Disaster process operation window fragmentation feature extraction and evaluation method and system

Abstract

The present application relates to the technical field of agricultural meteorological disaster assessment, and more particularly to a method and system for extracting and evaluating the fragmentation characteristics of the operation window in a disaster process. The method comprises the following steps: obtaining meteorological time series data of winter wheat suitable for harvesting, detecting and segmenting continuous rain and waterlogging events, and extracting multi-dimensional feature parameters; generating a disaster impact index based on historical quantile benchmarks for standardized mapping; sliding and dividing candidate operation windows and calculating the effective operation time, continuity coefficient, fragmentation index and efficiency constraint coefficient of the window, and weighting and aggregating them into a structure impact comprehensive score; selecting the window with the lowest score as the recommended machine harvesting window, and combining time parameters, agricultural machinery ownership and workable area to generate a scheduling scheme. The method improves the scientificity of machine harvesting window selection and operation efficiency, and effectively reduces disaster losses.

Description

Technical Field

[0001] This invention relates to the field of agricultural meteorological disaster assessment technology, and in particular to a method and system for extracting and evaluating fragmented features of disaster process operation windows. Background Technology

[0002] Mechanical harvesting of winter wheat during its optimal harvesting period is often affected by adverse weather conditions such as continuous rain and waterlogging, resulting in fragmented and dispersed harvesting windows. Current technologies typically rely on short-term weather forecasts or historical statistics of sunny days for the same period to determine suitable harvesting windows, using the absence of precipitation or soil moisture content below a threshold as the condition for operation, and then manually or semi-automatically selecting consecutive rainless periods. This conventional approach focuses on simple threshold discrimination of single meteorological elements, lacking the ability to systematically extract and quantify the structural characteristics of disaster events in the temporal dimension (such as the start and end times of the event, duration, precipitation intensity, and soil moisture dynamics).

[0003] Using the presence or absence of precipitation or excessive soil moisture as operational conditions alone is insufficient to reflect the complex impacts of continuous rain and waterlogging events. For example, a short-duration but intense rainfall event may cause soil moisture content to rise rapidly and remain elevated for several days, while a weak rainfall event may have an effect equivalent to severe waterlogging due to prior soil saturation. Conventional practices cannot distinguish the actual operational feasibility in such situations and may easily misjudge windows with theoretically no precipitation but actually excessively wet soil as operational windows.

[0004] Existing methods lack objective evaluation indicators for the "fragmentation" of work windows. Farmers or agricultural machinery dispatchers often rely on personal experience to judge the merits of multiple candidate windows, making it difficult to quantify and compare the effective working time, work continuity, and the distribution of discontinuities caused by disasters within each window on a global scale. This experience-based decision-making can easily lead to a mismatch between recommended windows and the actual number of agricultural machines and the available workable area, resulting in unreasonable dispatching plans, delayed harvests, or idle machinery. Therefore, there is an urgent need for a technical solution that can standardize the measurement and evaluation of window fragmentation based on the structural characteristics of disaster events. Summary of the Invention

[0005] This invention provides a method and system for extracting and evaluating fragmented features of disaster process operation windows, which can solve the problems in the prior art.

[0006] A first aspect of this invention provides a method for extracting and evaluating fragmented features of disaster process operation windows, comprising: Meteorological time-series data of a specified spatial range during the winter wheat harvest period are obtained. Based on the continuous rain and waterlogging judgment rules, continuous rain and waterlogging events are detected and identified. Event-based segmentation is performed on each disaster event, and feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence are extracted. Based on the historical quantile benchmark dataset of the same period in this spatial range, quantile standardization mapping is performed on the feature parameters of each disaster event in different dimensions, transforming the feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster impact index. Candidate operation windows are divided along the time axis of the harvest period with a preset operation window duration as the step size. Within each candidate window, the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient are calculated for each window using a multidimensional standardized impact index as input. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The structural impact scores of each window are sorted from low to high. The candidate window with the lowest score is selected as the recommended machine harvesting window. The machine harvesting operation scheduling plan is generated and output by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

[0007] Quantile normalization mapping was performed on the feature parameters of each disaster event across different dimensions, including: Extract from the historical quantile benchmark dataset of the same period in the spatial range the sample sets of duration of continuous rain events, average precipitation intensity, duration of waterlogging events, and hourly soil moisture content change sequence of waterlogging events for each historical year within the same period as the current harvest period. After sorting the duration sample set in ascending order, the cumulative empirical quantile level of each sorting position is calculated based on the total number of samples. A duration quantile mapping table is constructed with the quantile of the sorting position as the vertical axis and the corresponding duration sample value as the horizontal axis. A mean precipitation intensity quantile mapping table is constructed in the same way. The arithmetic mean of each sequence in the hourly soil moisture content change sequence sample set is calculated to obtain the moisture content mean sample set. The moisture content mean quantile mapping table is constructed in the same sorting and quantile method. At the same time, the proportion of times in each sequence that exceed the preset waterlogging threshold is counted to construct the threshold duration quantile mapping table. The actual characteristic values ​​of each disaster event are determined by linear interpolation between adjacent quantile nodes in the corresponding mapping table, and each quantile value is used as the standardized impact index component of the event in the corresponding dimension. All components are organized into a multidimensional standardized impact index according to the time series dimension, duration dimension, precipitation intensity dimension, and soil moisture content dimension.

[0008] The quantile values ​​of the actual characteristic values ​​of waterlogging events are determined by linear interpolation in the mean quantile mapping table of water content and the quantile mapping table of the proportion of time exceeding the threshold, including: For the hourly soil moisture content change sequence of a waterlogging event, the hourly soil moisture content values ​​of each time point in the sequence are read sequentially. All hourly soil moisture content values ​​are summed one by one and divided by the total number of time points in the sequence to obtain the arithmetic mean of the hourly soil moisture content of the waterlogging event. The arithmetic mean is then linearly interpolated between two adjacent quantile nodes in the hourly soil moisture content mean quantile mapping table to obtain the mean quantile value. The relationship between the hourly soil moisture content value and the preset waterlogging soil moisture content threshold is scanned hourly within the start and end time of the waterlogging event. The number of times the hourly soil moisture content exceeds the threshold is counted to obtain the cumulative number of times the threshold is exceeded. The cumulative number of times the threshold is exceeded is divided by the total number of times the waterlogging event lasts from the start to the end to obtain the percentage of time exceeding the threshold. The percentage is then calculated by linear interpolation between two adjacent quantile nodes in the percentage of time exceeding the threshold quantile mapping table to obtain the quantile value of the duration percentage. The mean quantile and the duration quantile are weighted and summed according to the preset mean contribution weight and duration contribution weight. The weighted summation result is used as the standardized impact index component of the waterlogging event in the soil moisture dimension.

[0009] The effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient of each window are calculated and weighted to form a comprehensive score of window structure impact, including: The window time range is defined by the total number of hours of the candidate operation window. For each hour in the window, the maximum value of each dimension component of the multidimensional standardized impact index is taken as the representative impact index of that hour. Each representative impact index is compared with the preset impact judgment threshold one by one. Hours with representative impact indices lower than the threshold are marked as workable hours. The total number of hours of all workable hours is accumulated to obtain the effective operation time of the window. Perform a continuous segment scan operation on all available workable time slots arranged in chronological order within the window. Identify the continuous workable time slots composed of consecutive workable time slots from the scan results, and take the ratio of the longest duration of the longest duration to the total duration of the window as the work continuity coefficient. The fragmentation index is obtained by dividing the total number of consecutive rainy and waterlogging events within the statistical window time range by the total number of disaster events. The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function; The window's effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient are each normalized using the min-max method. The four normalized indicators are then weighted and summed according to a preset indicator weight vector to obtain the comprehensive score of the window structure impact.

[0010] The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function, including: Establish a continuous mapping relationship between the mean value of the impact index component of the soil moisture content dimension and the soil bearing capacity attenuation coefficient. Map the lower limit value of the impact index component corresponding to the dry soil state to the full-capacity coefficient that characterizes the agricultural machinery to maintain rated operating efficiency, and map the upper limit value of the impact index component corresponding to the fully saturated soil state to the minimum allowable operating capacity coefficient that characterizes the agricultural machinery to maintain only the minimum operating efficiency. All intermediate state nodes between the lower limit value and the upper limit value are connected in sequence by a monotonically decreasing smooth continuous curve to form a mapping relationship curve. The arithmetic mean of the standardized impact index components of soil moisture content at each time step in the window is calculated to obtain the mean value of the impact index components. The mean value of the impact index components is then substituted into the above continuous mapping relationship. The specific value of the soil bearing capacity attenuation coefficient corresponding to the mean value of the impact index components is determined by linear interpolation between the two adjacent mapping nodes that are closest to the mean value of the impact index components. The equivalent number of working hours is obtained by multiplying the determined soil bearing capacity attenuation coefficient by the total number of working hours in the window. The difference between the total number of working hours in the window and the equivalent number of working hours is calculated, and the ratio obtained by dividing the difference by the total number of working hours in the window is used as the work efficiency constraint coefficient.

[0011] The machine harvesting operation scheduling plan is generated and output by combining the recommended window's time parameters, available agricultural machinery inventory, and the workable area within the window, including: Obtain the total number of registered and callable winter wheat harvesting machines within the spatial range and the rated hourly operating area of ​​each machine. Multiply the rated hourly operating area of ​​a single machine by the number of hours that can be worked within the recommended harvesting window to obtain the theoretical operating area of ​​a single machine within the recommended window period. Obtain the total planting area of ​​winter wheat to be harvested within the spatial range, compare the total planting area with the theoretically operable area of ​​a single agricultural machine, and divide into two processing scenarios. If the total planting area is not greater than the theoretically operable area of ​​a single agricultural machine, the entire planting area is allocated to one agricultural machine to complete the entire harvesting task independently. If the total planting area is greater than the theoretically operable area of ​​a single agricultural machine, the total planting area is evenly distributed according to the total number of agricultural machines, and the geographical range of the corresponding working field is marked for each agricultural machine. The machine departure time is obtained by shifting forward the preset time required for the agricultural machinery to assemble, based on the start time of the recommended harvesting window. The start and end times of each agricultural machine are allocated and calculated one by one according to the departure time of the agricultural machine plus the number of working hours in the window. The output is a complete scheduling scheme with the agricultural machine number as the primary key and including the geographical range of the corresponding field, the start and end times of the operation, and the departure time of the agricultural machine.

[0012] A second aspect of this invention provides a system for extracting and evaluating fragmented features of disaster process operation windows, comprising: The event detection unit is used to acquire meteorological time series data of a specified spatial range during the winter wheat harvest period, detect and identify continuous rain events and waterlogging events based on the continuous rain judgment rules and waterlogging judgment rules, perform event segmentation on each disaster event, and extract feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence. The shock standard unit is used to perform quantile standardization mapping on the feature parameters of each disaster event in different dimensions based on the historical quantile benchmark dataset of the same period in this spatial range, transforming feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster shock index. The window scoring unit is used to slide along the harvest period time axis to divide candidate operation windows with a preset operation window duration as the step size. Within each candidate window, the multidimensional standardized impact index is used as input to calculate the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient of the window window one by one. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The scheduling generation unit sorts the structural impact comprehensive scores of each window from low to high, selects the candidate window with the lowest score as the recommended machine harvesting window, and generates and outputs a machine harvesting operation scheduling plan by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

[0013] A third aspect of the present invention provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0015] This method can accurately identify continuous rain and waterlogging events from meteorological time-series data, and extract multi-dimensional feature parameters such as start and end times, duration, average precipitation intensity, and hourly soil moisture content changes, enabling refined event-based segmentation of disaster processes. By leveraging historical multi-year quantile benchmark datasets, feature parameters of different dimensions and magnitudes are uniformly mapped to the quantile scale, generating a multi-dimensional standardized disaster impact index. This effectively eliminates the interference of data heterogeneity on subsequent evaluations and improves the comparability of cross-regional and cross-year analyses.

[0016] After dividing candidate operation windows along the harvestable period time axis with a preset step size, the effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient are calculated for each window based on a multidimensional standardized impact index. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive structural impact score. This score can quantitatively characterize the degree of fragmentation impact of disaster events on mechanized harvesting operations within each window, providing a scientific basis for ranking the windows and avoiding the limitations of traditional methods that rely solely on a single indicator or subjective experience.

[0017] The system automatically selects the candidate window with the lowest comprehensive score as the recommended harvesting window, and generates a harvesting operation scheduling plan by combining window time parameters, available agricultural machinery inventory, and operable area. This plan can minimize the impact of disasters during high-impact periods, improve the continuity and efficiency of agricultural machinery operations, and reduce problems such as equipment idling and increased costs of cross-regional relocation caused by window fragmentation. It provides quantitative and operable technical support for agricultural machinery scheduling during the winter wheat harvesting season. Attached Figure Description

[0018] Figure 1 A flowchart illustrating the method for extracting and evaluating fragmented features of disaster process operation windows. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0020] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0021] Figure 1 This is a flowchart illustrating the method for extracting and evaluating fragmented features of disaster process operation windows according to an embodiment of the present invention, including: Meteorological time-series data of a specified spatial range during the winter wheat harvest period are obtained. Based on the continuous rain and waterlogging judgment rules, continuous rain and waterlogging events are detected and identified. Event-based segmentation is performed on each disaster event, and feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence are extracted. Based on the historical quantile benchmark dataset of the same period in this spatial range, quantile standardization mapping is performed on the feature parameters of each disaster event in different dimensions, transforming the feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster impact index. Candidate operation windows are divided along the time axis of the harvest period with a preset operation window duration as the step size. Within each candidate window, the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient are calculated for each window using a multidimensional standardized impact index as input. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The structural impact scores of each window are sorted from low to high. The candidate window with the lowest score is selected as the recommended machine harvesting window. The machine harvesting operation scheduling plan is generated and output by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

[0022] Quantile normalization mapping was performed on the feature parameters of each disaster event across different dimensions, including: Extract from the historical quantile benchmark dataset of the same period in the spatial range the sample sets of duration of continuous rain events, average precipitation intensity, duration of waterlogging events, and hourly soil moisture content change sequence of waterlogging events for each historical year within the same period as the current harvest period. After sorting the duration sample set in ascending order, the cumulative empirical quantile level of each sorting position is calculated based on the total number of samples. A duration quantile mapping table is constructed with the quantile of the sorting position as the vertical axis and the corresponding duration sample value as the horizontal axis. A mean precipitation intensity quantile mapping table is constructed in the same way. The arithmetic mean of each sequence in the hourly soil moisture content change sequence sample set is calculated to obtain the moisture content mean sample set. The moisture content mean quantile mapping table is constructed in the same sorting and quantile method. At the same time, the proportion of times in each sequence that exceed the preset waterlogging threshold is counted to construct the threshold duration quantile mapping table. The actual characteristic values ​​of each disaster event are determined by linear interpolation between adjacent quantile nodes in the corresponding mapping table, and each quantile value is used as the standardized impact index component of the event in the corresponding dimension. All components are organized into a multidimensional standardized impact index according to the time series dimension, duration dimension, precipitation intensity dimension, and soil moisture content dimension.

[0023] From historical meteorological archives, for the specific calendar period corresponding to the current suitable harvest period (e.g., the suitable harvest period for winter wheat in a certain region is from late May to early June each year), all identified and archived records of continuous rain events and waterlogging events from the same period over many years are extracted and compiled into four types of sample sets: a sample set of the duration of continuous rain events, a sample set of the average precipitation intensity of continuous rain events, a sample set of the duration of waterlogging events, and a sample set of hourly soil moisture content change sequences of waterlogging events. These four types of sample sets together constitute the benchmark reference system for subsequent quantile standardization mapping. Its core function is to place the characteristic values ​​of the current year's disaster events in the historical distribution context for relative intensity assessment, thereby eliminating the incomparability problem caused by differences in different physical dimensions and magnitudes.

[0024] After sorting the duration sample set in ascending order, let the total number of samples be . , No. The cumulative empirical quantile level corresponding to each sorting position is determined as follows: ,in The value range is 1 to , That is, the first The empirical quantile level corresponding to small sample values. The vertical axis represents the sample values ​​corresponding to the duration. The x-axis represents all node pairs recorded sequentially. This forms a duration quantile mapping table. This mapping table is essentially a discrete lookup table of the empirical cumulative distribution function, covering the complete distribution range from the minimum historical observation duration to the maximum historical observation duration. The same sorting and quantile calculation method is used for the mean precipitation intensity sample set, with precipitation intensity values ​​as the basis for the calculation. x-axis, corresponding quantile level Using the vertical axis as the ordinate, construct a quantile mapping table for average precipitation intensity. The construction logic of the two mapping tables is completely symmetrical, both using the empirical quantile formula at the midpoint to avoid boundary singularity problems caused by quantiles at the endpoints being 0 or 1.

[0025] The processing of the hourly soil moisture content variation sequence sample set involves two parallel statistical dimensions. For each hourly sequence in the sample set, the arithmetic mean of the soil moisture content values ​​at all times in the sequence is calculated to obtain the mean moisture content for that sequence. These mean moisture content samples are then compiled to form a mean moisture content sample set. Let the total number of samples in this set be... , No. The mean sample corresponding to each sorting position is quantile level is ,by x-axis A quantile mapping table of mean water content is constructed for the ordinate. Simultaneously, for each sequence, the number of sequences exceeding a preset waterlogging threshold is statistically analyzed. The number of times the threshold is exceeded is divided by the total number of times in the sequence to obtain the percentage of time exceeding the threshold. After summing and sorting the percentages of time exceeding the threshold for all sequences, a quantile mapping table for the percentage of time exceeding the threshold is constructed in the same way. Preset waterlogging threshold. The specific values ​​are determined based on local soil type and crop waterlogging tolerance, typically using a fixed percentage of field capacity as a reference standard. These values ​​are configured and written into the parameter file by agronomic experts during system initialization. These two mapping tables characterize the comprehensive impact of waterlogging events on soil operating conditions from two perspectives: the degree of sustained soil moisture and the proportion of periods of extreme waterlogging. Together, they form the basis for the quantile mapping of soil moisture content.

[0026] After constructing the four quantile mapping tables, for each disaster event detected and identified within the current harvest period, its actual feature value is used to perform quantile localization in the corresponding mapping table. The localization process uses a linear interpolation method: let a certain feature value be... Falling into two adjacent nodes in the mapping table and Between, that is Then the quantile value corresponding to this eigenvalue according to Calculation. When the actual eigenvalue exceeds the maximum node of the mapping table, the quantile value is set to 1.0, indicating that the event intensity exceeds all historical records of the same period; when the actual eigenvalue is below the minimum node, the quantile value is set to 0.0, indicating that the event intensity is weaker than all historical records of the same period. The introduction of linear interpolation allows the quantile value to change continuously between adjacent nodes, avoiding the step distortion caused by discrete table lookup, and improving the smoothness and comparability of the impact index.

[0027] After the above quantile positioning, each disaster event obtains a standardized quantile value between 0 and 1 in each characteristic dimension. This value is the standardized impact index component of the event in the corresponding dimension. The closer the component value is to 1, the higher the intensity percentile of the event in the historical distribution of the same period, and the stronger the impact on working conditions; the closer the component value is to 0, the lower the intensity of the event in history, and the relatively limited the impact on operations.

[0028] The standardized impact index components of various disaster events across all dimensions are organized according to a unified dimensional framework to form a multidimensional standardized impact index. This multidimensional structure includes four dimensions: a temporal dimension (recording the relative positions of the event's start and end times on the harvestable time axis, used to support subsequent sliding window time alignment calculations), a duration dimension (the quantile components corresponding to the duration of continuous rain or waterlogging events, reflecting the degree of sustained impact of the disaster), a precipitation intensity dimension (the quantile components corresponding to the average precipitation intensity, reflecting the immediate impact of precipitation on field access conditions), and a soil moisture content dimension (including the mean quantile components of moisture content and the quantile components of the proportion of time exceeding the threshold, jointly reflecting the degree of damage to soil carrying capacity). The temporal dimension serves as an index axis and does not participate in subsequent weighted aggregation calculations. The component values ​​of the other three physical dimensions are used as inputs during the sliding window scoring stage to participate in the calculation of the fragmentation index and operational efficiency constraint coefficient. This organization of the multidimensional standardized impact index preserves the independent physical meaning of each dimension while providing a unified dimensional basis for subsequent multidimensional weighted aggregation, ensuring the horizontal comparability of impact intensity between different types of disaster events.

[0029] The quantile values ​​of the actual characteristic values ​​of waterlogging events are determined by linear interpolation in the mean quantile mapping table of water content and the quantile mapping table of the proportion of time exceeding the threshold, including: For the hourly soil moisture content change sequence of a waterlogging event, the hourly soil moisture content values ​​of each time point in the sequence are read sequentially. All hourly soil moisture content values ​​are summed one by one and divided by the total number of time points in the sequence to obtain the arithmetic mean of the hourly soil moisture content of the waterlogging event. The arithmetic mean is then linearly interpolated between two adjacent quantile nodes in the hourly soil moisture content mean quantile mapping table to obtain the mean quantile value. The relationship between the hourly soil moisture content value and the preset waterlogging soil moisture content threshold is scanned hourly within the start and end time of the waterlogging event. The number of times the hourly soil moisture content exceeds the threshold is counted to obtain the cumulative number of times the threshold is exceeded. The cumulative number of times the threshold is exceeded is divided by the total number of times the waterlogging event lasts from the start to the end to obtain the percentage of time exceeding the threshold. The percentage is then calculated by linear interpolation between two adjacent quantile nodes in the percentage of time exceeding the threshold quantile mapping table to obtain the quantile value of the duration percentage. The mean quantile and the duration quantile are weighted and summed according to the preset mean contribution weight and duration contribution weight. The weighted summation result is used as the standardized impact index component of the waterlogging event in the soil moisture dimension.

[0030] The impact of waterlogging on winter wheat harvesting is not only reflected in the intensity of rainfall, but also more profoundly in the accumulation and persistence of soil moisture. Excessive soil moisture content leads to a decrease in the bearing capacity of the field surface, increased resistance to harvesting machinery, and even machine stagnation. Therefore, it is necessary to quantify waterlogging events in detail from two dimensions: the average level of soil moisture content and the degree of persistence above the threshold, and ultimately integrate them into a unified standardized impact index component.

[0031] When calculating the mean value of the hourly soil moisture content change sequence of a waterlogging event, the hourly soil moisture content values ​​are read sequentially from the start to the end of the event. Let the total number of waterlogging events be... , No. The hourly soil moisture content values ​​for each time period are as follows: Then the hourly soil moisture content arithmetic mean of the waterlogging event. Calculate as follows:

[0032] During the traversal, the data read at each time step... Add them sequentially to the cumulative amount to complete the process. After each iteration, divide the cumulative amount by . That is, to obtain the hourly arithmetic mean of soil moisture content for this waterlogging event. This arithmetic mean reflects the overall soil moisture saturation during the duration of a waterlogging event; a higher value indicates a more severe average soil moisture condition throughout the event.

[0033] get Then, locate the mean quantile mapping table of hourly soil moisture content. The interval between two adjacent quantile nodes. The mapping table is constructed from historical samples of average water content over many years. Each record in the table contains a sample value of the average water content and its corresponding cumulative empirical quantile level. Let... The mean water content of the lower boundary node that falls into is The corresponding quantile level is The mean water content at the upper bound node is The corresponding quantile level is Then for Perform linear interpolation to obtain the mean quantile values:

[0034] when When the sample value is exactly equal to the sample value of a node in the mapping table, the quantile level corresponding to that node is directly taken as the quantile level. No interpolation is needed. When When the maximum node in the mapping table is exceeded, Take 1.0; when When the value is below the minimum node in the mapping table, Take 0.0.

[0035] The calculation of the percentage of time exceeding the threshold is independent of the mean calculation process, and is also based on the hourly soil moisture content change sequence of the waterlogging event. The soil moisture content values ​​at each time point in this sequence are scanned hourly. Compare it with the preset waterlogging soil moisture content threshold. Perform a comparison one by one. When When a time exceeds the threshold, the time counter is incremented by 1. At that time, this instance is not included in the threshold exceeding statistics. Complete all After a number of scans, let the cumulative number of times the threshold is exceeded be... The percentage of time exceeding the threshold Defined as:

[0036] The range of values ​​is The closer the value is to 1, the more likely the soil moisture content is in a moist state that is above the threshold for a long period of time during the waterlogging event, and the higher the proportion of time that restricts mechanical operations; the closer the value is to 0, the more likely the state of exceeding the threshold occurs only for a short period of time, and the overall impact is relatively limited.

[0037] Will Linear interpolation is performed by substituting the values ​​into the quantile mapping table for the duration exceeding the threshold. This mapping table is also constructed based on historical data of the percentage of flooding events exceeding the threshold over multiple years. Each node in the table records the percentage sample value and its corresponding cumulative empirical quantile level. Let... The percentage of lower bound nodes that fall into the boundary is: The corresponding quantile level is The proportion of the upper bound node is The corresponding quantile level is Then the quantile value of the duration percentage The following linear interpolation is used to obtain:

[0038] Boundary handling rules are consistent with mean quantile interpolation: when The value is 1.0 when it exceeds the maximum node in the mapping table, and 0.0 when it is below the minimum node.

[0039] After obtaining the mean quantile values ​​respectively Percentile values ​​of duration Then, the contribution weights are calculated according to the preset average. Weighting based on duration By performing a weighted summation on the two, the standardized impact index component of the waterlogging event in the soil moisture dimension is obtained. :

[0040] in Both weights are non-negative real numbers, and their specific values ​​are determined based on regional agronomic experience and historical data verification results. In areas with heavy clay texture and poor drainage, the sustained high soil moisture content has a more significant impact on mechanical movement; in such cases, the weights can be appropriately increased. The proportion of water content; however, in sandy loam soils with better water conductivity, the average water content level is more indicative of surface bearing capacity, and can be appropriately increased. The weight.

[0041] The value range also falls within Within the interval, the standardized shock index components of other dimensions maintain the same dimensions, thereby supporting the unified fusion calculation of subsequent multidimensional shock indices. The closer it is to 1, the higher the intensity of the waterlogging event is compared to the historical data for the same period, and the stronger the constraint on machine harvesting operations within the candidate operation window. The closer the value is to 0, the lower the soil moisture content of the waterlogging event is compared to historical data, indicating a relatively minor impact on mechanized harvesting operations. This two-dimensional quantile standardization method can simultaneously capture the intensity of the waterlogging event's impact from two different perspectives: average moisture content and duration of high moisture content. This avoids overlooking the substantial interference of short-term extreme high moisture content events or long-term low-amplitude waterlogging events on the operational window, which is common in single-dimensional assessments.

[0042] The effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient of each window are calculated and weighted to form a comprehensive score for window structure impact, including: The window time range is defined by the total number of hours of the candidate operation window. For each hour in the window, the maximum value of each dimension component of the multidimensional standardized impact index is taken as the representative impact index of that hour. Each representative impact index is compared with the preset impact judgment threshold one by one. Hours with representative impact indices lower than the threshold are marked as workable hours. The total number of hours of all workable hours is accumulated to obtain the effective operation time of the window. Perform a continuous segment scan operation on all available workable time slots arranged in chronological order within the window. Identify the available continuous workable time slots composed of consecutive available workable time slots from the scan results, and take the ratio of the longest duration of the longest duration to the total duration of the window as the work continuity coefficient. The total number of consecutive rainy and waterlogging events within the statistical window time range is used to divide the total number of disaster events by the total window time to obtain the fragmentation index; The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function; The window's effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient are each normalized using the min-max method. The four normalized indicators are then weighted and summed according to a preset indicator weight vector to obtain the comprehensive score of the window structure impact.

[0043] After generating the multidimensional standardized impact indices for each disaster event, a structural assessment is needed for each candidate operational window, which is divided by sliding partitioning, to quantify the comprehensive impact level of each window in terms of temporal distribution, continuity, fragmentation, and soil carrying capacity. The candidate operational windows are divided along the harvestable period time axis with a preset operational window duration as the step size. Each window covers several consecutive hourly periods, and the total window duration is denoted as [missing information]. This refers to the total number of hourly observation records contained in the window from the start time to the end time.

[0044] For each hourly occurrence within the window, the maximum value of each dimension component in the multidimensional standardized impact index corresponding to that hourly occurrence is taken as the representative impact index for that hourly occurrence. Specifically, continuous rain events form standardized components in the dimensions of duration and average precipitation intensity, while waterlogging events form standardized components in the dimensions of mean soil moisture content and the proportion of time exceeding the threshold. If an hourly occurrence falls within the coverage period of a certain disaster event, then that hourly occurrence inherits the impact index components of that disaster event in each dimension. If the hourly occurrence is covered by multiple disaster events, then the maximum value is taken after summing all dimension components of all covered events. If an hourly occurrence does not fall within the coverage period of any disaster event, its representative impact index is set to 0. The representative impact index of each hourly occurrence is then compared with a preset impact judgment threshold. A point-by-point comparison shows that the impact index is significantly lower than... Each available time slot is marked as a workable time slot. The total time of all workable time slots within the window is then added together to obtain the effective work duration of the window. . It reflects the total amount of available time within the candidate window period where both land conditions and weather conditions meet the requirements for machine harvesting operations. The larger the value, the more abundant the time supply for machine harvesting operations within the window.

[0045] The calculation of the job continuity coefficient relies on the scanning analysis of the spatial distribution structure of available job times. A continuous segment scanning operation is performed on all available job times arranged chronologically within the window: starting from the first job time and proceeding sequentially, any two adjacent job times that are both available and temporally consecutive (i.e., their job numbers differ by 1) are grouped into the same continuous available job time period; otherwise, the period is cut off and a new continuous segment is created. After the scan is complete, the segment with the longest duration is selected from all identified continuous available job time periods, and its duration is recorded as [missing value]. ,Will Total window hours The ratio is defined as the job continuity coefficient. ,Right now . The range of values ​​is , The larger the value, the longer the uninterrupted working period within the window, allowing for concentrated and continuous machine harvesting operations and high scheduling efficiency; The smaller the value, the more frequently the work can be interrupted by disaster events, making it difficult to form an effective continuous work rhythm.

[0046] Fragmentation Index This describes the density of disaster events within a given time window, reflecting the frequency with which the work window is disrupted by disasters. It also counts the total number of consecutive rainy days and waterlogging events occurring within the time window, denoted as [missing information]. ,Will Divide by the total number of hours in the window The fragmentation index is obtained, i.e. The number of disaster events here is based on the number of independent events after event segmentation. The same event, regardless of its duration, is counted as 1 time and is not counted repeatedly. The larger the window, the more frequent the disasters within it, the more fragmented the work window is, and the worse the continuity of the machine harvesting operation in terms of time scheduling; A value close to 0 indicates that there are few disaster disturbances within the window and the operation time structure is relatively complete.

[0047] Operational efficiency constraint coefficient This study characterizes the overall constraint of soil moisture status within a window on the efficiency of mechanized harvesting. The standardized impact index components of soil moisture content at each time interval within the window are taken, and their arithmetic mean is calculated to obtain the window-level mean soil moisture impact. For hourly data points not covered by any waterlogging event, the standardized impact index component of the soil moisture content dimension is set to 0 for averaging. Substitute into the preset conversion function By performing mapping, the operation efficiency constraint coefficients are obtained. Conversion function It is usually designed as a monotonically increasing function, when When the soil is relatively small (low overall soil moisture) Taking a smaller value indicates that the soil has a weaker constraint on the efficiency of machine harvesting; when When larger A larger value indicates that soil wetting impact significantly constrains the efficiency of mechanized harvesting, potentially leading to problems such as machinery getting stuck, reduced operating speed, or increased harvest loss. The specific form of the conversion function can be determined based on actual agricultural machinery operation test data. Common forms include linear functions or piecewise linear functions to ensure the physical validity of the mapping results.

[0048] Obtain the effective working time of the window. Operation continuity coefficient Fragmentation Index and work efficiency constraint coefficient After establishing the four structural indicators, they need to be standardized to a comparable scale before weighted aggregation is performed. This is done separately for each of the four indicators. Normalization: For and For these two indicators, larger values ​​are generally more favorable; after normalization, larger values ​​indicate better window conditions. Therefore, a positive score is used in the subsequent weighted summation. and The larger these two indicators are, the more detrimental they are. Their normalized values ​​are directly included in the weighted summation, and the larger the values ​​are, the greater their negative contribution to the overall score. Normalization uses the minimum and maximum values ​​of all candidate windows on the same metric as the endpoints of the normalization interval, linearly mapping the original metric values ​​of each window to... Interval. After normalization, according to the preset index weight vector. A weighted summation of the four normalized indicators yields the comprehensive score for window structure impact. ,in The negative contribution weight corresponding to the normalized effective job duration. The negative contribution weight corresponding to the normalized continuity coefficient, The positive contribution weight corresponding to the normalized fragmentation index, The positive contribution weights corresponding to the normalized operational efficiency constraint coefficients, the sum of the four weights is 1. Overall Score The lower the value, the better the candidate window performs overall in terms of sufficient effective working time, strong operational continuity, minimal disaster disturbance, and weak soil constraints, making it more suitable as a recommended machine harvesting window. Weight Vector Different configurations can be made based on the agronomic requirements, equipment characteristics, or historical scheduling experience of different production areas to adapt to the subjective judgment of the importance of each structural indicator under different operating scenarios.

[0049] The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function, including: Establish a continuous mapping relationship between the mean value of the impact index component of the soil moisture content dimension and the soil bearing capacity attenuation coefficient. Map the lower limit value of the impact index component corresponding to the dry soil state to the full-capacity coefficient that characterizes the agricultural machinery to maintain rated operating efficiency, and map the upper limit value of the impact index component corresponding to the fully saturated soil state to the minimum allowable operating capacity coefficient that characterizes the agricultural machinery to maintain only the minimum operating efficiency. All intermediate state nodes between the lower limit value and the upper limit value are connected in sequence by a monotonically decreasing smooth continuous curve to form a mapping relationship curve. The arithmetic mean of the standardized impact index components of soil moisture content at each time step in the window is calculated to obtain the mean value of the impact index components. The mean value of the impact index components is then substituted into the above continuous mapping relationship. The specific value of the soil bearing capacity attenuation coefficient corresponding to the mean value of the impact index components is determined by linear interpolation between the two adjacent mapping nodes that are closest to the mean value of the impact index components. The equivalent number of working hours is obtained by multiplying the determined soil bearing capacity attenuation coefficient by the total number of working hours in the window. The difference between the total number of working hours in the window and the equivalent number of working hours is calculated, and the ratio obtained by dividing the difference by the total number of working hours in the window is used as the work efficiency constraint coefficient.

[0050] The core of calculating the operational efficiency constraint coefficient lies in quantifying the inhibitory effect of soil moisture content on agricultural machinery operation capacity into a dimensionless index that can be used in comprehensive scoring. The decrease in soil bearing capacity with increasing moisture content is a common physical law in agricultural machinery operations: when the soil is dry or has suitable moisture content, agricultural machinery can maintain its rated operating speed and rated harvesting efficiency; when the soil moisture content is close to or exceeds the field capacity, the soil's support force on the machinery's tires or tracks significantly decreases, leading to machinery sinking, slipping, increased resistance, and a substantial reduction in actual operational efficiency, sometimes even preventing it from operating in the field. Therefore, it is necessary to establish a continuous mapping curve from the standardized impact index component of soil moisture content to the soil bearing capacity attenuation coefficient, in order to transform discrete, hourly impact index information into a quantitative constraint on the overall operational efficiency of the window.

[0051] When establishing this continuous mapping relationship, the impact index component standardized by soil moisture content is used as the horizontal axis input variable, and the soil bearing capacity attenuation coefficient is used as the vertical axis output variable. After quantile standardization, the impact index component is constrained to a range of 0 to 1: the lower bound of 0 corresponds to dry soil, indicating extremely low soil moisture content and no inhibition on the bearing capacity of agricultural machinery. In this case, the soil bearing capacity attenuation coefficient is mapped to the full-capacity coefficient, meaning that the agricultural machinery can maintain its rated operating efficiency, and this coefficient is 1. The upper bound of 1 corresponds to fully saturated soil, indicating extremely high soil moisture content and maximum constraint on agricultural machinery operation. In this case, the soil bearing capacity attenuation coefficient is mapped to the minimum allowable operating capacity coefficient, which is greater than 0 but much less than 1. The specific value can be calibrated based on field test data of different models. In a typical case, it can be set to 0.3, indicating that the agricultural machinery can only maintain about 30% of its rated operating efficiency. Between the lower and upper bounds, all intermediate state nodes are sequentially connected by a monotonically decreasing, smooth, continuous curve. This curve reflects the physical process of the gradual decline in agricultural machinery operating capacity as the impact of soil moisture content increases. The monotonically decreasing characteristic of the curve ensures that the larger the impact index component, the smaller the bearing capacity attenuation coefficient, which conforms to the physical laws of soil. The smooth, continuous characteristic ensures that a unique attenuation coefficient can be obtained by interpolation at any impact index component value, avoiding discontinuities in scoring caused by node jumps.

[0052] In its implementation, this mapping relationship is stored as a discrete node table. Nodes can be calibrated based on actual field test data or soil bearing capacity attenuation patterns in literature. For example, nodes can be set at the impact index components of 0, 0.2, 0.4, 0.6, 0.8, and 1.0, corresponding to soil bearing capacity attenuation coefficients of 1.00, 0.92, 0.80, 0.65, 0.48, and 0.30, respectively. The nodes are connected by linear interpolation, thus forming a piecewise linear, approximately smooth curve over the entire interval. For a smoother curve, cubic spline interpolation can be used to fit the nodes, ensuring the curve has a continuous first derivative at each node, more accurately reflecting the gradual change in soil bearing capacity with water content. The number of nodes and the coefficient values ​​at the nodes are adjustable parameters, allowing for localized calibration based on soil type, agricultural machinery model, and historical operational data in different production areas, thereby improving the adaptability of the evaluation results to specific operational scenarios.

[0053] After obtaining the mapping curve, perform an arithmetic mean operation on the standardized impact index components of soil moisture content for all time periods within the current candidate job window to obtain the mean of the impact index components. This mean comprehensively reflects the overall level of soil moisture content shocks within the window: if soil moisture content is low for most of the time intervals within the window, then... The value is too small, close to the lower limit; if waterlogging or excessively wet soil occurs frequently within the window, then... Larger, close to the upper limit. Substituting the above mapping relationship, in relation to Linear interpolation is performed between the two adjacent mapping nodes with the closest values. Let the values ​​of the impact exponential components of the two adjacent nodes be respectively... and (satisfy The corresponding soil bearing capacity attenuation coefficients are as follows: and The soil bearing capacity attenuation coefficient is then determined by linear interpolation. for:

[0054] This interpolation process guarantees that, in any... Each value yields a unique attenuation coefficient, and the computational complexity is extremely low, making it suitable for quickly executing on a large number of candidate windows one by one during window sliding traversal.

[0055] Obtain the soil bearing capacity attenuation coefficient Then multiply it by the total number of window hours. The converted workable hours are obtained. ,Right now The meaning of converted working hours is: after considering the inhibitory effect of excessive soil moisture on agricultural machinery efficiency, the actual effective working hours that can be utilized within the window. The difference between the total working hours in the calculation window and the converted working hours is calculated. This difference represents the equivalent number of hours of efficiency loss due to excessive soil moisture content. Divide this difference by the total window hours. The work efficiency constraint coefficient is obtained. ,Right now:

[0056] Operational efficiency constraint coefficient The value ranges from 0 to 1: when the soil is dry, Approaching 1, A value approaching 0 indicates that the soil exerts almost no constraint on operational efficiency; when the soil is fully saturated... Take the lowest permissible work capacity coefficient (e.g., 0.3). Taking the maximum value (e.g., 0.7) indicates that 70% of the hours within the window are severely impaired due to excessively wet soil. The larger the value, the stronger the constraint of soil moisture content on the window, and the lower the actual output capacity of machine harvesting. Therefore, a higher penalty weight should be given in the comprehensive score of window structure impact.

[0057] In Before being incorporated into a multidimensional structural indicator system and participating in weighted aggregation, the indicators need to be normalized to ensure they are on the same dimensional scale as other indicators such as effective window operation time, operation continuity coefficient, and fragmentation index. The normalization method is consistent with other indicators, employing range normalization or quantile normalization based on all candidate windows from the same historical period. The actual values ​​are mapped to the range of 0 to 1, and then the preset index weight vector is used. The corresponding weight components Participate in the overall scoring The calculation of the efficiency constraint coefficient ensures that its contribution to the overall score matches its importance in actual operations, avoiding the overemphasis or neglect of a particular indicator in the aggregated results due to differences in dimensionality.

[0058] Through the above complete process, the physical inhibition mechanism of soil moisture content on the efficiency of agricultural machinery operation is embedded into the window evaluation system in a standardized numerical form. This ensures that the selection of recommended harvesting windows not only considers the feasibility of operation in terms of time, but also fully considers the constraints of soil conditions on actual operation output, thereby improving the actual executability of the scheduling plan and the level of operation efficiency guarantee.

[0059] The machine harvesting operation scheduling plan is generated and output by combining the recommended window's time parameters, available agricultural machinery inventory, and the workable area within the window, including: Obtain the total number of registered and callable winter wheat harvesting machines within the spatial range and the rated hourly operating area of ​​each machine. Multiply the rated hourly operating area of ​​a single machine by the number of hours that can be worked within the recommended harvesting window to obtain the theoretical operating area of ​​a single machine within the recommended window period. Obtain the total planting area of ​​winter wheat to be harvested within the spatial range, compare the total planting area with the theoretically operable area of ​​a single agricultural machine, and divide into two processing scenarios. If the total planting area is not greater than the theoretically operable area of ​​a single agricultural machine, the entire planting area is allocated to one agricultural machine to complete the entire harvesting task independently. If the total planting area is greater than the theoretically operable area of ​​a single agricultural machine, the total planting area is evenly distributed according to the total number of agricultural machines, and the geographical range of the corresponding working field is marked for each agricultural machine. The machine departure time is obtained by shifting forward the preset time required for the agricultural machinery to assemble, based on the start time of the recommended harvesting window. The start and end times of each agricultural machine are allocated and calculated one by one according to the departure time of the agricultural machine plus the number of working hours in the window. The output is a complete scheduling scheme with the agricultural machine number as the primary key and including the geographical range of the corresponding field, the start and end times of the operation, and the departure time of the agricultural machine.

[0060] After determining the recommended harvesting window, the time parameters of the window need to be combined with the actual agricultural machinery resources and the information on the plots to be harvested to generate a directly executable harvesting operation scheduling plan. The generation of the scheduling plan uses the start time of the recommended window and the number of working hours within the window as the core time parameters, the number of registered and callable agricultural machines and the rated hourly working area of ​​each agricultural machine as resource parameters, and the total area of ​​winter wheat to be harvested as the task scale parameter. These three types of parameters jointly drive the subsequent area allocation and time calibration calculations.

[0061] Obtain the total number of registered and available agricultural machines for winter wheat harvesting within this spatial range. and the The rated hourly operating area of ​​a single agricultural machine in Taiwan (Unit: mu / hour), of which The range of values ​​is to The recommended working hours within the machine harvesting window are recorded as follows: (Unit: hours). This value is directly derived from the calculation result of the effective working time within the window, that is, the cumulative number of consecutive working times when the comprehensive impact index is lower than the preset judgment threshold within the window. The theoretical workable area of ​​a single agricultural machine within the recommended window period. Calculate using the following formula:

[0062] This formula directly multiplies the rated operating efficiency of a single machine by the number of available operating hours to obtain the maximum theoretical area that a single agricultural machine can complete under the recommended window constraint. It should be noted that... This is an upper limit estimate under ideal conditions that do not consider losses during field relocation. In actual scheduling, the upper limit can be adjusted according to the degree of dispersion of the plots. A transition reduction factor is introduced for correction, but in the basic scheduling scheme generation stage, As a benchmark value for area allocation.

[0063] Obtain the total area of ​​winter wheat planted to be harvested within this spatial range. (Unit: mu). [The rest of the text appears to be incomplete and requires further context.] The theoretical workable area of ​​a single agricultural machine Numerical comparisons are performed to distinguish between the two processing scenarios. When the agricultural machinery models are the same or the parameters of the first agricultural machinery are used as representative, if... In this case, the entire planting area does not require multiple agricultural machines to work together; a single agricultural machine can independently complete the entire harvesting task. The workload of this single agricultural machine is [missing information]. The geographical area of ​​the corresponding work field covers all plots of land to be harvested. This situation usually occurs in scenarios with a small space, limited planting scale, or sufficient recommended working hours. A single agricultural machine can complete all the work within the working hours without the need for additional multi-machine coordination costs.

[0064] when At times, the theoretical capacity of a single agricultural machine is insufficient to harvest the entire planted area within the recommended timeframe, necessitating the use of all available machinery. Taiwanese agricultural machinery operates in a coordinated manner. The total planting area is allocated equally according to the total number of agricultural machines. Taiwan agricultural machinery allocated operating area Determine by the following formula:

[0065] The premise of equal distribution is that the rated hourly working area of ​​each agricultural machine is similar or the same. If the area of ​​each agricultural machine... If differences exist, the allocation should be weighted according to the proportion of each agricultural machine's rated operating efficiency to the total efficiency. Weighted allocation area of ​​agricultural machinery in Taiwan for:

[0066] The allocation method fully utilizes the efficiency differences between different machine models, avoiding situations where high-efficiency machines are idle and low-efficiency machines are overloaded. After calculating the area allocation, based on the assigned operating area for each machine and the spatial distribution information of the plots, the geographical range of the corresponding operating field for each machine is determined. Plot division uses administrative villages or plot numbers as the smallest unit, prioritizing the allocation of spatially adjacent and contiguous plots to the same machine to reduce the number of field moves and improve actual operating efficiency. The geographical range of the operating field corresponding to each machine is recorded in the form of a list of plot numbers or a polygon with geographical coordinate boundaries, serving as the spatial attribute field output of the scheduling plan.

[0067] In the time calibration stage, the start time of the recommended receiver window is used. Based on the baseline, shift forward by the preset time required for agricultural machinery assembly. (Unit: hours) Obtain the departure time of the agricultural machinery. :

[0068] The value is preset based on the average transfer distance and travel speed from the agricultural machinery depot to the work site, usually between 0.5 hours and 2 hours. The specific value is pre-configured by the regional agricultural machinery management department according to the actual road conditions and type of agricultural machinery. Agricultural machinery departure time. This is the moment when all agricultural machinery participating in this dispatch will uniformly execute the departure command, and each piece of agricultural machinery will... They depart from their respective parking locations at all times and head towards their assigned work plots.

[0069] The start time for all agricultural machines is uniformly set to the start time of the recommended window. Operation termination time The calculations are based on the allocated operating area for each agricultural machine and the rated hourly operating area per machine:

[0070] When equal distribution is used When using weighted allocation, Take the corresponding weighted value. It must not exceed the end time of the recommendation window. If the calculated result This indicates the current number of agricultural machines. If the recommended window is insufficient to complete all tasks, an early warning will be triggered. It is recommended to add more agricultural machinery or postpone the excess tasks to the suboptimal window for processing.

[0071] The final complete scheduling scheme is based on the agricultural machinery number as the primary key. Each record contains the following fields: agricultural machinery number, geographical range of the corresponding work field (list of field numbers or boundary coordinates), and start time of the operation. , Operation termination time Agricultural machinery departure time and the allocated work area The scheduling plan is output in a structured data format, supporting import into the agricultural machinery scheduling management platform or being pushed to the agricultural machinery driver's terminal in report form, realizing a closed loop of the entire process from disaster process analysis to the issuance of work instructions. For situations where multiple agricultural machines are operating in parallel, the scheduling plan also includes suggestions for the transfer routes of each agricultural machine to different plots, using the shortest transfer route as the optimization objective and automatically generating the transfer sequence based on the spatial distribution of the plots, further improving the actual work completion rate within the window period.

[0072] A second aspect of this invention provides a system for extracting and evaluating fragmented features of disaster process operation windows, comprising: The event detection unit is used to acquire meteorological time series data of a specified spatial range during the winter wheat harvest period, detect and identify continuous rain events and waterlogging events based on the continuous rain judgment rules and waterlogging judgment rules, perform event segmentation on each disaster event, and extract feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence. The shock standard unit is used to perform quantile standardization mapping on the feature parameters of each disaster event in different dimensions based on the historical quantile benchmark dataset of the same period in this spatial range, transforming feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster shock index. The window scoring unit is used to slide along the harvest period time axis to divide candidate operation windows with a preset operation window duration as the step size. Within each candidate window, the multidimensional standardized impact index is used as input to calculate the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient of the window window one by one. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The scheduling generation unit sorts the structural impact comprehensive scores of each window from low to high, selects the candidate window with the lowest score as the recommended machine harvesting window, and generates and outputs a machine harvesting operation scheduling plan by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

[0073] A third aspect of the present invention provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0074] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0075] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for extracting and evaluating fragmented features of disaster process operation windows, characterized in that, include: Meteorological time-series data of a specified spatial range during the winter wheat harvest period are obtained. Based on the continuous rain and waterlogging judgment rules, continuous rain and waterlogging events are detected and identified. Event-based segmentation is performed on each disaster event, and feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence are extracted. Based on the historical quantile benchmark dataset of the same period in this spatial range, quantile standardization mapping is performed on the feature parameters of each disaster event in different dimensions, transforming the feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster impact index. Candidate operation windows are divided along the time axis of the harvest period with a preset operation window duration as the step size. Within each candidate window, the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient are calculated for each window using a multidimensional standardized impact index as input. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The structural impact scores of each window are sorted from low to high. The candidate window with the lowest score is selected as the recommended machine harvesting window. The machine harvesting operation scheduling plan is generated and output by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

2. The method according to claim 1, characterized in that, Quantile normalization mapping was performed on the feature parameters of each disaster event across different dimensions, including: Extract from the historical quantile benchmark dataset of the same period in the spatial range the sample sets of duration of continuous rain events, average precipitation intensity, duration of waterlogging events, and hourly soil moisture content change sequence of waterlogging events for each historical year within the same period as the current harvest period. After sorting the duration sample set in ascending order, the cumulative empirical quantile level of each sorting position is calculated based on the total number of samples. A duration quantile mapping table is constructed with the quantile of the sorting position as the vertical axis and the corresponding duration sample value as the horizontal axis. A mean precipitation intensity quantile mapping table is constructed in the same way. The arithmetic mean of each sequence in the hourly soil moisture content change sequence sample set is calculated to obtain the moisture content mean sample set. The moisture content mean quantile mapping table is constructed in the same sorting and quantile method. At the same time, the proportion of times in each sequence that exceed the preset waterlogging threshold is counted to construct the threshold duration quantile mapping table. The actual characteristic values ​​of each disaster event are determined by linear interpolation between adjacent quantile nodes in the corresponding mapping table, and each quantile value is used as the standardized impact index component of the event in the corresponding dimension. All components are organized into a multidimensional standardized impact index according to the time series dimension, duration dimension, precipitation intensity dimension, and soil moisture content dimension.

3. The method according to claim 2, characterized in that, The quantile values ​​of the actual characteristic values ​​of waterlogging events are determined by linear interpolation in the mean quantile mapping table of water content and the quantile mapping table of the proportion of time exceeding the threshold, including: For the hourly soil moisture content change sequence of a waterlogging event, the hourly soil moisture content values ​​of each time point in the sequence are read sequentially. All hourly soil moisture content values ​​are summed one by one and divided by the total number of time points in the sequence to obtain the arithmetic mean of the hourly soil moisture content of the waterlogging event. The arithmetic mean is then linearly interpolated between two adjacent quantile nodes in the hourly soil moisture content mean quantile mapping table to obtain the mean quantile value. The relationship between the hourly soil moisture content value and the preset waterlogging soil moisture content threshold is scanned hourly within the start and end time of the waterlogging event. The number of times the hourly soil moisture content exceeds the threshold is counted to obtain the cumulative number of times the threshold is exceeded. The cumulative number of times the threshold is exceeded is divided by the total number of times the waterlogging event lasts from the start to the end to obtain the percentage of time exceeding the threshold. The percentage is then calculated by linear interpolation between two adjacent quantile nodes in the percentage of time exceeding the threshold quantile mapping table to obtain the quantile value of the duration percentage. The mean quantile and the duration quantile are weighted and summed according to the preset mean contribution weight and duration contribution weight. The weighted summation result is used as the standardized impact index component of the waterlogging event in the soil moisture dimension.

4. The method according to claim 1, characterized in that, The effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient of each window are calculated and weighted to form a comprehensive score of window structure impact, including: The window time range is defined by the total number of hours of the candidate operation window. For each hour in the window, the maximum value of each dimension component of the multidimensional standardized impact index is taken as the representative impact index of that hour. Each representative impact index is compared with the preset impact judgment threshold one by one. Hours with representative impact indices lower than the threshold are marked as workable hours. The total number of hours of all workable hours is accumulated to obtain the effective operation time of the window. Perform a continuous segment scan operation on all available workable time slots arranged in chronological order within the window. Identify the continuous workable time slots composed of consecutive workable time slots from the scan results, and take the ratio of the longest duration of the longest duration to the total duration of the window as the work continuity coefficient. The fragmentation index is obtained by dividing the total number of consecutive rainy and waterlogging events within the statistical window time range by the total number of disaster events. The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function; The window's effective operation time, operation continuity coefficient, fragmentation index, and operation efficiency constraint coefficient are each normalized using the min-max method. The four normalized indicators are then weighted and summed according to a preset indicator weight vector to obtain the comprehensive score of the window structure impact.

5. The method according to claim 4, characterized in that, The arithmetic mean of the standardized impact index components of each hourly soil moisture content dimension in the window is mapped to the operation efficiency constraint coefficient through a preset conversion function, including: Establish a continuous mapping relationship between the mean value of the impact index component of the soil moisture content dimension and the soil bearing capacity attenuation coefficient. Map the lower limit value of the impact index component corresponding to the dry soil state to the full-capacity coefficient that characterizes the agricultural machinery to maintain rated operating efficiency, and map the upper limit value of the impact index component corresponding to the fully saturated soil state to the minimum allowable operating capacity coefficient that characterizes the agricultural machinery to maintain only the minimum operating efficiency. All intermediate state nodes between the lower limit value and the upper limit value are connected in sequence by a monotonically decreasing smooth continuous curve to form a mapping relationship curve. The arithmetic mean of the standardized impact index components of soil moisture content at each time step in the window is calculated to obtain the mean value of the impact index components. The mean value of the impact index components is then substituted into the above continuous mapping relationship. The specific value of the soil bearing capacity attenuation coefficient corresponding to the mean value of the impact index components is determined by linear interpolation between the two adjacent mapping nodes that are closest to the mean value of the impact index components. The equivalent number of working hours is obtained by multiplying the determined soil bearing capacity attenuation coefficient by the total number of working hours in the window. The difference between the total number of working hours in the window and the equivalent number of working hours is calculated, and the ratio obtained by dividing the difference by the total number of working hours in the window is used as the work efficiency constraint coefficient.

6. The method according to claim 1, characterized in that, The machine harvesting operation scheduling plan is generated and output by combining the recommended window's time parameters, available agricultural machinery inventory, and the workable area within the window, including: Obtain the total number of registered and callable winter wheat harvesting machines within the spatial range and the rated hourly operating area of ​​each machine. Multiply the rated hourly operating area of ​​a single machine by the number of hours that can be worked within the recommended harvesting window to obtain the theoretical operating area of ​​a single machine within the recommended window period. Obtain the total planting area of ​​winter wheat to be harvested within the spatial range, compare the total planting area with the theoretically operable area of ​​a single agricultural machine, and divide into two processing scenarios. If the total planting area is not greater than the theoretically operable area of ​​a single agricultural machine, the entire planting area is allocated to one agricultural machine to complete the entire harvesting task independently. If the total planting area is greater than the theoretically operable area of ​​a single agricultural machine, the total planting area is evenly distributed according to the total number of agricultural machines, and the geographical range of the corresponding working field is marked for each agricultural machine. The machine departure time is obtained by shifting forward the preset time required for the agricultural machinery to assemble, based on the start time of the recommended harvesting window. The start and end times of each agricultural machine are allocated and calculated one by one according to the departure time of the agricultural machine plus the number of working hours in the window. The output is a complete scheduling scheme with the agricultural machine number as the primary key and including the geographical range of the corresponding field, the start and end times of the operation, and the departure time of the agricultural machine.

7. A system for extracting and evaluating fragmented features of disaster process operation windows, used to implement the method as described in any one of claims 1-6, characterized in that, include: The event detection unit is used to acquire meteorological time series data of a specified spatial range during the winter wheat harvest period, detect and identify continuous rain events and waterlogging events based on the continuous rain judgment rules and waterlogging judgment rules, perform event segmentation on each disaster event, and extract feature parameters of each event in the dimensions of start and end time, duration, average precipitation intensity and hourly soil moisture content change sequence. The shock standard unit is used to perform quantile standardization mapping on the feature parameters of each disaster event in different dimensions based on the historical quantile benchmark dataset of the same period in this spatial range, transforming feature parameters of different dimensions and magnitudes to the same quantile scale, and generating a multidimensional standardized disaster shock index. The window scoring unit is used to slide along the harvest period time axis to divide candidate operation windows with a preset operation window duration as the step size. Within each candidate window, the multidimensional standardized impact index is used as input to calculate the effective operation duration, operation continuity coefficient, fragmentation index and operation efficiency constraint coefficient of the window window one by one. A multidimensional structural index system is constructed and weighted and aggregated into a comprehensive score of window structural impact. The scheduling generation unit sorts the structural impact comprehensive scores of each window from low to high, selects the candidate window with the lowest score as the recommended machine harvesting window, and generates and outputs a machine harvesting operation scheduling plan by combining the time parameters of the recommended window, the number of available agricultural machines, and the workable area within the window.

8. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 6.

Images