile top directional force and incident angle path under wind disturbance, compare the arrangement continuity according to the numbering order, filter the module column area with consistent numbering spacing, and obtain the structural arrangement mapping segment.

[0066] To obtain the column number and arrangement order of each column in the high-pile photovoltaic module array, the number sequence can be obtained by manually comparing the numbered nameplates with the installation drawings. The changes in angle, pile top directional force, and incident angle generated by the modules during operation under wind disturbance are collected using angle sensors, strain gauges, and illumination angle measurement devices, respectively. The angle change can be expressed as the amplitude of the angle between the leading and trailing edges of the module relative to the ground. The strain gauge records the axial and radial strain changes at the pile top, and the illumination angle device records the deflection of the incident direction on the light-receiving surface. By comparing the module number with its actual arrangement position, each number in the module column is arranged sequentially, and the differences between the numbers are extracted. If a sequence with a fixed difference between numbers exists, this part is considered... For a continuously arranged region, for example, if there is a numbering sequence A01, A02, A03, A04, then the spacing between the numbers is 1 and they are continuous. By extracting the corresponding component column positions of this type of numbering region, and then collecting the collected angle change, directional force amplitude and incident angle offset data, and comparing them with the arrangement order, it is determined whether there are continuous disturbance propagation characteristics. At the same time, non-continuous information such as number jumps, number repetitions, and number missings are removed. If a numbering sequence has, for example, A03, A06, A07, then its jump part is not included in the subsequent analysis. On this basis, a region that can be tracked and continuously expressed in terms of arrangement order is formed, which is used for subsequent interference propagation path analysis and response sequence calling to obtain the structural arrangement mapping segment.

[0067] S102: Based on the structural arrangement mapping segment, determine the attitude change direction of the components corresponding to adjacent numbers, compare the trend of angle change before and after wind direction disturbance, filter out number segments with alternating and continuous direction to obtain the direction offset alternating region.

[0068] First, extract the angle change direction data for each component corresponding to its number. The angle change direction can be determined by the change in the angle between the front edge of the component panel and the horizontal reference. For example, components numbered B05 and B06 show angle fluctuations from 30 degrees to 35 degrees and from 35 degrees to 33 degrees before and after the disturbance, respectively, indicating that their angle change directions are opposite. Then, compare the angle change directions of all adjacent numbers pairwise. If the change directions alternate (one increases and one decreases), it indicates that there is an alternating direction behavior in that number segment. At the same time, it is necessary to check whether the adjacent numbers are arranged in a continuous sequence. If there is a number jump, such as B05 and B07, the data segment should be excluded and not included in the subsequent comparison. Perform a positional consistency check on the retained continuous number segments. To ensure that the arrangement is free from skew interference, for example, a continuous numbering arrangement with an L-shaped position distribution is not adopted. Only numbering groups with linear extension characteristics in the horizontal or vertical direction are retained. The difference in attitude change direction before and after the perturbation is calculated for the selected component columns. The direction difference can be expressed as the change relationship between positive and negative values ​​before and after the angle change. If there are multiple direction differences with alternating positive and negative distributions in the numbering segment and the arrangement order is continuous without jumps, it can be judged that there is a significant trend of alternating and reversing directions in the numbering sequence. For example, the component columns numbered B10 to B15 have an attitude change sequence of +2, -3, +1, -2, +4, -1, which is regarded as the alternating direction situation under continuous numbering. Finally, the alternating direction offset area is obtained.

[0069] S103: Based on the alternating region of directional offset, extract the correspondence between the numbering order and the direction of torsion angle change, determine whether the consistent direction continues during the interference period, filter out numbering segments with consecutive numbers and the angle change maintaining the order, and obtain the structural response offset group.

[0070] First, extract the torsional angle change values ​​corresponding to each component in each numbered sequence, and use the numbering order as a reference for sequence arrangement. Arrange the components from left to right or top to bottom, and collect the attitude change angles of the components during the disturbance phase. The angle change values ​​are obtained by comparing the panel tilt angle before and after the disturbance. For example, for components numbered A11 to A16, the corresponding torsional angle changes are +1.2, +1.5, +1.7, +1.9, +2.1, and +2.3 degrees, respectively. It can be seen that the change direction of this numbering segment is consistent and continuously increases during the disturbance phase. Next, the continuity of the extracted numbering sequence needs to be checked, that is, whether it is consecutively numbered and arranged in the same direction. For example, if there is a jump in the numbering segment (A12, A14, A15), it is a discontinuous arrangement and must be removed. The direction of angle change in a continuous sequence is determined. If all numbers correspond to the same direction of angle change (i.e., all are positive or all are negative), it indicates that their directions are consistent. Then, the trend direction is determined by combining the numbering order. For example, if the numbering direction is from top to bottom and the angle change value gradually increases, it can be determined that the consistent direction continues. If there are reverse changes or numerical fluctuations, such as +1.3, +1.7, +0.5, +2.1, +2.5, the numbers at the fluctuation points need to be excluded, and only the trend continuous segments are retained. Finally, the numbering segments with continuous numbers, consistent arrangement direction, and continuous angle change are selected. For example, the sequence numbered A20 to A24 has angle change values ​​of +1.1, +1.5, +1.8, +2.0, and +2.4, respectively. The order is stable and the direction is consistent, thus obtaining the structural response offset group.

[0071] Please see Figure 3 The specific steps of S2 are as follows:

[0072] S201: Based on the number set of the structural response offset group, extract the time point of the first occurrence of structural angle offset within the disturbance start stage, identify the number of attitude response that occurs within the disturbance period, filter the continuous number segment according to the number order, and obtain the attitude response number range.

[0073] First, the attitude change information of the component corresponding to each number needs to be extracted. The number range can be, for example, A01 to A20. The corresponding angle offset value is recorded as the angle change details for different time periods. For example, the tilt angle change of A01 first appears in the 10th frame of data, A02 appears in the 12th frame, and A03 appears in the 10th frame. Here, the frame number is identified by sorting according to the index recorded in the monitoring sequence. Then, the first angle offset time of each number needs to be filtered, and the first offset point data of all numbers from the start of the disturbance to the response phase needs to be extracted to determine whether it is in the segment of the disturbance influence phase. For example, the angle offset of number A01 appears in the 8th frame of monitoring, and the start of the disturbance can be located within the 3 frames after the 5th frame. Therefore, A01 should be included in the object that needs to be responded to. Next, confirm whether A02, A03 to A20 meet the condition of the first angle change within the interference range and record their location. Then, organize the set of numbers with response data according to the ascending order of the numbers, for example, A01, A02, A03, A05, A06. It is necessary to determine whether there are any number jumps. For example, if A04 is missing between A03 and A05, it is judged as a discontinuous segment and removed. Or further confirm whether it is caused by data anomalies. If it is a true no-response state, no supplementary recording is made. Finally, retain the set of numbers with continuous numbering and response time within the disturbance period, such as the two segments A01 to A03 and A05 to A06, as continuous response areas for summary processing to obtain the attitude response number range.

[0074] S202: Based on the attitude response number range, extract the response time sequence corresponding to adjacent numbers, determine whether there is cross-response in time changes, identify number pairs whose time response and number arrangement are not synchronized, extract the interleaved number segments, and obtain a set of time sequence abnormal numbers;

[0075] First, extract the response time data between adjacent numbers. Then, extract the corresponding first frame time of the response according to the numbering order, forming a number-time sequence. For example, numbers B01, B02, B03, and B04 correspond to response frame sequences 10, 12, 11, and 13 respectively. In this case, the response time order between B02 and B03 does not satisfy the numbering order logic. This type of inconsistency needs to be judged and compared item by item. Specifically, for any pair of adjacent numbers, such as B02 and B03, first extract their numbering order from smallest to largest (B02 first, B03 last), then extract their response times to compare whether B02 is earlier than B03. If this time order is not satisfied, it is determined that the time response is inconsistent with the numbering order. Simultaneously, a positional relationship judgment is performed to confirm whether they are in spatial arrangement. For directly adjacent components, such as B02 and B03, if there are no other numbers between them and the response times are in reverse order, then this number pair is recorded as an asynchronous data pair and the process continues to traverse. For example, if the response times of B03 and B04 are 11 and 13 respectively, and the order is normal, then it is skipped. Finally, it can be concluded that B02 and B03 are an interleaved sequence. If there are subsequent segments such as B06, B07, and B08 where the number increases and the response time decreases, such as B06 responding to frame 14, B07 to 13, and B08 to 15, then B06 and B07 still need to be added to the inconsistent sequence. After sorting out all mismatched number pairs, the complete number segments involving these numbers are extracted, and the numbers are filtered to remove duplicates and non-interleaved combinations, resulting in a set of time-sequence abnormal numbers.

[0076] S203: Based on the set of time-series abnormal numbers, filter structural segments with adjacent response times and consecutive numbers, determine whether the numbers have no jumps in the arrangement direction and the time intervals are adjacent, extract the part with consecutive numbers and adjacent time, and obtain the response behavior time sequence chain.

[0077] First, extract the response frame positions corresponding to all numbers and arrange them in numerical order. Simultaneously, include the physical arrangement position of each number. For each pair of adjacent numbers, perform a dual-judgment operation: First, check if the number order is continuous, i.e., whether numbers B21, B22, B23, etc., satisfy the requirement of no skipped numbers. If there are missing numbers, such as B21 and B23, they are considered discontinuous and not included in subsequent judgments. Second, check if the difference between response frame positions is within the allowable range. For example, set the time difference judgment benchmark to ±1 frame. If response frame 35 for number B21, 36 for B22, and 38 for B23, then B21 and B22 meet the condition, while B22 and B23 exceed the range and need to be checked. During the judgment process, two conditions must be met simultaneously: consecutive numbering and compliant time difference. The qualified numbering segments will be merged into the traceable numbering set, and the judgment will continue to proceed in sequence. All segments that do not meet the conditions will be screened out, and finally a sequence containing consecutive numbers such as B21, B22, B23, B24 will be formed. Such segments must meet two criteria: first, the numbers must be consecutive without gaps in the arrangement; second, the corresponding response times must be consecutive frames or the frame difference must be within ±1. Such numbering segments represent spatially continuous and temporally similar event processes. Through the above dual conditions, segments are selected one by one, and the numbering range of multiple numbering segments is unified and consolidated to obtain the temporal chain of response behavior.

[0078] Please see Figure 4 The specific steps of S3 are as follows:

[0079] S301: Based on the arrangement structure of consecutive numbers in the temporal chain of response behavior, extract the arrangement direction of adjacent numbers, compare the spatial continuity relationship, filter the structure sequence with consistent direction and consecutive numbers, and obtain the consecutive number segment of arrangement direction.

[0080] First, using the numerical order as the initial index, the arrangement position corresponding to each number is extracted from the graphic coordinate data. Adjacent number pairs are then paired sequentially. For each pair, the arrangement direction angle is calculated using the direction of the line connecting their corresponding coordinate points. The angle difference is used as the judgment criterion. For example, if the coordinates corresponding to number A01 are (5.0, 3.0), A02 are (6.0, 3.0), and A03 are (7.0, 3.0), then the direction angles between A01 and A02, and between A02 and A03, are both 0°, and they are considered to have the same direction. If a sudden change in direction angle occurs in the middle segment, for example, the angle of A04 is 45°, then A03 and A04 are judged to have a change in direction and must be removed from the continuous segment. The direction difference is used to determine the direction. The judgment benchmark is ±5°. Values ​​outside this range are considered inconsistent. The process continues to compare the direction of all adjacent numbers, and the consistency of the direction of each number segment is judged. All consecutive number segments with direction angles within the allowable error range are retained. At the same time, pairs with skipped numbers in the numbering order are removed. For example, a jump from A07 to A10 is considered a non-consecutive number pair and is discarded. The judgment starting point is continuously advanced and each set of numbers that meets the conditions is recorded. Inconsistent numbers are filtered out by combining the numbering order and the continuity of direction angles. The spatial arrangement relationship is also judged to see if it meets the continuation standard. Finally, the number segments with consecutive numbers and stable arrangement direction are used as the screening result to obtain consecutive number segments in the arrangement direction.

[0081] S302: Based on the continuous numbering segment of the arrangement direction, extract the trend of the arrangement direction between the first and last numbers, filter the positions where the direction deviates, extract the numbers whose connection direction changes according to the changes in the arrangement before and after, and obtain the set of direction offset trigger numbers.

[0082] First, extract the first and last numbers of each numbered segment in numerical order. Calculate the connection direction vector using their corresponding coordinates. Extract the straight line between the first and last numbers in the graphic space to obtain the directional trend of each consecutive numbered segment. Combine this with point-by-point comparison of the directional angle changes between adjacent number pairs to identify abrupt changes in the connection direction at points where the angle changes. For example, if the directional angle of numbers D01 to D05 gradually remains at 90°, while there is a deviation towards the northeast between D06 and D07, then D06 is taken as the directional offset point. Simultaneously, determine whether the arrangement relationship before and after the numbered segment still maintains continuity, i.e., whether the numbers are in sequence. There are no jumps in the column. When the direction changes, the node number is extracted and added to the direction trigger judgment range. This operation is performed sequentially for multiple number segments. For each group of number nodes where the direction trend changes, a coordinate difference and angle change comparison are performed. The difference criterion is the spatial position change angle threshold between pairs of numbers. If the direction angle change is greater than 10°, it is included in the range of changed nodes. Combining the sequential continuity between numbers and the trend of direction angle change, nodes with number jumps or no angle deflection are filtered out. Finally, the set of numbers where the connection direction changes abruptly is extracted from all positions where the direction deflection occurs, and the set of direction offset trigger numbers is obtained.

[0083] S303: Based on the set of trigger numbers for directional offset, compare the corresponding situation with the direction of the disturbance path according to the number order, extract the continuous numbered paths with the same directional trend, and obtain the response path structure linked list;

[0084] First, extract the coordinates of each number in sequence. Calculate the direction angle of the connecting line based on the orientation of that position. Construct connecting lines between each pair of numbers and measure their azimuth angles to obtain the spatial extension direction of each pair of numbers. Call the disturbance path direction angle sequence and compare the extracted connecting directions with the disturbance direction sequentially according to the number order. If the angle difference is within the allowable range, the directions are considered consistent. The allowable range can be set to ±15°. Taking the comparison of numbers D05-D06 with the disturbance direction vector as an example, if the direction difference is 12°, the consistency condition is met. Continue to judge whether subsequent numbers within adjacent number segments that meet the condition still meet the angle direction difference requirement. If several consecutive number segments meet the condition, retain them as continuous path segments with consistent direction trends. After completely traversing all numbers, filter out numbers with skipped positions, excessively large angle differences, or discontinuous numbers, retaining all number paths that meet the disturbance direction consistency characteristics. Connect the adjacent number relationships sequentially into a continuous structural path. Finally, output the complete path sequence according to the original number order to obtain the response path structure linked list.

[0085] Please see Figure 5 The specific steps of S4 are as follows:

[0086] S401: Based on the position of the end node in the response path structure linked list, collect the coordinate sequence of the node in the component array diagram, extract the arrangement direction and direction change of adjacent numbers, judge the continuity of the direction trend based on the connection state of the node arrangement in the array structure, and obtain the end arrangement direction distribution.

[0087] First, extract the coordinate points corresponding to all end numbers in the component array diagram in sequence, and construct a coordinate sequence according to the ascending order of the numbers. Calculate the arrangement direction vector between adjacent numbers using the horizontal and vertical axis values ​​of the coordinate points. Calculate the direction change between each segment using sequential segments such as AB, BC, etc. The direction change value is obtained by the angle difference between two adjacent vectors. Then, aggregate the direction change values ​​of each segment in the sequence. Determine if the angle value is within the set allowable range of direction difference (e.g., within 5° is considered continuous direction). If the angle value of a segment exceeds the set range, it is determined that there is a direction deflection. Based on this, determine the segment... Whether the continuity of the point arrangement is continuous, i.e. whether the numbering jumps or whether there are interrupted areas in space, is determined by the difference in numbering and the distance between coordinate points. The stable and offset segments of the directional trend are marked in combination with the directional change. Taking the numbering end segments D17-D18 and D18-D19 as examples, if D17-D18 is eastward and D18-D19 is southeastward with an angle of 40°, then the continuity requirement is not met. Based on such segments, the end number of the continuous segment is traced forward. Finally, the summary distribution of the directional trend is extracted from the continuous segments with stable directional differences to obtain the directional distribution of the end arrangement.

[0088] S402: Based on the distribution of the end arrangement direction, extract the directional connection status of the component numbers in adjacent areas, identify whether there is directional offset and arrangement distortion position between the numbers, and determine whether the arrangement direction is continuously misaligned by the trend of the number direction change to obtain the deflection and misalignment feature segment.

[0089] First, extract the coordinate data of adjacent numbers. Determine if the direction of the adjacent numbers' arrangement is consistent with the direction of the last number by analyzing the increase or decrease in the horizontal and vertical displacement directions. Then, retrieve the coordinate vector angle between each pair of numbers and calculate the angle between adjacent vectors. Based on the sequence of consecutive numbers, obtain the angle change segment by segment. Check if the angle between the connecting line segments of the numbers deviates by more than 20 degrees. If the angle of a segment in a consecutive numbering segment exceeds this value, it is identified as a direction deflection node. The numbers before and after this node, along with the current number, are grouped into segments to be identified. Finally, combine the horizontal coordinates of the numbers with the horizontal coordinates of the spatial coordinates. The longitudinal offset is used to determine whether the segment has crisscrossing or tilting. If there is a sudden change in the longitudinal coordinate of the horizontal numbering sequence exceeding 0.6 meters, it is judged as a misalignment and the segment is recorded as a potentially distorted numbering segment. The projection trend of adjacent numbers in the array diagram is used to determine whether there is a reverse bend. If three points in the numbering sequence form an angle reversal in the two-dimensional coordinates and the angle between the turning points is less than 150 degrees, it is judged as a distorted arrangement. Finally, the above numbering segments are integrated into a region according to spatial continuity, and all numbering segment sequences that meet the above conditions are output to obtain the deflection and misalignment feature segment.

[0090] S403: Based on the deflection and staggered feature segments, analyze the continuity of the numbering arrangement, extract the edge position of the deflection structure in space, filter the areas where the numbering is interrupted and the structural direction is turned, and obtain a list of evolutionary end boundary regions.

[0091] First, the numbers are arranged in ascending order. The coordinates of each number in the array diagram are extracted. The continuity of the numbering arrangement is determined by comparing the interval between adjacent numbers to see if it is 1. When a number is not continuous with the previous number and the spatial coordinates are more than 1.2 meters apart, it is temporarily stored as the initial break point. Then, it is further determined whether there is an angular change in the arrangement direction on both sides of the break point. Specifically, the three numbers on each side of the break point are formed into a line segment, and the angle value of the connecting vector in the two-dimensional plane is extracted. If the angle at the corner is less than 160 degrees, it is determined that the direction has been twisted. By aggregating the positions with the above characteristics, multiple sub-regions are formed. Then, the starting and ending numbers of each region are extracted. The correspondence between the number and the coordinates is extracted again at these edge positions. The overall extension direction of the numbering arrangement trajectory is calculated by the spatial coordinate displacement. Combined with the criteria of whether there are missing numbers or abrupt changes in direction before and after the break point, these regions are divided into characteristic segments with arrangement changes and numbering interruptions. Finally, these segments are reorganized in ascending order of number, and the set of regions with discontinuous numbers and changes in spatial arrangement trend is summarized to obtain the list of evolutionary end boundary regions.

[0092] Please see Figure 6 The specific steps of S5 are as follows:

[0093] S501: Based on the structure numbers in the list of boundary regions at the end of the evolution, extract the attitude change sequence of the corresponding number position during the disturbance influence stage, determine whether the number arrangement order has the same direction as the response sequence, filter out continuous segments with the same number change direction, and obtain the number group corresponding to the attitude response.

[0094] First, the sequence of attitude parameter changes at each position corresponding to each number during the interference process is extracted. Specifically, the angle information of each position is extracted in chronological order and then represented as an array to indicate the attitude orientation at each moment. In a real-world scenario, for example, for the numbering segment A01 to A05, the attitude angles at five time points are sampled from the time sequence corresponding to each number. For example, A01 is (12, 14, 18, 23, 27), A02 is (15, 17, 22, 26, 30), A03 is (16, 18, 23, 27, 31), A04 is (14, 17, 21, 24, 28), and A05 is (10, 13, 18, 22, 25). Then, the direction of increase or decrease between adjacent points in the attitude angle curve is categorized as rising, falling, or remaining constant, resulting in the trend sequence corresponding to each number. For example, A01 shows a continuous increase, A02 shows a continuous increase, A03 shows a continuous increase, and so on. If A04 and A05 are continuously increasing, then this numbering segment exhibits a consistent attitude change direction during the disturbance phase. The numbers are then compared in order to determine if adjacent numbers satisfy the condition of a 1-digit interval and consistent arrangement direction. If the five numbers are in the order of A01 to A05 without any jumps or reversals, the condition is met. Further analysis is performed to determine if the change time points of the attitude response curves are reversed. For example, if A01 starts responding at second 1, A02 at second 2, A03 at second 3, A04 at second 4, and A05 at second 5, then the response order matches the numbering order. Conversely, if A03 starts responding earlier than A01, the filtering condition is not met. Finally, the numbering segments that simultaneously satisfy the conditions of continuous numbers, consistent arrangement direction, and sequential response start times are extracted to obtain the corresponding attitude response numbering groups.

[0095] S502: Based on the attitude response corresponding number group, extract the correspondence between the number arrangement direction and the wind direction interference path, determine whether the number continues to expand along the wind direction in space, filter out the number positions with inconsistent expansion direction and path offset, and obtain the set of numbers with consistent direction.

[0096] First, the position coordinates of each number in the numbering group are read sequentially, and their arrangement order in the component array diagram is displayed in a two-dimensional coordinate format. Then, the disturbance path direction vector is determined and denoted as the overall disturbance trend direction. For example, if the wind direction in the component array is expanding from west to east, the positive X-axis is used as the wind direction reference. Next, the spatial arrangement relationship between every two adjacent numbers is obtained, and the angle between their vector direction and the disturbance direction vector is calculated. It is determined whether it is less than 30 degrees. If the angle is large or the directions are reversed, it is considered that the directions are inconsistent. This is used as a criterion to remove nodes with obvious deviations in direction from the numbering. Furthermore, the remaining numbers are re-compared with the expansion trend according to the arrangement order. If there is a pair of numbers in the middle of the path... If a sudden directional deviation or reversal occurs, its number is marked as a directional deviation item and excluded. In a real-world scenario, for example, under a west-to-east wind disturbance, the original numbering sequence is A01, A02, A03, A04, and A05. A01 to A03 are arranged along the X-axis. Due to an installation deviation, A04 is positioned towards the northeast, causing the direction of the connection vector between A03 and A04 to deviate from the disturbance direction by more than a set judgment value. Therefore, A04 and its subsequent numbers are removed and no longer included in the continuous extension path. Through the above operation, the consistency of the numbered path direction is filtered, and finally, the numbered segments that are consistent with the disturbance direction on the spatial path and have not undergone a sudden change in direction are extracted, resulting in a set of numbered segments with consistent direction.

[0097] S503: Call the set of numbers with consistent direction, analyze whether the numbering has direction interruption and sequence jump during the response phase, filter the structural segments with continuous numbering and continuous response direction, and obtain the high pile structure disturbance trend prediction sequence.

[0098] First, extract the order of all the numbers. For each pair of adjacent numbers, determine the difference in numbers to see if the numbers increase sequentially in natural number order with a step size of 1. If there is a difference between adjacent numbers greater than 1, mark it as an interruption point. Next, obtain the time series data of the response behavior corresponding to each number during the interference phase. Analyze the changes in the response time corresponding to each pair of adjacent numbers. By sorting the time differences, we can identify whether there is a reverse jump in the response direction. Then, match the number arrangement direction with the response sequence one by one to determine whether there is a phenomenon where the number order increases but the response time advances. Taking numbers A01 to A06 as an example, ideally, their response time should be gradually delayed from early to late. If the actual response time sequence is A01, A03, A06, A07, A08, A09, A0 ... If A02, A04, A05, and A06 are selected, it can be determined that there is a jump in response direction between A03 and A02. Then, the spatial coordinate data is used to check whether there is an anomaly in the positional distance between the numbers in the array arrangement. If the numbers A03 and A02 are not adjacent in space, it is further confirmed that they are the break point in response direction. At the same time, this abnormal segment is excluded from the number set. The number continuity and response direction continuity are checked for the remaining number set. If there is no number jump in the number sequence, the response time is monotonically changing, and the response direction is consistent with the arrangement trend, it is determined to be a continuous response segment that meets the conditions. Finally, based on the above, these continuous numbers with stable response direction and no number jump in the number arrangement are connected into a set of behavior prediction clue paths to obtain the high pile structure disturbance trend prediction sequence.

[0099] Please see Figure 7 A high-pile photovoltaic disaster prediction system integrating big data includes:

[0100] The structural offset identification module obtains the structural number, arrangement order, angle change trend, force trajectory and incident angle offset path, compares the attitude change direction according to the number, filters out the reverse offset and torsional synchronous segments, and obtains the structural response offset group.

[0101] The temporal behavior extraction module retrieves disturbance state records based on the structural response offset group, extracts the time point from the disturbance to the attitude change, determines whether the numbers overlap, and delineates the continuous proximity region in sequence to obtain the temporal chain of response behavior.

[0102] The response path construction module analyzes the numbering connection relationship and direction based on the temporal chain of response behavior, determines whether there are points of directional change, and traces the direction of disturbance according to the connection order to obtain the response path structure linked list.

[0103] The terminal boundary determination module is based on the response path structure linked list. It retrieves the terminal position and adjacent trends, judges the trend deflection and misalignment, delineates the trend interruption area, and obtains a list of evolution terminal boundary areas.

[0104] The disturbance trend prediction module is based on the list of boundary regions at the end of the evolution. By comparing the attitude change sequence, it judges the consistency between the numbering sequence and the response direction, extracts continuous and consistent regions, and obtains the disturbance trend prediction sequence of the high pile structure.

[0105] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A disaster prediction method for high-pile photovoltaic systems integrating big data, characterized in that, Includes the following steps: S1: Obtain the structural number and arrangement order of the high-pile photovoltaic modules, collect the angle change, pile top force and incident angle path under disturbance, analyze the direction of adjacent attitude change, screen the continuous offset structure, and obtain the structural response offset group. S2: Based on the number set of the structural response offset group, extract the time point from the disturbance to the attitude change, determine whether adjacent numbered responses are interleaved, identify the continuous time response region according to the numbering order, and obtain the temporal chain of response behavior. S3: Based on the arrangement structure of consecutive numbers in the temporal chain of the response behavior, analyze the connection direction of the numbers, identify the position of directional change in the path, track the direction of interference according to the number order, and obtain the response path structure linked list; S4: Based on the position of the end node in the response path structure chain list, extract the arrangement trend and direction change path of adjacent regions, determine whether the end structure is misaligned or deflected, divide the arrangement interruption region, and obtain the list of evolution end boundary regions. S5: Based on the structure number of the boundary region list at the end of the evolution, compare the number arrangement with the interference path, filter regions with the same order and direction, extract the structural response trend, and obtain the high pile structure disturbance trend prediction sequence.

2. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The structural response offset group includes structural number, arrangement relationship, reverse offset component, and torsional synchronous structural segment. The response behavior time sequence chain includes response time point sequence, reaction order, structural number correspondence, and temporal continuous region. The response path structure chain list includes start number, end number, connection direction, and response direction change node. The evolution end boundary region list includes end concentration node, arrangement trend, structural misalignment region, and arrangement trend interruption edge. The high pile structure disturbance trend prediction sequence includes numbering order, disturbance direction consistency region, and continuous numbering path.

3. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The incident angle path refers to the angle change between the incident light direction and the component surface obtained by the illumination angle measuring device and the angle change trajectory formed in time sequence; The attitude change direction refers to the directional change of the panel tilt angle over time during wind disturbances.

4. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The direction of interference tracking refers to analyzing the spatial connection direction of the structure numbers pair by pair in the component sequence with a determined response time to identify the disturbance propagation path; The so-called arrangement interruption region refers to the region of connection break and direction change caused by the interruption of the continuity of structural numbering and spatial arrangement during the extension of the disturbance path.

5. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the structural number and arrangement order of each column in the high-pile photovoltaic module array, collect the angle change, pile top directional force and incident angle path under wind disturbance, compare the arrangement continuity according to the numbering order, filter the module column area with consistent numbering spacing, and obtain the structural arrangement mapping segment. S102: Based on the structure arrangement mapping segment, determine the attitude change direction of the components corresponding to adjacent numbers, compare the trend of angle change before and after wind direction disturbance, filter out number segments with alternating and continuous arrangement of directions, and obtain the direction offset alternating region. S103: Based on the alternating directional offset region, extract the correspondence between the numbering order and the direction of torsion angle change, determine whether the consistent direction continues during the interference period, filter out numbering segments with consecutive numbers and the angle change maintaining order, and obtain the structural response offset group.

6. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the number set of the structural response offset group, extract the time point when the structural angle offset first occurs within the disturbance start stage, identify the number of the attitude response that occurs within the disturbance period, filter the continuous number segment according to the number order, and obtain the attitude response number range. S202: Based on the attitude response number range, extract the response time sequence corresponding to adjacent numbers, determine whether there is cross-response in time change, identify number pairs whose time response and number arrangement are not synchronized, extract the interleaved number segment, and obtain a set of time sequence abnormal numbers; S203: Based on the set of time-sequence abnormal numbers, filter structural segments with adjacent response times and consecutive numbers, determine whether the numbers have no jumps in the arrangement direction and the time intervals are adjacent, extract the part with consecutive numbers and adjacent time, and obtain the response behavior time sequence chain.

7. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the arrangement structure of consecutive numbers in the temporal chain of the response behavior, extract the arrangement direction of adjacent numbers, compare the spatial continuity relationship, filter the structure sequence with consistent direction and consecutive numbers, and obtain the consecutive number segment of arrangement direction. S302: Based on the continuous numbering segment of the arrangement direction, extract the trend of the arrangement direction between the first and last numbers, filter the positions where the direction deviates, extract the numbers whose connection direction changes according to the changes in the arrangement before and after, and obtain the set of direction offset trigger numbers. S303: Based on the set of direction offset trigger numbers, compare the correspondence between the numbers and the direction of the disturbance path, extract the consecutive numbered paths with the same direction trend, and obtain the response path structure linked list.

8. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the position of the end node in the response path structure linked list, collect the coordinate sequence of the node in the component array diagram, extract the arrangement direction and direction change of adjacent numbers, judge the continuity of the direction trend according to the connection state of the node arrangement in the array structure, and obtain the end arrangement direction distribution. S402: Based on the distribution of the end arrangement direction, extract the directional connection status of the component numbers in adjacent regions, identify whether there is a directional offset and arrangement distortion position between the numbers, and determine whether the arrangement direction is continuously misaligned by the trend of the number direction change to obtain the deflection and misalignment feature segment. S403: Based on the deflection staggered feature segment, analyze the continuity of the numbering arrangement, extract the edge position of the deflection structure in space, filter the areas where the numbering is interrupted and the structure direction changes, and obtain a list of evolutionary end boundary regions.

9. The high-pile photovoltaic disaster prediction method integrating big data according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the structure number in the list of boundary regions at the end of the evolution, extract the attitude change sequence of the corresponding position during the disturbance influence stage, determine whether the number arrangement order has the same direction as the response sequence, filter out continuous segments with the same direction of number change, and obtain the number group corresponding to the attitude response. S502: Based on the attitude response corresponding number group, extract the correspondence between the number arrangement direction and the wind direction interference path, determine whether the number continues to expand along the wind direction in space, filter out the number positions with inconsistent expansion direction and path offset, and obtain a set of numbers with consistent direction. S503: Call the set of consistent direction numbers, analyze whether the numbering has direction interruption and sequence jump during the response phase, filter the structural segments with continuous numbering and continuous response direction, and obtain the high pile structure disturbance trend prediction sequence.

10. A high-pile photovoltaic disaster prediction system integrating big data, characterized in that, The system is used to implement the high-pile photovoltaic disaster prediction method integrating big data as described in any one of claims 1-9, and the system includes: The structural offset identification module obtains the structural number, arrangement order, angle change trend, force trajectory and incident angle offset path, compares the attitude change direction according to the number, filters out the reverse offset and torsional synchronous segments, and obtains the structural response offset group. The temporal behavior extraction module retrieves the disturbance state record based on the structural response offset group, extracts the time point from the disturbance to the attitude change, determines whether the numbers overlap, and delineates the continuous proximity region in sequence to obtain the temporal chain of response behavior. The response path construction module analyzes the numbering connection relationship and direction based on the temporal chain of the response behavior, determines whether there are points of directional change, tracks the direction of disturbance according to the connection order, and obtains the response path structure linked list. The terminal boundary determination module, based on the response path structure linked list, retrieves the terminal position and adjacent trends, determines the direction deviation and misalignment, delineates the trend interruption area, and obtains a list of evolution terminal boundary areas. The disturbance trend prediction module, based on the list of boundary regions at the end of the evolution, compares the order of attitude changes, determines the consistency between the numbering order and the response direction, extracts continuous and consistent regions, and obtains the disturbance trend prediction sequence of the high pile structure.

Images