Method and system for intelligent evaluation of meteorological observation capability

Abstract

The present application relates to the technical field of meteorological service, and discloses a kind of meteorological observation ability intelligent evaluation method and system, comprising the following steps: obtaining meteorological observation task type, scene parameter and standard observation process, the input data of training terminal, operation event, intermediate result and result are collected, and observation evidence chain is constructed according to the observation phase of presequence dependency relationship;Compare stage data with standard stage data, determine the first deviation stage and the stage of associated deviation;According to the result difference and stage association state, form result accuracy index, process reliability index and evidence consistency state;According to evidence integrity and evidence consistency state, correct evaluation weight, merge evaluation evidence and generate ability evaluation result, trigger artificial review or supplementary task when confirmation condition is not satisfied.The present application distinguishes root deviation and associated deviation by associating observation result and process data, so that the evaluation result has process basis.

Description

Technical Field

[0001] This invention relates to the field of meteorological service technology, and in particular to an intelligent assessment method and system for meteorological observation capabilities. Background Technology

[0002] In the field of meteorological service technology, the collection, transmission, identification, calculation, storage, retrieval, and display of meteorological observation data, images, operational records, and assessment results are typically carried out. Meteorological observation equipment, training terminals, and servers collaborate in the order of task configuration, data upload, result comparison, and record archiving. Traditional intelligent assessment methods and systems for meteorological observation capabilities evaluate observers' knowledge, target identification, parameter interpretation, calculation processing, and operational procedures in observations of cloud formation, cloud cover, sea state, waves, wind direction and speed, humidity, air pressure, and sea surface wind fields.

[0003] Traditional meteorological observation capability assessments often rely on answers to questions, parameter results, and completion time. Servers compare and score based on standard answers, numerical tolerances, or fixed steps. Images, operational events, intermediate results, and final results from different observation tasks are often saved separately, lacking a unified expression of pre- and post-processing relationships. Consequently, it is difficult to distinguish between accidental correct answers, process deviations, end-entry errors, and cascading deviations caused by preceding errors. Fixed score summarization can also cause the same source of error to be reflected repeatedly in multiple subsequent stages, making the correspondence between assessment conclusions and the observation process unclear. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent assessment method and system for meteorological observation capabilities. By establishing a capability assessment mechanism covering the entire meteorological observation process, the system performs correlation analysis on scenario condition data, trainee input data, operational events, intermediate results, and final observation results during meteorological observation training. Combining the pre-dependent relationships between observation stages, it constructs data correlation relationships that can reflect the entire observation process. Based on the consistency between the observation process and the observation results, the system comprehensively evaluates capability indicators to form an assessment result that reflects the actual observation capability.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for intelligently assessing meteorological observation capabilities includes the following steps: Obtain the task type, scenario parameters, and corresponding standard observation procedures for meteorological observation tasks. The standard observation procedures include multiple observation stages with pre-dependent relationships.

[0006] The system collects scene condition data, student input data, operation events, intermediate results, and final observation results generated when the training terminal performs meteorological observation tasks, and constructs an observation evidence chain according to the observation stage and prior dependencies.

[0007] Preferably, when constructing the observation evidence chain, a stage record is established for each observation stage according to the standard observation process. The stage record includes a task identifier, stage identifier, preceding stage identifier, data source identifier, input data, output data, time identifier, and validity status. Based on the preceding dependency relationship, a reference relationship is established between the output data of the preceding observation stage and the input data of the subsequent observation stage. When the input data, output data, and time identifier all meet the data requirements of the corresponding observation stage, the stage record is marked as valid. When any data requirement is not met, the stage record is marked as invalid. The valid stage records are then combined into the observation evidence chain according to the execution order.

[0008] As a further limitation, when collecting scenario condition data, student input data, operation events, intermediate results, and final observation results, corresponding data collection rules are established based on the meteorological observation task type; various types of data are recorded according to three collection points: the start of the observation phase, the completion of the phase, and the phase switch; all types of data are assigned a unified task identifier, phase identifier, and time identifier; when any observation phase has missing data or abnormal time sequence, the data of the corresponding observation phase is reacquired; if reacquisition fails, the corresponding observation phase is marked as missing and used as the basis for calculating the completeness of evidence.

[0009] The data from each observation stage in the observation evidence chain are compared sequentially with the corresponding standard stage data to determine the first deviation stage and the associated deviation stages caused by the first deviation stage.

[0010] Furthermore, according to the execution order of the stage records in the observation evidence chain, the input data and output data of each valid stage record are compared with the corresponding standard stage data; when the input data of the stage record meets the corresponding standard stage data but the output data does not meet the corresponding standard stage data, the corresponding observation stage is determined as a candidate deviation stage; the earliest observation stage is selected from the candidate deviation stages that appear in the execution order as the first deviation stage, and subsequent observation stages whose input data directly or indirectly references the output data of the first deviation stage and deviates from the corresponding standard stage data are determined as the associated deviation stages.

[0011] The accuracy index of the results is generated based on the difference between the final observation results and the standard results. The credibility index of the process is generated based on the correlation between the preceding inputs, intermediate results and the final observation results, and the consistency status of the evidence is determined.

[0012] Furthermore, the accuracy index of the results is calculated based on the difference between the final observation results and the standard results; the reliability index of the process is calculated based on the consistency of the input data, output data and the correlation between the stages of each observation stage; the consistency of the result accuracy index and the reliability index of the process is verified, and the deviation is calculated after normalizing the two indicators. When the deviation is not greater than the preset consistency threshold, the evidence consistency status is determined to be consistent; when the deviation is greater than the preset consistency threshold, the evidence consistency status is determined to be inconsistent.

[0013] Based on the evidence completeness of the observation evidence chain and the basic evaluation weight of the evidence consistency status correction capability index, the evaluation evidence is merged and evaluated according to the mapping relationship between the first deviation stage and the capability index to generate a meteorological observation capability assessment result; when the evidence completeness or the evidence consistency status does not meet the confirmation conditions, a manual review or supplementary observation task is triggered.

[0014] Preferably, the evidence completeness level is determined based on the proportion of valid stage records to all stage records in the observation evidence chain; the basic evaluation weights of each capability indicator are adjusted according to the evidence completeness level and evidence consistency status; the evaluation evidence is merged according to the capability indicator mapping relationship corresponding to the first deviation stage to form intermediate evaluation results for each capability indicator; the meteorological observation capability assessment result is generated based on the intermediate evaluation results of each capability indicator, and when the evidence consistency status is inconsistent, the meteorological observation capability assessment result is marked as a result to be reviewed.

[0015] As a further limitation, when generating the meteorological observation capability assessment results, candidate evaluation levels for corresponding capability items are formed based on the intermediate evaluation results of each capability indicator; the candidate evaluation levels are checked for consistency based on the consistency of evidence; when the candidate evaluation levels for each capability item meet the consistency check conditions, the final capability evaluation level is generated; when any candidate evaluation level for a capability item does not meet the consistency check conditions, the intermediate evaluation results for the corresponding capability item are maintained, and the corresponding capability item is marked as pending confirmation; the meteorological observation capability assessment results are generated together based on the final capability evaluation level and the pending confirmation status.

[0016] Furthermore, when the completeness of the evidence is lower than the preset completeness threshold or the consistency of the evidence is inconsistent, a supplementary observation stage is determined; a supplementary observation task is generated based on the supplementary observation stage, and the supplementary observation task is limited to the supplementary observation stage and its subsequent associated deviation stages; the new stage record corresponding to the supplementary observation task is received, the observation evidence chain of the corresponding stage is reconstructed, and a capability assessment is performed; when the regenerated meteorological observation capability assessment result meets the confirmation conditions, the updated meteorological observation capability assessment result is output; otherwise, a manual review instruction is output.

[0017] Furthermore, the mapping relationship between capability indicators and the first deviation stage includes observation preparation capability, equipment operation capability, observation execution capability, data processing capability, and result analysis capability. Each capability indicator corresponds to at least one observation stage. When the first deviation stage corresponds to multiple capability indicators, the main capability indicators are determined based on the degree of influence of the first deviation stage on each capability indicator. Only the main capability indicators are used as the merged evaluation objects for the corresponding related deviation stages to form intermediate evaluation results for each capability indicator. Finally, the meteorological observation capability assessment results are generated based on the intermediate evaluation results of each capability indicator.

[0018] The above method is equipped with a meteorological observation capability intelligent assessment system. The system is used to support the execution of the above method and includes a task structure configuration unit, a multi-source process data access unit, an evidence chain construction unit, a stage deviation tracing unit, an evidence consistency assessment unit, a weight correction unit, and a capability evidence merging unit. The system includes the following components: a task structure configuration unit for storing observation stages, prerequisite dependencies, standard stage data, and capability index mapping relationships corresponding to meteorological observation tasks; a multi-source process data access unit for receiving various process data generated by the training terminal; an evidence chain construction unit for forming an observation evidence chain; a stage deviation tracing unit for determining the first deviation stage and associated deviation stages; an evidence consistency assessment unit for forming an evidence consistency status; a weight correction unit for correcting the basic evaluation weights of capability indicators; and a capability evidence merging unit for merging and evaluating evidence based on capability index mapping relationships and outputting meteorological observation capability assessment results. When the evidence completeness or the evidence consistency status does not meet the confirmation conditions, a review instruction or supplementary task instruction is output to implement the above methods.

[0019] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by organizing scenario condition data, student input data, operational events, intermediate results, and final observation results according to the stage sequence and pre-determined dependencies in the standard observation process, a traceable chain of observation evidence is formed, enabling the final result to correspond to a specific observation stage. By comparing standard stage data stage by stage and identifying the first deviation stage and associated deviation stages, the root cause deviation and subsequent propagation deviation can be distinguished, reducing repeated evaluations of the same source of error. By combining the basic evaluation weights of result accuracy indicators, process credibility indicators, and evidence consistency state correction capability indicators, the evaluation criteria for different tasks are matched with the actual evidence completeness, and manual review or supplementary observation tasks are triggered when the confirmation conditions are not met. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A schematic diagram of the overall processing flow of the meteorological observation capability assessment method; Figure 2 A diagram illustrating the data structure and stage dependencies of the evidence chain; Figure 3 This is a schematic diagram illustrating the tracing relationship between the first deviation stage and subsequent deviation stages. Figure 4 A schematic diagram illustrating the state transition of the completeness and consistency of evidence in the evaluation of competence. Figure 5 This is a schematic diagram of the functional unit interactions of the meteorological observation capability assessment system. Detailed Implementation

[0022] The technical solution described in this invention will be further explained below with reference to the accompanying drawings. The meteorological observation task in this document refers to the task whereby a training terminal provides trainees with the observation object, scene parameters, input conditions, and operational requirements, and requires trainees to complete data reading, parameter input, step execution, intermediate processing, or result submission. Meteorological observation tasks include at least one of the following: cloud formation observation, cloud cover observation, sea state observation, wave observation, humidity observation, air pressure observation, wind direction and speed observation, true wind plotting, and sea surface wind field inversion.

[0023] A standard observation procedure refers to a phased execution benchmark corresponding to the type of meteorological observation task. Each standard observation procedure consists of multiple observation stages, and the observation stages are established with prerequisite dependencies based on data sources, processing order, and result referencing relationships. When data generated in a preceding observation stage is read by a subsequent observation stage, the preceding observation stage is a directly dependent stage of the subsequent observation stage; when a subsequent observation stage reads data generated in an earlier observation stage through one or more intermediate observation stages, the earlier observation stage is an indirectly dependent stage of the subsequent observation stage.

[0024] Standard stage data refers to the data benchmark used to determine the execution status of the corresponding observation stage, including standard input conditions, standard output results, allowable error range, necessary data items, and stage execution order. Standard stage data is not limited to a single numerical value; it can also be category labels, direction ranges, parameter intervals, combinations of operation events, stage-related conditions, or intermediate result statuses.

[0025] A phase record is a structured representation of the execution process of an observation phase. A phase record includes at least a task identifier, phase identifier, preceding phase identifier, data source identifier, input data, output data, time identifier, and validity status. The task identifier is used to group phase records generated from the same meteorological observation task into the same task process; the phase identifier is used to distinguish different observation phases; the preceding phase identifier indicates the preceding observation phase that the current observation phase depends on; the data source identifier distinguishes whether the data was generated from the training terminal, student input operations, the scene generation process, or the preceding observation phase; the time identifier indicates the time of data generation and the start, completion, or switching point of the observation phase; and the validity status indicates whether the phase record meets the data integrity and chronological order requirements of the corresponding observation phase.

[0026] An observational evidence chain is a data structure formed by connecting the records of effective stages in the same meteorological observation task in the order of execution. The connections within the observational evidence chain are not simply a chronological arrangement, but rather based on the identifiers of preceding stages and input-output references, indicating which preceding observation stage's data was used in subsequent observation stages. Through the observational evidence chain, one can trace back from the final observation result to the input data, intermediate results, and operational events that led to the final observation result.

[0027] The results accuracy index describes the degree of conformity between the final observation results and the standard results. The process reliability index describes whether the stages of the preceding inputs, intermediate results, operational events, and final observation results meet the requirements of the standard observation procedure. The evidence consistency status indicates whether the results accuracy index and the process reliability index support each other. The evidence completeness index indicates the completeness of the valid stage records in the observation evidence chain relative to all required stage records. The capability index summarizes the personnel capabilities reflected in different observation stages, including at least observation preparation capability, equipment operation capability, observation execution capability, data processing capability, and results analysis capability.

[0028] Example 1 Please see Figures 1-4 This embodiment provides a method for assessing meteorological observation capabilities, which is executed in conjunction with a training terminal and an assessment processing terminal. The training terminal is responsible for presenting the meteorological observation task and collecting scenario condition data, trainee input data, operational events, intermediate results, and final observation results during the task execution process. The assessment processing terminal is responsible for calling the standard observation procedure and standard stage data, establishing an observation evidence chain, determining the stage deviation status, forming intermediate evaluation results of capability indicators, and outputting the meteorological observation capability assessment result. This embodiment is executed sequentially according to S1 to S5.

[0029] S1, obtain the task type, scene parameters and corresponding standard observation procedures for meteorological observation tasks.

[0030] Before a meteorological observation mission begins, the training terminal generates a mission identifier and reads the mission type recorded in the mission configuration. The mission type is used to determine the observation object and the observation phase to be performed. Scene parameters are formed by the training terminal during mission loading and are associated with the mission identifier. For cloud observation, scene parameters include the scene identifier corresponding to the presented cloud image and the observation conditions; for sea state or wave observation, scene parameters include the virtual sea surface state and the observation perspective; for humidity, air pressure, and wind direction and speed observation, scene parameters include the original observation values ​​given in the task; for true wind plotting, scene parameters include the input wind parameters required for plotting; for sea surface wind field inversion, scene parameters include wind speed, wind direction, incident angle, observation azimuth, polarization mode, noise parameters, or configured data in the observation geometry.

[0031] The evaluation processing end uses task type as the search criterion to read the corresponding standard observation process from the task structure configuration data. The task structure configuration data, indexed by task type, stores the observation phases, phase execution order, prerequisite dependencies, standard phase data, and capability indicator mappings. Each observation phase corresponds to a phase identifier, and each non-initial observation phase stores at least one prerequisite phase identifier.

[0032] For tasks requiring only one input step and one result submission step, a standard observation workflow may include an observation object presentation stage, a parameter or category input stage, and a result submission stage. For tasks with intermediate calculation processes, a standard observation workflow also includes a raw data reading stage, an intermediate calculation stage, and a result verification stage. For tasks with continuous operation processes, a standard observation workflow also includes a scene parameter setting stage, an operation execution stage, an intermediate state generation stage, a result interpretation stage, and a result submission stage.

[0033] After the evaluation processing terminal completes the reading of the standard observation process, it verifies whether all stage identifiers in the standard observation process are unique, and verifies whether non-initial observation stages can be traced back to the initial observation stage through the identifier of the preceding stage. If the above conditions are met, the standard observation process proceeds to the subsequent acquisition process; if the above conditions are not met, the current task does not enter the capability evaluation process, and the evaluation processing terminal outputs the task configuration confirmation status.

[0034] S2 collects data generated when the training terminal performs meteorological observation tasks and constructs an observation evidence chain according to the observation stage and prior dependencies.

[0035] The training terminal records the task process at three data acquisition points: the start of the observation phase, the completion of the observation phase, and the switching of observation phases. At the start of the observation phase, the training terminal records the task identifier, phase identifier, start time, and the input data read in the current observation phase. At the completion of the observation phase, the training terminal records the completion time, the output data generated by the trainee in the current observation phase, and the operation events that occurred during the current observation phase. When switching observation phases, the training terminal marks the output data of the previous observation phase as the input source for the subsequent observation phase and records the identifier of the previous phase.

[0036] Student input data includes the selected category, entered values, set scene parameters, orientation and position of the drawn object, selected observation viewpoint, input fuzzy solution, and set quality control conditions. Operation events include parameter modification, step entry, step exit, result submission, and prompt recall. Intermediate results include humidity calculation data, air pressure correction data, intermediate vectors generated from real wind plotting, and solution space or filtered results generated from sea surface wind field inversion. Final observation results include the final submitted category, values, direction, vector results, fuzzy solution interpretation results, or wind field processing results.

[0037] Data source identifiers are determined based on the data's origin. Data generated during the task loading process is labeled as scenario source; data generated by student input operations is labeled as input source; data generated by calculations or scenario updates performed by the training terminal is labeled as process source; and data transmitted during the preceding observation phase is labeled as phase reference source. The training terminal sends all types of data, along with task identifiers, phase identifiers, and time identifiers, to the evaluation processing terminal.

[0038] The evaluation processing unit establishes a stage record for each observation stage according to the standard observation procedure. First, the evaluation processing unit assigns the received data to the corresponding task based on the task identifier, and then writes the data to the corresponding stage record based on the stage identifier. For input data, the evaluation processing unit checks the data source identifier; when the data source identifier is a stage reference source, it reads the previous stage identifier and establishes a reference relationship between the output data of the corresponding previous observation stage and the input data of the current observation stage.

[0039] The validity status is determined based on data integrity and chronological order. The evaluation processing unit reads the necessary data items recorded in the standard stage data and checks the input data, output data, and time markers in the stage record against the necessary data items one by one. When all necessary input data, necessary output data, and time markers are present, the start time is no later than the completion time, and the start time of the current observation stage is no earlier than the start time of the directly dependent stage, the stage record is marked as valid. When necessary data items are missing, the chronological order is abnormal, or the stage marker does not correspond to the standard observation procedure, the stage record is marked as invalid.

[0040] When missing data or chronological anomalies exist during the observation phase, the evaluation processing unit sends a reacquisition request to the training terminal. The reacquisition request includes the task identifier, phase identifier, and the name of the missing data item. The training terminal reads the corresponding data from the current task process record and retransmits it. If the data requirements are met after reacquisition, the evaluation processing unit updates the phase record and changes the valid status to valid; if the data requirements are still not met after reacquisition, the evaluation processing unit marks the corresponding observation phase as missing. A missing status does not directly indicate a trainee error; it only serves as a basis for subsequent evidence completeness calculations.

[0041] The evaluation process begins with the initial observation phase and connects the valid phase records according to their preceding dependencies. Each connection includes a preceding phase identifier, a subsequent phase identifier, and the referenced data item. When a subsequent observation phase depends on multiple preceding observation phases, corresponding reference relationships are established for each. The valid phase records connected in the execution order constitute the observation evidence chain.

[0042] After the observation evidence chain is constructed, the evaluation processing end performs a chain integrity check. The chain integrity check starts from the observation stage where the final observation result is located and reads the stage records backward along the previous stage identifiers. When the backward read can reach the initial observation stage in the standard observation process, the observation evidence chain is marked as traceable; when the backward read is interrupted at a certain observation stage, the observation stage corresponding to the interruption position is marked as missing, and the identifiers of the two stages before and after the interruption are recorded.

[0043] S3, compare the stage data in the observation evidence chain with the standard stage data in turn to determine the first deviation stage and the associated deviation stages.

[0044] The evaluation processing unit reads each valid stage record according to the execution order of the observation evidence chain, and compares the input and output data in the stage record with the corresponding standard stage data. The comparison method is determined by the data type of the standard stage data.

[0045] For categorical data, the comparison is between the student's selected category and the standard category. If they are the same, the output status is "Stage Compliance"; otherwise, the output status is "Stage Deviation". For numerical data, the comparison is between the student's output value and the standard value or standard value range. If the student's output value is within the allowable error range, the output status is "Stage Compliance"; otherwise, the output status is "Stage Deviation". For directional data, the comparison is between the student's output direction and the standard direction range. If the student's output direction is within the standard direction range, the output status is "Stage Compliance"; otherwise, the output status is "Stage Deviation". For operational events, the comparison is between the necessary operational events actually generated in the current observation stage and the set of necessary operational events recorded in the standard stage data. If all necessary operational events occur and their execution order meets the standard observation procedure, the output status is "Stage Compliance"; otherwise, the output status is "Stage Deviation".

[0046] For intermediate results, the evaluation process compares the intermediate result itself with the preceding inputs it reads. When the input data of a stage record meets the corresponding standard stage data, but the output data of the stage record does not, the corresponding observation stage is identified as a candidate deviation stage. When the input data of a stage record has deviated from the corresponding standard stage data, and the deviation of the output data can be formed by the input data deviation along the preceding dependency, the corresponding observation stage is not directly identified as a new candidate deviation stage, but is first marked as a stage to be attributed.

[0047] The evaluation processing unit selects the earliest observation stage from the candidate deviation stages appearing in the execution order as the first deviation stage. The first deviation stage represents the earliest position in the current observation evidence chain where the input conditions still meet the standard stage data, but the output result begins to deviate from the standard stage data.

[0048] After identifying the first deviation stage, the evaluation processing unit reads which subsequent observation stages directly reference the output data of the first deviation stage, and continues to read the reference destinations of the output data of subsequent observation stages. When the input data of a subsequent observation stage directly or indirectly references the output data of the first deviation stage, and the output data of the subsequent observation stage deviates from the data of the corresponding standard stage, the subsequent observation stage is identified as a dependent deviation stage.

[0049] If subsequent observation phases use the output data from the first deviation phase, but the trainee corrects the data by re-entering, recalculating, or adjusting the results to make the output data of subsequent observation phases meet the corresponding standard phase data, then the subsequent observation phases are not marked as deviation phases. The evaluation processing end retains the operation events and output data before and after the correction to indicate that the subsequent observation phases have blocked the continued propagation of the previous deviation.

[0050] For observation evidence chains that do not have candidate bias stages, the evaluation processing end records the first bias stage status as "no bias stage". For standard observation processes with multiple independent branches, the evaluation processing end determines the branch candidate bias stages within each branch, and then determines the first bias stage corresponding to the entire task according to the time marker of the first occurrence of bias in each branch and the execution order of the branches in the standard observation process.

[0051] Through the above processing, the evaluation process generates stage deviation results. The stage deviation results include the first deviation stage identifier, the set of associated deviation stage identifiers, the compliance status of each stage, and the data items of the deviation for each stage.

[0052] S4. Based on the final observation results, an accuracy index is formed; based on the stage-related status, a process credibility index is formed; and the consistency status of evidence is determined.

[0053] The evaluation process first determines the accuracy index based on the data type of the final observation results, following the comparison rules in the standard stage data. To ensure that different data types can be included in the same evaluation process, the accuracy index adopts a unified directional expression: the closer the final observation results are to the standard results, the higher the degree of conformity corresponding to the accuracy index.

[0054] For categorical results, the accuracy index is determined based on category consistency. For numerical results, the evaluation process first determines the error boundary based on the standard value and allowable error range, then maps the absolute difference between the final observed result and the standard value to a unified evaluation interval. When the final observed result is within the allowable error range, the accuracy index is in the compliance range; when it exceeds the allowable error range, the accuracy index decreases as the deviation increases. For directional or vector results, the evaluation process forms the accuracy index based on directional deviation, positional deviation, or the deviation between the vector endpoint and the standard result. Data of different dimensions are not directly added; they are first converted to dimensionless compliance based on their respective allowable error ranges before subsequent comparisons.

[0055] The process reliability index is formed based on the proportion of valid stage records in the observation evidence chain, the stage execution order, the input-output reference relationship, and the continuity of intermediate results. The evaluation processing end checks whether each valid stage record is executed at the position specified in the standard observation procedure, checks whether the input data of subsequent observation stages comes from the output data of the corresponding preceding observation stages, and checks whether the intermediate results can be formed by the preceding inputs according to the processing relationship of the corresponding observation tasks.

[0056] The process reliability index also adopts a unified direction of expression: the more complete the valid stage records, the more the execution order conforms to the standard observation process, and the more continuous the input-output reference relationship, the higher the degree of conformity of the corresponding process reliability index. Invalid stage records and missing states are not directly used as trainee bias, but rather serve as evidence that reduces the process reliability index.

[0057] After the result accuracy index and process reliability index are formed, the evaluation processing end maps the two indices to the same evaluation interval. The mapping boundary is determined by the standard stage data and allowable error range for the corresponding task type. For the result accuracy index, the boundary of the allowable error range is mapped to the boundary between result conformity and result deviation; for the process reliability index, a complete process state is mapped when all necessary observation stages specified in the standard observation procedure are valid and all preceding dependencies are established. The number of missing necessary observation stages and the number of broken dependencies are used to reduce the process reliability index.

[0058] The consistency threshold is used to determine whether the differences between the result accuracy indicator and the process reliability indicator are still within the same evaluation state. The consistency threshold is configured according to the task type and is determined based on the allowable result error and allowable process omission range in the standard stage data. The evaluation processing end calculates the difference between the two indicators within a unified evaluation interval. When the difference between the two indicators is not greater than the consistency threshold, the evidence consistency status is determined to be consistent; when the difference between the two indicators is greater than the consistency threshold, the evidence consistency status is determined to be inconsistent.

[0059] When the state of evidence consistency is "consistent," it includes two categories: results that are consistent and the process is continuous, and results that deviate but the process evidence can explain the source of the deviation. When the state of evidence consistency is "inconsistent," it includes two categories: the final observation result is within the consistent range but the intermediate process is missing or discontinuous, and the final observation result deviates from the standard result but the input-output relationship in the observation process does not show a corresponding deviation.

[0060] The evaluation processing unit writes the result accuracy index, process credibility index, evidence consistency status, and stage deviation results into the task intermediate evaluation record. The task intermediate evaluation record serves as input for the next step of adjusting the evaluation weights of capability indicators and generating capability assessment results.

[0061] S5. Based on the basic evaluation weights of the evidence completeness and evidence consistency status correction capability indicators, the evaluation evidence is merged and the meteorological observation capability assessment results are generated.

[0062] The evaluation processing unit determines the completeness of evidence based on the number of valid stage records in the observation evidence chain and the total number of stage records required by the standard observation procedure. When calculating the completeness of evidence, the numerator only counts stage records marked as valid, while the denominator counts all stage records that the current meteorological observation task should generate according to the standard observation procedure. Stage records marked as missing or invalid are not included in the numerator.

[0063] The completeness level of evidence is determined based on the completeness boundaries in the task configuration. The completeness boundaries are determined by the number of necessary observation stages for each task type. When all necessary observation stages are valid, the completeness level of evidence is complete; when there are missing stages but it is still possible to trace back to the initial observation stage from the final observation result, the completeness level of evidence is partially complete; when it is impossible to trace back to the initial observation stage from the final observation result, the completeness level of evidence is incomplete.

[0064] Each capability indicator is pre-associated with a basic evaluation weight. The basic evaluation weight represents the initial level of participation of the corresponding capability indicator in the current task. The evaluation processing end adjusts the basic evaluation weight based on the evidence completeness level and evidence consistency status. When the evidence completeness level is complete and the evidence consistency status is consistent, the basic evaluation weight of the corresponding capability indicator is maintained; when the evidence completeness level is partially complete, the evaluation weight of capability indicators dependent on missing stages is reduced, and capability indicators in the task that can still be recorded and verified by valid stages are retained in the evaluation process; when the evidence completeness level is incomplete or the evidence consistency status is inconsistent, the evaluation results of the capability indicators formed by the corresponding task are marked as pending confirmation.

[0065] The evaluation processing end reads the mapping relationship between the first deviation stage and the capability indicators. The observation preparation stage corresponds to the observation preparation capability, the equipment parameter setting or observation condition configuration stage corresponds to the equipment operation capability, the observation object reading and operation execution stage corresponds to the observation execution capability, the intermediate calculation and result conversion stage corresponds to the data processing capability, and the result interpretation and conclusion submission stage corresponds to the result analysis capability.

[0066] When the first deviation phase corresponds to only one capability indicator, the evaluation evidence for that phase is merged into the corresponding capability indicator. When the first deviation phase corresponds to multiple capability indicators, the evaluation processing unit reads the degree of correlation between each capability indicator and the deviation data items in the current observation phase. The degree of correlation is determined by the correspondence between phase data items and capability indicators recorded in the task structure configuration data. Capability indicators that have a direct correspondence with deviation data items are identified as primary capability indicators, and the remaining capability indicators are considered as associated capability indicators. Evaluation evidence for associated deviation phases is only merged into the primary capability indicator to avoid the deviation formed in the same first deviation phase being repeatedly counted in multiple subsequent observation phases.

[0067] For tasks that do not have a first deviation phase, the assessment processing end merges the compliance evidence into the corresponding capability indicators based on the compliance status of each observation phase. For tasks that have a first deviation phase, the assessment processing end merges the deviation data items, phase execution status, result accuracy indicators, and process reliability indicators of the first deviation phase into the main capability indicators, and retains the associated deviation phase identifier to indicate that subsequent deviations are formed by the propagation of previous deviations.

[0068] After the evaluation evidence is consolidated, the assessment processing unit generates intermediate evaluation results for each capability indicator. The intermediate evaluation results include at least the number of valid evidences corresponding to the capability indicator, the status of compliant evidence, the status of biased evidence, the status of the first bias stage, the level of evidence completeness, and the status of evidence consistency.

[0069] The evaluation processing unit generates candidate evaluation levels for corresponding capability items based on the intermediate evaluation results. The candidate evaluation levels are determined according to the evaluation interval in which the intermediate evaluation results fall. The evaluation interval is determined by the capability evaluation boundaries in the task configuration. Before entering the evaluation interval, different capability indicators use dimensionless evaluation results converted to allowable error ranges.

[0070] After candidate evaluation levels are generated, the assessment processing unit performs verification based on the consistency of evidence. When the evaluation evidence corresponding to a capability item is complete, and the result accuracy index, process reliability index, and the status of the first deviation stage correspond to each other, the candidate evaluation level meets the consistency verification conditions, and the assessment processing unit determines the candidate evaluation level as the final capability evaluation level. When the candidate evaluation level of any capability item does not meet the consistency verification conditions, the assessment processing unit retains the intermediate evaluation result of the corresponding capability item and marks the corresponding capability item as pending confirmation.

[0071] The meteorological observation capability assessment result consists of a final capability evaluation level and a pending confirmation status. For capability items that meet the consistency verification conditions, the final capability evaluation level is output; for capability items that do not meet the consistency verification conditions, intermediate evaluation results and a pending confirmation status are output, without directly outputting a definitive level. Thus, the underlying collected data first forms stage deviation results, result accuracy indicators, and process reliability indicators, then forms intermediate evaluation results and candidate evaluation levels, and finally, consistency verification forms the meteorological observation capability assessment result.

[0072] When the completeness of evidence is lower than the completeness threshold in the task configuration, or when the evidence consistency status is inconsistent, the evaluation processing end determines the observation stage to be supplemented. The observation stage to be supplemented is preferentially selected from those corresponding to the missing status; if no missing status exists, the first observation stage where the input-output relationship is discontinuous is selected; if it still cannot be determined, the first observation stage directly corresponding to the capability item to be confirmed is selected.

[0073] The evaluation processing unit generates supplementary observation tasks based on the observation phase to be supplemented. These supplementary observation tasks retain the original task type and are limited to the observation phase to be supplemented and subsequent associated bias phases affected by it. Supplementary observation tasks do not require the re-execution of observation phases unrelated to the capability to be confirmed that already have valid phase records.

[0074] After receiving the supplementary observation task, the training terminal re-collects new stage records for the corresponding observation phase and sends the new stage records to the evaluation and processing terminal. The evaluation and processing terminal replaces the missing or inconsistent stage records in the original task with the new stage records, reconstructs the observation evidence chain for the corresponding part, and re-executes stage comparison, consistency judgment, and capability evaluation.

[0075] If the regenerated meteorological observation capability assessment result meets the criteria for evidence completeness and consistency, the assessment processing terminal outputs the updated meteorological observation capability assessment result. If the regenerated meteorological observation capability assessment result still does not meet the criteria, the assessment processing terminal outputs a manual review instruction. The manual review instruction includes the task identifier, capability items to be confirmed, observation stages to be supplemented, relevant stage records, and evidence inconsistency status, allowing reviewers to confirm based on the same chain of observation evidence.

[0076] In cloud observation tasks, the training terminal presents cloud images and records scene identifiers. After the trainee inputs the cloud category, the final observation result is generated. The standard observation process can include an image presentation stage, an observation object reading stage, a category selection stage, and a result submission stage. The evaluation processing terminal generates an accuracy index based on whether the categories are consistent, and a reliability index based on whether image presentation and observation object reading were completed before category selection, and whether the result submission references the corresponding category selection result.

[0077] In humidity or air pressure observation tasks, the standard observation process can include a raw observation reading stage, an intermediate calculation or correction stage, and a result submission stage. When the raw observations meet the standard stage data but the intermediate calculation results deviate from the standard stage data, the intermediate calculation stage is identified as a candidate deviation stage; when the final result continues to reference the deviated intermediate calculation results, the result submission stage is identified as a dependent deviation stage.

[0078] In the real wind mapping task, the training terminal records the input wind parameters, vector drawing operations, intermediate vector states, and the final wind vector. The standard observation process may include the input parameter reading stage, the vector direction determination stage, the vector length determination stage, the intersection point formation stage, and the result submission stage. The evaluation processing terminal compares the direction, length, intersection point, and final result respectively, and determines the first deviation stage along the reference relationship.

[0079] In the sea surface wind field inversion task, the training terminal records the observation geometric parameters, solution space, fuzzy solution interpretation results, noise conditions, filtering settings, and quality control results. The standard observation process can include the observation condition configuration stage, solution space generation stage, fuzzy solution interpretation stage, filtering processing stage, and quality control result confirmation stage. The evaluation processing terminal determines whether the deviation is passed from the observation condition configuration to subsequent stages based on the input-output relationship of each stage, avoiding the repeated inclusion of the same preceding configuration deviation in multiple subsequent stages.

[0080] This embodiment connects different types of data generated by the training terminal into a traceable chain of observational evidence through stage records and prior dependencies, enabling the final observation results to correspond to the input data, intermediate results, and operational events that led to the final observation results. Stage deviation comparison first identifies the first deviation stage, then identifies the associated deviation stages affected by the first deviation stage, thus creating different states for root cause deviation and propagation deviation. Result accuracy indicators and process reliability indicators first establish a state of evidence consistency, then participate in the adjustment of capability indicator evaluation weights and the verification of candidate evaluation levels, ensuring that the meteorological observation capability assessment results retain the corresponding process evidence.

[0081] Example 2 Please see Figure 5 This embodiment provides a meteorological observation capability assessment system. Compared to the method-flow-based implementation of Embodiment 1, this embodiment differs in that it employs multiple functional units with clearly defined input-output boundaries to carry out the processing procedures of Embodiment 1. Each functional unit transmits data through task identifiers, stage identifiers, stage records, observation evidence chains, stage deviation results, and meteorological observation capability assessment results. The system includes a task structure configuration unit, a multi-source process data access unit, an evidence chain construction unit, a stage deviation tracing unit, an evidence consistency assessment unit, a weight correction unit, and a capability evidence merging unit.

[0082] The task structure configuration unit stores task structure configuration data. This data is indexed by task type and includes at least the observation phase, phase execution order, prerequisite dependencies, standard phase data, capability index mapping relationships, basic evaluation weights, completeness boundaries, and consistency thresholds. The task structure configuration unit receives the task type of the meteorological observation task, queries the corresponding task structure configuration data, and sends the observation phase and prerequisite dependencies to the evidence chain construction unit, the standard phase data to the phase deviation tracing unit, and the capability index mapping relationships and basic evaluation weights to the weight correction unit and the capability evidence merging unit.

[0083] The multi-source process data access unit communicates with the training terminal. It receives scenario condition data, student input data, operation events, intermediate results, and final observation results from the training terminal. Each received data set includes a task identifier, a stage identifier, a time identifier, and a data source identifier. The multi-source process data access unit performs task aggregation based on the task identifier, stage aggregation based on the stage identifier, and then sends the aggregated data to the evidence chain construction unit.

[0084] When the data received by the multi-source process data access unit lacks a task identifier or stage identifier, the multi-source process data access unit does not send the corresponding data to the evidence chain construction unit, but instead returns a field supplementation request to the training terminal. When the received data has a task identifier and a stage identifier but lacks a time identifier, the multi-source process data access unit marks the corresponding data as data to be verified and sends it to the evidence chain construction unit along with the data to be verified.

[0085] The evidence chain construction unit receives the observation stages and preceding dependencies sent by the task structure configuration unit, and receives the aggregated data sent by the multi-source process data access unit. The evidence chain construction unit establishes a stage record for each observation stage, writes the aggregated data into the corresponding stage record, and establishes a reference relationship between the output data of the preceding observation stage and the input data of the subsequent observation stage based on the preceding stage identifier.

[0086] The evidence chain construction unit checks the necessary data items and time sequence in the stage records. Stage records that meet the data requirements of the corresponding observation stage are marked as valid, while stage records that do not meet the data requirements are marked as invalid. For data to be verified, the evidence chain construction unit determines whether the corresponding data can be included in the current observation stage based on the time identifiers of adjacent observation stages in the same task; if it cannot be determined, it outputs a re-acquisition request to the multi-source process data access unit, which then forwards it to the training terminal.

[0087] The evidence chain construction unit connects the valid stage records according to the execution order and prior dependencies to form an observation evidence chain, and sends the observation evidence chain to the stage deviation tracing unit and the evidence consistency assessment unit. The evidence chain construction unit also outputs the number of valid stage records required for evidence completeness calculation, the total number of stage records, and the missing stage identifiers.

[0088] The stage deviation tracing unit receives the observation evidence chain and standard stage data. Following the execution order, the stage deviation tracing unit compares the input and output data recorded in each valid stage with the corresponding standard stage data. When the input data conforms to the standard stage data but the output data deviates from it, the stage deviation tracing unit marks the corresponding observation stage as a candidate deviation stage.

[0089] The stage deviation tracing unit determines the earliest execution order observation stage from the candidate deviation stages as the first deviation stage, and reads the direct and indirect references to the output data of the first deviation stage along the observation evidence chain. When a subsequent observation stage references the output data of the first deviation stage and the output data deviates from the corresponding standard stage data, the subsequent observation stage is marked as a related deviation stage. The stage deviation tracing unit generates the stage deviation result and sends the stage deviation result to the evidence consistency assessment unit and the capability evidence merging unit.

[0090] The Evidence Consistency Assessment Unit receives the observation evidence chain, stage deviation results, and standard stage data. Based on the difference between the final observation results and the standard results, the Evidence Consistency Assessment Unit generates an accuracy index for the results; and based on valid stage records, stage execution order, input-output reference relationships, and the continuity of intermediate results, it generates a process credibility index.

[0091] The evidence consistency assessment unit maps the result accuracy index and process credibility index to the same evaluation interval and compares their differences. When the difference is no greater than the consistency threshold, a consistent status is output; when the difference is greater than the consistency threshold, an inconsistent status is output. The evidence consistency assessment unit then sends the result accuracy index, process credibility index, and evidence consistency status to the weight correction unit and the capability evidence merging unit.

[0092] The weighting adjustment unit receives the number of valid stage records, the total number of stage records, missing stage identifiers, evidence consistency status, capability indicator mapping relationships, and basic evaluation weights. The weighting adjustment unit calculates the evidence completeness level based on the number of valid stage records and the total number of stage records, and adjusts the basic evaluation weights according to the evidence completeness level and evidence consistency status.

[0093] When the evidence completeness level is complete and the evidence consistency status is consistent, the weight correction unit maintains the basic evaluation weight. When the evidence completeness level is partially complete, the weight correction unit reduces the evaluation weight of the capability indicator in the dependency missing stage. When the evidence completeness level is incomplete or the evidence consistency status is inconsistent, the weight correction unit outputs a confirmation flag to the capability evidence merging unit. The weight correction unit then sends the corrected evaluation weight and the confirmation flag to the capability evidence merging unit.

[0094] The capability evidence merging unit receives the stage deviation results, result accuracy indicators, process reliability indicators, evidence consistency status, corrected evaluation weights, confirmation identifiers, and capability indicator mapping relationships. Based on the mapping relationship between the first deviation stage and the capability indicators, the capability evidence merging unit merges the evaluation evidence of the first deviation stage into the corresponding capability indicator, and records the associated deviation stage identifiers as deviation propagation status records, preventing associated deviation stages from being repeatedly formed as independent root cause deviations.

[0095] When the first deviation stage corresponds to multiple capability indicators, the capability evidence merging unit reads the correspondence between the deviation data items and the capability indicators, determines the main capability indicators directly corresponding to the deviation data items, and merges the evaluation evidence corresponding to the related deviation stages into the main capability indicators. The capability evidence merging unit generates intermediate evaluation results for each capability indicator, and then completes the candidate evaluation level verification based on the consistency of the evidence.

[0096] When a candidate evaluation level meets the consistency verification condition, the capability evidence merging unit outputs the final capability evaluation level; when a candidate evaluation level does not meet the consistency verification condition, the capability evidence merging unit outputs intermediate evaluation results and a pending confirmation status. The final capability evaluation level and the pending confirmation status together constitute the meteorological observation capability assessment result.

[0097] When the capability evidence merging unit receives a pending confirmation flag, it determines the observation stage to be supplemented based on the missing stage flag, the first deviation stage, and the location of evidence inconsistency, and outputs a supplementary task instruction to the training terminal. After the training terminal completes the supplementary observation task, the new stage record is re-entered into the evidence chain construction unit via the multi-source process data access unit. Each functional unit re-executes its corresponding processing with the new stage record as input. When the re-formed meteorological observation capability assessment result is still in a pending confirmation state, the capability evidence merging unit outputs a manual review instruction.

[0098] Data transfer between functional units is based on task identifiers. Stage records are generated in the evidence chain construction unit, stage deviation results in the stage deviation tracing unit, evidence consistency status in the evidence consistency assessment unit, corrected evaluation weights in the weight correction unit, and meteorological observation capability assessment results in the capability evidence merging unit. The same task identifier and stage identifier are maintained when the same data object is transferred between different functional units, enabling the output results to be traced back to the corresponding stage record.

[0099] The specific processes, stage divisions, field states, processing order, judgment conditions, data boundaries, and functional unit collaboration relationships in Examples 1 and 2 are used to illustrate the possible implementations of the present invention and should not be limited to the specific embodiments listed. Without departing from the technical solutions described in this invention, any equivalent substitutions, modifications, or combinations made by those skilled in the art regarding the observation stage names, stage record field names, data carrying formats, functional unit execution subjects, execution order, and carrier formats should fall within the scope of protection of this patent. However, changes in names or carrier formats should not alter the corresponding technical objects, prior dependencies, observation evidence chains, the first deviation stage, the associated deviation stage, the evidence consistency status, and the substantive relationship between the meteorological observation capability assessment results.

Claims

1. A method for intelligently assessing meteorological observation capabilities, characterized in that, include: Obtain the task type, scenario parameters, and corresponding standard observation process for meteorological observation tasks. The standard observation process includes multiple observation stages with pre-dependent relationships. The system collects scene condition data, student input data, operation events, intermediate results, and final observation results generated when the training terminal performs the meteorological observation task, and constructs an observation evidence chain according to the observation stage and prior dependencies. The data of each observation stage in the observation evidence chain are compared with the corresponding standard stage data in turn to determine the first deviation stage and the associated deviation stages caused by the first deviation stage. The accuracy index of the results is generated based on the difference between the final observation results and the standard results. The credibility index of the process is generated based on the correlation between the preceding inputs, intermediate results and the final observation results. The consistency status of the evidence is also determined. Based on the evidence completeness of the observation evidence chain and the basic evaluation weight of the evidence consistency status correction capability index, the evaluation evidence is merged and evaluated according to the mapping relationship between the first deviation stage and the capability index to generate a meteorological observation capability assessment result; when the evidence completeness or the evidence consistency status does not meet the confirmation conditions, a manual review or supplementary observation task is triggered.

2. The intelligent assessment method for meteorological observation capabilities according to claim 1, characterized in that, The steps for constructing the observation evidence chain include: establishing a stage record for each observation stage according to the standard observation process. The stage record includes a task identifier, stage identifier, preceding stage identifier, data source identifier, input data, output data, time identifier, and valid status. Based on the aforementioned prior dependencies, a reference relationship is established between the output data of the preceding observation stage and the input data of the subsequent observation stage; When the input data, output data, and time stamp all meet the data requirements of the corresponding observation stage, the stage record is marked as valid. When any data requirement is not met, the stage record is marked as invalid, and the valid stage records are arranged in the execution order to form the observation evidence chain.

3. The intelligent assessment method for meteorological observation capabilities according to claim 2, characterized in that, The steps for determining the first deviation stage and the associated deviation stages include: comparing the input and output data of each valid stage record with the corresponding standard stage data according to the execution order of the stage records in the observation evidence chain; When the input data recorded in a stage meets the corresponding standard stage data but the output data does not meet the corresponding standard stage data, the corresponding observation stage is determined as a candidate deviation stage. The earliest observation stage is selected from the candidate deviation stages that appear in the execution order as the first deviation stage, and the subsequent observation stages that directly or indirectly reference the output data of the first deviation stage and deviate from the corresponding standard stage data are determined as the associated deviation stages.

4. The intelligent assessment method for meteorological observation capabilities according to claim 3, characterized in that, The steps for generating result accuracy indicators, process credibility indicators, and evidence consistency status include: calculating result accuracy indicators based on the difference between the final observation results and the standard results; calculating process credibility indicators based on the consistency of input data, output data, and stage correlations at each observation stage; verifying the consistency between the result accuracy indicators and the process credibility indicators, normalizing the two indicators, calculating the deviation, and determining the evidence consistency status as consistent when the deviation is not greater than a preset consistency threshold, and as inconsistent when the deviation is greater than the preset consistency threshold.

5. The intelligent assessment method for meteorological observation capabilities according to claim 4, characterized in that, The steps for correcting the basic evaluation weights of capability indicators and generating meteorological observation capability assessment results include: determining the evidence completeness level based on the proportion of valid stage records to all stage records in the observation evidence chain; adjusting the basic evaluation weights of each capability indicator according to the evidence completeness level and evidence consistency status; merging the evaluation evidence according to the capability indicator mapping relationship corresponding to the first deviation stage to form intermediate evaluation results for each capability indicator; generating meteorological observation capability assessment results based on the intermediate evaluation results of each capability indicator; and marking the meteorological observation capability assessment results as pending review when the evidence consistency status is inconsistent.

6. The intelligent assessment method for meteorological observation capabilities according to claim 5, characterized in that, The steps for generating meteorological observation capability assessment results include: forming candidate evaluation levels for corresponding capability items based on the intermediate evaluation results of each capability indicator; verifying the candidate evaluation levels according to the consistency status of evidence, and generating the final capability evaluation level when the candidate evaluation levels of each capability item meet the consistency verification conditions; maintaining the intermediate evaluation results of the corresponding capability item and marking the corresponding capability item as pending confirmation when any candidate evaluation level of a capability item does not meet the consistency verification conditions; and generating the meteorological observation capability assessment results based on the final capability evaluation level and the pending confirmation status.

7. The intelligent assessment method for meteorological observation capabilities according to claim 6, characterized in that, The steps to trigger manual review or supplementary observation tasks include: when the completeness of evidence is lower than a preset completeness threshold or the consistency of evidence is inconsistent, determining the observation stage to be supplemented; generating supplementary observation tasks based on the observation stage to be supplemented, and limiting the supplementary observation tasks to correspond only to the observation stage to be supplemented and its subsequent associated deviation stages; receiving the new stage record corresponding to the supplementary observation task, reconstructing the observation evidence chain of the corresponding stage and performing a capability assessment; when the regenerated meteorological observation capability assessment result meets the confirmation conditions, outputting the updated meteorological observation capability assessment result; otherwise, outputting a manual review instruction.

8. The intelligent assessment method for meteorological observation capabilities according to claim 1, characterized in that, The steps for collecting scenario condition data, student input data, operation events, intermediate results, and final observation results include: establishing corresponding data collection rules based on the type of meteorological observation task; recording various types of data at three collection points: the start of the observation phase, the completion of the phase, and the phase transition; assigning unified task identifiers, phase identifiers, and time identifiers to various types of data; and re-acquiring the data for the corresponding observation phase when any observation phase has missing data or abnormal time sequence. If re-acquisition fails, the corresponding observation phase is marked as missing and used as the basis for calculating the completeness of evidence.

9. The intelligent assessment method for meteorological observation capabilities according to claim 1, characterized in that, The mapping relationship between capability indicators and the first deviation stage includes observation preparation capability, equipment operation capability, observation execution capability, data processing capability, and result analysis capability. Each capability indicator corresponds to at least one observation stage. When the first deviation stage corresponds to multiple capability indicators, the main capability indicators are determined based on the degree of influence of the first deviation stage on each capability indicator. Only the main capability indicators are used as the merged evaluation objects for the corresponding related deviation stages to form intermediate evaluation results for each capability indicator. Finally, the meteorological observation capability assessment results are generated based on the intermediate evaluation results of each capability indicator.

10. A meteorological observation capability intelligent assessment system, used to support the meteorological observation capability intelligent assessment method according to any one of claims 1 to 9, characterized in that, It includes a task structure configuration unit, a multi-source process data access unit, an evidence chain construction unit, a stage deviation tracing unit, an evidence consistency assessment unit, a weight correction unit, and a capability evidence merging unit; The task structure configuration unit stores the observation stages, prerequisite dependencies, standard stage data, and capability index mapping relationships corresponding to meteorological observation tasks. The multi-source process data access unit communicates with the training terminal to receive scenario condition data, student input data, operation events, intermediate results and final observation results; The evidence chain construction unit associates the received data with the observation stage and the preceding dependency relationship to form an observation evidence chain; The stage deviation tracing unit compares the stage data in the observation evidence chain with the standard stage data to determine the first deviation stage and the associated deviation stages. The evidence consistency assessment unit determines the evidence consistency status based on the differences in results and the stage-related status. The weighting correction unit uses the basic evaluation weights of the evidence completeness and the evidence consistency status correction capability index; The capability evidence merging unit merges and evaluates the evidence based on the mapping relationship between the first deviation stage and the capability index, and outputs the meteorological observation capability assessment result. When the evidence completeness or the evidence consistency status does not meet the confirmation conditions, it outputs a review instruction or a supplementary task instruction.

Images