Job permission information consistency checking method based on multi-source field observation, electronic device and medium

Abstract

This invention relates to the field of industrial operation safety technology. It provides a consistency verification method, electronic device, and medium based on work permit information and multi-source on-site observations. The method includes: determining permitted process information based on the electronic work permit corresponding to the target operation; determining candidate processes based on first observation data of the work area where the target operation is located; determining an observation data type based on the permitted process information and the candidate processes; the observation data type is used to distinguish between the permitted process and the candidate processes; acquiring second observation data corresponding to the observation data type; and determining whether the actual process implemented by the target operation is consistent with the permitted process based on the second observation data. This addresses the technical problem in related technologies where the consistency between the actually implemented process and the permitted process is difficult to accurately determine, thus affecting the accuracy and reliability of operation process supervision.

Description

Technical Field

[0001] This invention relates to the field of industrial operation safety technology, and more specifically, to a consistency verification method, electronic device, and medium based on work permit information and multi-source field observation. Background Technology

[0002] With the improvement of digital and intelligent management in industrial production sites, hot work, cutting, welding, and other operations typically require electronic work permits before commencement to record information such as the work content, permitted processes, work area, personnel, permitted time period, and safety measures. Simultaneously, inspection robots, fixed camera equipment, and various on-site sensors can acquire on-site observation data such as images, videos, sounds, and equipment operating status within the work area, providing a data foundation for operational process supervision.

[0003] In existing work safety management solutions, electronic work permits can be used to complete work applications, approvals, authorizations, and record-keeping. On-site inspection equipment can monitor workers, tools, work behaviors, and the on-site environment based on preset image recognition, behavior recognition, equipment status detection, or environmental parameter detection rules. In some solutions, on-site detection results can also be compared with the work information in the electronic work permit to identify situations such as personnel violations, boundary crossings, or abnormal work status.

[0004] However, in related technologies, it is difficult to accurately determine the consistency between the actual implemented process and the licensed process, which in turn affects the accuracy and reliability of operational process supervision. Summary of the Invention

[0005] Therefore, it is necessary to address the above-mentioned problems by providing a consistency verification method, electronic device, and medium based on work permit information and multi-source field observations. This invention aims to solve the technical problem in related technologies where it is difficult to accurately determine the consistency between the actual implemented process and the permitted process, thereby affecting the accuracy and reliability of work process supervision.

[0006] In a first aspect, the present invention provides a consistency verification method based on work permit information and multi-source field observations. The method includes: determining permitted process information based on the electronic work permit corresponding to the target work; determining candidate processes based on first observation data of the work area where the target work is located; determining an observation data type based on the permitted process information and the candidate processes; the observation data type is used to distinguish between the permitted process and the candidate processes; acquiring second observation data corresponding to the observation data type; and determining whether the actual process implemented by the target work is consistent with the permitted process based on the second observation data.

[0007] Optionally, determining the observation data type based on the licensed process information and the candidate process includes: for each candidate observation data type, determining a first reference feature space and a second reference feature space for the licensed process and the candidate process under the feature dimension corresponding to the candidate observation data type; determining the portion of the second reference feature space that does not overlap with the first reference feature space as the distinguishing feature region of the corresponding candidate observation data type; determining the candidate process distinguishability of each candidate observation data type based on the distinguishing feature region corresponding to each candidate observation data type; and determining the observation data type based on the candidate process distinguishability.

[0008] In the above implementation process, the electronic device can establish reference feature spaces for the licensed process and candidate processes under the same feature dimensions, and can determine the unique feature value range of the candidate process as the distinguishing feature region. Since the determination of the observation data type can revolve around the non-overlapping features between the licensed process and the candidate process, the observation data type can more directly serve the distinction between the licensed process and the candidate process, thereby improving the effectiveness of subsequent field observation data.

[0009] Optionally, acquiring the second observation data corresponding to the observation data type includes: determining acquisition parameters for acquiring the second observation data based on the feature dimension and value range of the distinguishable feature region corresponding to the observation data type; acquiring the second observation data according to the acquisition parameters, so that the second observation data contains the observation information corresponding to the feature dimension and value range; wherein, the acquisition parameters include at least one of acquisition frequency band, spectral window, temporal resolution, exposure parameters, acquisition position, acquisition angle, and acquisition duration.

[0010] In the above implementation process, distinguishing the feature region can limit the feature dimensions and value ranges that need to be acquired, and the acquisition parameters can convert the feature dimensions and value ranges into control quantities that can be executed by the field observation equipment. Since the second observation data can be acquired around the distinguishing feature region, the second observation data can contain effective observation information for distinguishing between permitted processes and candidate processes, thereby reducing the impact of irrelevant information on process verification.

[0011] Optionally, determining whether the actual process implemented by the target operation is consistent with the licensed process based on the second observation data includes: determining actual process characteristics based on the second observation data; determining a first matching degree between the actual process characteristics and the first reference feature space, and a second matching degree between the actual process characteristics and the second reference feature space; determining that the actual process corresponds to the candidate process and that the actual process is inconsistent with the licensed process when the actual process characteristics are located within the distinguishing feature region, the second matching degree is greater than the first matching degree, and the difference between the second matching degree and the first matching degree is greater than a preset judgment threshold.

[0012] In the above implementation process, actual process features can be matched with both licensed and candidate processes simultaneously. Distinguishing feature regions can provide exclusive judgment conditions that differentiate candidate processes from licensed processes. Since the determination of process inconsistency simultaneously satisfies three conditions—distinguishing feature regions, a second matching degree greater than a first matching degree, and a matching degree difference greater than a preset judgment threshold—electronic devices can reduce the situation where process deviations are determined solely based on a single similarity or a single threshold, thereby improving the accuracy of process inconsistency judgment.

[0013] Secondly, the present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing program instructions, and the processor executing the steps in the above method when running the program instructions.

[0014] Thirdly, embodiments of the present invention provide a readable storage medium storing computer program instructions, which, when executed by a processor, perform the steps of the above-described method.

[0015] In this invention, by determining the permitted process information based on the electronic work permit corresponding to the target operation, and determining candidate processes based on the first observation data of the work area where the target operation is located, the on-site process determination is no longer based solely on the preset permitted information in the electronic work permit, or solely on a single on-site observation result. By determining the type of observation data used to distinguish between the permitted and candidate processes based on the permitted process information and candidate processes, and obtaining the corresponding second observation data, supplementary on-site observation information with distinguishing function can be obtained to address potential process differences between the permitted and candidate processes. Since the second observation data can provide further basis for the judgment between the permitted and candidate processes, it is possible to determine whether the actual process is consistent with the permitted process based on the second observation data. This helps to solve the technical problem in related technologies where it is difficult to accurately determine the consistency between the actually implemented process and the permitted process, thus affecting the accuracy and reliability of work process supervision. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is an exemplary flowchart of a consistency verification method based on work permit information and multi-source field observations provided in one embodiment of the present invention; Figure 2 This is a structural schematic diagram of an electronic device provided in another embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will now be described with reference to the accompanying drawings. For example, the flowcharts and block diagrams in the drawings illustrate the architecture, functions, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, program segment, or part of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions. In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0019] The following describes a consistency verification method based on work permit information and multi-source field observations, with reference to specific implementation methods. This consistency verification method can be executed by an electronic device. The electronic device can be communicatively connected to an electronic work permit system, field observation equipment, and data storage equipment. The field observation equipment can include fixed observation equipment, mobile observation equipment, or a group of observation equipment consisting of fixed and mobile observation equipment. The electronic device can receive electronic work permits, first observation data, and second observation data, and can process permitted process information, candidate processes, and actual processes.

[0020] For ease of understanding, the target operation can refer to an operation requiring work permit verification, such as welding, cutting, grinding, or other high-risk operations with identifiable technological characteristics. The operation source can refer to the operational entity or location within the target operation area that generates sound, light, images, vibration, or other physical information that can be acquired by observation equipment, such as the location of an electric arc, the location of a tool's action, or a processing point. The first and second observation data can represent two sets of field observation data used for different processing purposes. The first observation data can be used to form candidate processes, and the second observation data can be used to further distinguish between permitted and candidate processes. The first and second observation data do not need to be collected by the same equipment, nor do they need to be collected at fixed time intervals.

[0021] Please see Figure 1 , Figure 1 A flowchart illustrating a consistency verification method based on work permit information and multi-source field observations provided for embodiments of the present invention; the method includes: Step S101: Determine the permitted process information based on the electronic work permit corresponding to the target operation; Step S102: Determine candidate processes based on the first observation data of the work area where the target operation is located; Step S103: Determine the observation data type based on the licensed process information and the candidate process; the observation data type is used to distinguish between the licensed process and the candidate process. Step S104: Obtain the second observation data corresponding to the observation data type; Step S105: Based on the second observation data, determine whether the actual process implemented for the target operation is consistent with the permitted process.

[0022] The following sections provide a detailed explanation of each step.

[0023] In step S101 above, the electronic device can obtain the electronic work permit through interface query, message subscription, file parsing, or database reading. The electronic work permit may include at least one of the following: work permit identifier, work area of ​​the target work, permit time window, permit process name, permit process code, and permit process parameters. The permit process information may represent the process category permitted by the electronic work permit, or it may represent the corresponding work conditions. The electronic device can convert the permit process name in the electronic work permit into a unified process code to reduce the impact of different permit filling formats on subsequent processing. For example, when the electronic work permit states "carbon dioxide gas shielded welding," the electronic device can map "carbon dioxide gas shielded welding" to a predefined gas shielded welding process code.

[0024] In step S102 above, the first observation data can represent the acquired field data that can be used to preliminarily identify the on-site process status. The first observation data can originate from a fixed camera, inspection robot, production equipment, environmental sensors, or a field data platform. The first observation data can include at least one of the following: images, videos, sounds, light signals, equipment operation signals, tool status information, or operator action information. Electronic equipment can extract preliminary process features based on the first observation data and can determine candidate processes using process classification rules, process identification models, or process feature comparison tables. Candidate processes can include one process type or multiple process types sorted according to candidate probability or candidate score.

[0025] In step S103 above, the observation data type can represent the category to which the subsequent field data to be acquired belongs. The observation data type can include acoustic data types, optical data types, image data types, equipment status data types, or other data types that can reflect process differences. The electronic equipment can select the observation data type based on the differences to be verified between the permitted process and the candidate process, without having to indiscriminately collect all field data types. For example, when the permitted process is welding and the candidate process is grinding, the electronic equipment can determine the observation data type that reflects the differences between welding and grinding.

[0026] In step S104 above, the electronic device can send a data acquisition request to the field observation device corresponding to the observation data type, or it can acquire second observation data corresponding to the observation data type from the established data stream. The second observation data may include raw data collected by the field observation device, or data output by the field observation device after preprocessing the raw data. The electronic device can associate the second observation data with the target operation identifier, acquisition time, and observation location to facilitate subsequent determination of the actual process.

[0027] In step S105 above, the electronic device can identify the process characteristics corresponding to the actual process from the second observation data, and can determine the actual process based on the process identification result. The electronic device can compare the actual process with the licensed process. When the actual process and the licensed process are the same, the electronic device can determine that the processes are consistent; when the actual process and the licensed process are different, the electronic device can determine that the processes are inconsistent. The electronic device can also output a confirmation result when the second observation data is insufficient to determine the actual process.

[0028] In the above implementation process, by determining the permitted process information based on the electronic work permit corresponding to the target operation, and determining candidate processes based on the first observation data of the work area where the target operation is located, the on-site process determination is no longer based solely on the preset permitted information in the electronic work permit, or solely on a single on-site observation result. Furthermore, by determining the type of observation data used to distinguish between the permitted and candidate processes based on the permitted process information and candidate processes, and obtaining the corresponding second observation data, on-site observation information with distinguishing function can be obtained to supplement any process differences that may exist between the permitted and candidate processes. Since the second observation data can provide further basis for the judgment between the permitted and candidate processes, it is possible to determine whether the actual process is consistent with the permitted process based on the second observation data. This helps to solve the technical problem in related technologies where the consistency between the actually implemented process and the permitted process is difficult to accurately determine, thus affecting the accuracy and reliability of work process supervision.

[0029] Optionally, in some embodiments, determining the observation data type based on the permitted process information and the candidate processes, i.e., step S103, may include steps S1031 to S1034. Steps S1031 to S1034 are used to select an observation data type with process-discriminating capabilities from a variety of candidate observation data types. Candidate observation data types can represent data categories that can be acquired by field observation equipment and can characterize process features. Step S1031: For each candidate observation data type, determine the first reference feature space and the second reference feature space of the licensed process and the candidate process under the feature dimension corresponding to the candidate observation data type. The feature dimension can represent one or more numerical dimensions used to describe the candidate observation data type. The first reference feature space can represent the set of feature values ​​allowed for the licensed process under the feature dimension corresponding to the candidate observation data type. The second reference feature space can represent the set of feature values ​​allowed for the candidate process under the same feature dimension. The first and second reference feature spaces can be established using feature intervals, feature vector sets, probability distribution regions, cluster boundaries, hyperspheres, hyperrectangles, or other forms that can describe the range of feature values. The electronic equipment can obtain the first and second reference feature spaces based on the process reference model library, historical compliant operation data, process samples processed by equipment calibration, or process samples confirmed manually. For the same candidate observation data type, the electronic equipment can use the same feature extraction rules, missing value handling rules, and normalization rules to process the licensed process reference feature samples and candidate process reference feature samples respectively, and establish the first and second reference feature spaces based on the processed feature samples.

[0030] For example, when the candidate observation data type is acoustic, the feature dimensions may include normalized energy, frequency band energy variation rate, or frequency band energy standard deviation for different frequency bands. The first reference feature space corresponding to the licensed process can represent the common energy range of the licensed process in each frequency band, and the second reference feature space corresponding to the candidate process can represent the common energy range of the candidate process in the same frequency band. When the candidate observation data type is optical, the feature dimensions may include at least one of spectral peak value, light intensity fluctuation, and scintillation frequency.

[0031] Step S1032: The portion of the second reference feature space that does not overlap with the first reference feature space is determined as the distinguishing feature region for the corresponding candidate observation data type. The distinguishing feature region can represent the feature value region where the candidate process can occur, while the permitted process is unlikely to occur or should not occur. The electronic device can determine the distinguishing feature region through set difference operation, interval difference operation, boundary segmentation, density difference calculation, or classification boundary. When the first reference feature space and the second reference feature space are multi-dimensional spaces, the electronic device can determine non-overlapping regions on multiple feature dimensions.

[0032] Step S1033: Based on the distinguishing feature regions corresponding to each candidate observation data type, determine the candidate process distinguishability of each candidate observation data type. The candidate process distinguishability can represent the degree to which the candidate observation data type can distinguish between candidate processes and licensed processes. The candidate process distinguishability can be represented by a score, grade, or ranking value. When the candidate process distinguishability is high, the candidate observation data type can more easily provide observation information for excluding licensed processes or confirming candidate processes.

[0033] Step S1034: Determine the observation data type based on the candidate process distinguishability. The electronic device can determine the observation data type from candidate observation data types whose candidate process distinguishability meets preset conditions, or it can determine the candidate observation data type with the highest candidate process distinguishability as the observation data type. The observation data type may include one candidate observation data type or multiple candidate observation data types.

[0034] In the above implementation process, the electronic device can establish reference feature spaces for the licensed process and candidate processes under the same feature dimensions, and can determine the unique feature value range of the candidate process as the distinguishing feature region. Since the determination of the observation data type can revolve around the non-overlapping features between the licensed process and the candidate process, the observation data type can more directly serve the distinction between the licensed process and the candidate process, thereby improving the effectiveness of subsequent field observation data.

[0035] Optionally, in some embodiments, determining the candidate process distinguishability of each candidate observation data type based on the distinguishing feature regions corresponding to each candidate observation data type, i.e., step S1033 may include steps S10331 to S10333. Steps S10331 to S10333 are used to calculate the candidate process distinguishability based on the proportion of candidate process reference feature samples in the distinguishing feature regions. Permitted process reference feature samples may represent feature samples used to establish a first reference feature space, and candidate process reference feature samples may represent feature samples used to establish a second reference feature space. Permitted process reference feature samples and candidate process reference feature samples may respectively come from historical operation records, test acquisition records, verified process datasets, or feature samples generated based on a process model.

[0036] Step S10331: For each of the candidate observation data types, determine the total number of candidate process reference feature samples in the corresponding second reference feature space; the electronic device can classify and statistically analyze the candidate process reference feature samples according to the candidate observation data type. When the candidate observation data type is an acoustic data type, the total number of candidate process reference feature samples can represent the number of samples used to describe the acoustic features of the candidate process. When the candidate observation data type is an optical data type, the total number of candidate process reference feature samples can represent the number of samples used to describe the optical features of the candidate process.

[0037] Step S10332: Determine the number of candidate process reference feature samples falling into the distinguishable feature regions corresponding to each of the candidate observation data types; the electronic device can compare the feature values ​​of the candidate process reference feature samples with the corresponding distinguishable feature regions. When the feature value of a candidate process reference feature sample is located within a distinguishable feature region, the electronic device can include the candidate process reference feature sample in the number of candidate process reference feature samples. When the feature value of a candidate process reference feature sample is located within the overlapping region of the first reference feature space and the second reference feature space, the electronic device may not include the candidate process reference feature sample in the number of candidate process reference feature samples.

[0038] Step S10333: Based on the number of candidate process reference feature samples and the total number of candidate process reference feature samples, determine the candidate process discriminative power for the corresponding candidate observation data type. Electronic devices can be configured according to the formula... Determine the first Candidate process distinguishability for various candidate observation data types, among which, Indicates the first The number of candidate process reference feature samples falling into the distinguishable feature region under various candidate observation data types. Indicates the first Total number of candidate process reference feature samples under various candidate observation data types It can take values ​​between 0 and 1. When it is zero, electronic devices can Set it to zero, or you can set the first one to zero. The candidate observation data types are removed from the candidate observation data types. When the value is large, the proportion of candidate process reference feature samples that can fall into the distinguishable feature region is relatively high.

[0039] In the above implementation process, the candidate process discrimination can be determined by the proportion of samples falling into the discrimination feature region in the candidate process reference feature sample. Since the candidate process discrimination can reflect the stability of the features of the candidate process outside the permitted process reference feature space, electronic devices can use the candidate process discrimination to quantify the theoretical discrimination ability of different candidate observation data types, thereby reducing the situation of relying solely on a single empirical rule to select the observation data type.

[0040] Optionally, in some embodiments, determining the observation data type based on the candidate process distinguishability may include steps S10341 to S10343. Steps S10341 to S10343 are used to further consider field acquisition conditions based on the candidate process distinguishability to determine the observation data type. The acquisition device status information can indicate whether the field observation device corresponding to the candidate observation data type is currently suitable for outputting usable observation data.

[0041] Step S10341: Based on the first observation data and the acquisition device status information corresponding to each of the candidate observation data types, determine the acquisition quality coefficient corresponding to each candidate observation data type. The acquisition quality coefficient can represent the reliability of acquiring the data corresponding to the candidate observation data type under the current operating environment. The electronic device can determine the acquisition quality coefficient based on at least one of the following in the first observation data: signal-to-noise ratio, target area visibility, background brightness, occlusion degree, target distance, signal integrity, and environmental interference degree. The electronic device can also determine the acquisition quality coefficient based on at least one of the following in the acquisition device status information: device online status, device fault status, lens cleaning status, sensor calibration status, remaining battery power, or communication status.

[0042] Signal quality can be determined based on the signal-to-noise ratio and signal integrity of the corresponding candidate observation data type; coverage quality can be determined based on the coverage ratio of the target operation area by the field observation equipment and the target distance; equipment availability can be determined based on at least one of the equipment online status, equipment fault status, and sensor calibration status. Electronic equipment can use preset threshold mapping or minimum-maximum normalization to map signal quality, coverage quality, and equipment availability to a range of 0 to 1.

[0043] For example, electronic devices can... The signal quality, coverage quality, and equipment availability corresponding to the candidate observation data types are normalized as follows: and And can be in accordance with Determine the data acquisition quality coefficient ,in, Indicates non-negative weights, and , and They represent the first The signal quality, coverage quality, and equipment availability corresponding to each candidate observation data type. When the acquisition equipment is faulty or cannot cover the target operation area, the electronic equipment can set the corresponding acquisition quality coefficient to zero.

[0044] Step S10342: Based on the candidate process discrimination index and acquisition quality coefficient corresponding to each candidate observation data type, determine the discrimination priority corresponding to each candidate observation data type; the discrimination priority can represent the comprehensive priority of the candidate observation data type in distinguishing between licensed processes and candidate processes under the current conditions. Electronic equipment can be configured according to... Determine the first Priority of distinguishing candidate observation data types Alternatively, a weighted summation of the candidate process discrimination index and the acquisition quality coefficient can be used to determine the discrimination priority.

[0045] Step S10343: The candidate observation data type with the highest priority is determined as the observation data type. When the priority of the candidate observation data types is the same, the electronic device can determine the observation data type in order of higher acquisition quality coefficient, shorter device response time, or lower resource consumption.

[0046] In the above implementation process, the candidate process discrimination index reflects the theoretical discrimination capability of the candidate observation data type, while the acquisition quality coefficient reflects the collectability of the candidate observation data type in the current field environment. Since the observation data type can be determined jointly based on the candidate process discrimination index and the acquisition quality coefficient, electronic equipment can avoid selecting candidate observation data types with high theoretical discrimination capability but low field data quality, thereby improving the utilization efficiency of field observation resources and the reliability of process verification results.

[0047] Optionally, in some embodiments, acquiring the second observation data corresponding to the observation data type, i.e., step S104, may include steps S1041 and S1042. Steps S1041 and S1042 are used to convert the distinguishable feature region corresponding to the observation data type into acquisition parameters that can be executed by the field observation equipment. The acquisition parameters may represent a set of parameters used to control the field observation equipment to acquire the second observation data. Step S1041: Based on the feature dimensions and value range of the distinguishable feature region corresponding to the observation data type, determine the acquisition parameters used to acquire the second observation data; the electronic device can pre-store the parameter mapping relationship between feature dimensions, value ranges, and acquisition parameters. When the observation data type is acoustic data type, the feature dimension may include frequency or frequency band energy, and the electronic device can determine the acquisition frequency band and acquisition duration based on the frequency range corresponding to the distinguishable feature region. When the observation data type is optical data type, the feature dimension may include peak wavelength, scintillation frequency, or light intensity variation, and the electronic device can determine the spectral window based on the wavelength range corresponding to the distinguishable feature region, determine the time resolution based on the scintillation frequency range, and determine the exposure parameters based on the light intensity range. When the observation data type is image data type, the feature dimension may include target contour, motion trajectory, or spatial morphology, and the electronic device can determine the acquisition position and acquisition angle based on the work area where the target operation is located.

[0048] Step S1042: Acquire the second observation data according to the acquisition parameters, so that the second observation data contains the observation information corresponding to the feature dimension and value range; the electronic device can send the acquisition parameters to the field observation device, and the field observation device can adjust the data acquisition range, sampling method, or observation position according to the acquisition parameters. The acquisition parameters may include at least one of the following: acquisition frequency band, spectral window, temporal resolution, exposure parameters, acquisition position, acquisition angle, and acquisition duration. The electronic device can attach acquisition parameter identifiers to the second observation data for subsequent interpretation of the observation information in the second observation data.

[0049] In the above implementation process, distinguishing the feature region can limit the feature dimensions and value ranges that need to be acquired, and the acquisition parameters can convert the feature dimensions and value ranges into control quantities that can be executed by the field observation equipment. Since the second observation data can be acquired around the distinguishing feature region, the second observation data can contain effective observation information for distinguishing between permitted processes and candidate processes, thereby reducing the impact of irrelevant information on process verification.

[0050] Optionally, in some embodiments, determining whether the actual process implemented for the target operation is consistent with the licensed process based on the second observation data, i.e., step S105 may include steps S1051 to S1053. Steps S1051 to S1053 are used to convert the second observation data into actual process characteristics, and to use the actual process characteristics to compete and match with the licensed process and candidate processes respectively.

[0051] Step S1051: Based on the second observation data, determine the actual process features. The electronic device can extract the actual process features from the second observation data according to the feature extraction rules corresponding to the observation data type. When the second observation data is acoustic data, the actual process features may include at least one of frequency band energy, frequency band energy change rate, or frequency band energy standard deviation. When the second observation data is optical data, the actual process features may include at least one of peak wavelength, scintillation frequency, or light intensity fluctuation. When the second observation data is image data, the actual process features may include at least one of contour morphology, motion continuity, or spatial distribution features. The electronic device can determine normalization parameters based on reference feature samples of the corresponding candidate observation data type, and normalize the actual process features according to the normalization parameters, so that the actual process features, the first reference feature space, and the second reference feature space have the same feature dimension and numerical scale.

[0052] Step S1052: Determine the first matching degree between the actual process feature and the first reference feature space, and the second matching degree between the actual process feature and the second reference feature space. The first matching degree indicates the degree to which the actual process feature belongs to the permitted process reference feature space. The second matching degree indicates the degree to which the actual process feature belongs to the candidate process reference feature space. The electronic device can first determine the minimum distance from the actual process feature to the first reference feature space. Minimum distance from actual process features to the second reference feature space Then you can follow and Determine the first matching degree respectively Second matching degree ,in, and This can represent the scale parameter used for normalizing the distance. For different candidate observation data types, the electronic device can determine the scale parameter based on the corresponding reference feature samples, and map both the first and second matching degrees to a unified interval of 0 to 1. The difference between the candidate process matching degree and the permitted process matching degree corresponding to different candidate observation data types can be used for subsequent calculation and weighted fusion of process deviation evidence values. The electronic device can also use cosine similarity, Mahalanobis distance, probability density, or the output probability of a classification model to determine the first and second matching degrees.

[0053] Step S1053: When the actual process feature is located within the distinguishing feature region, the second matching degree is greater than the first matching degree, and the difference between the second matching degree and the first matching degree is greater than a preset judgment threshold, it is determined that the actual process corresponds to the candidate process, and that the actual process is inconsistent with the permitted process. The preset judgment threshold can be preset based on historical verification data, process safety level, or allowable false positive rate. When the actual process feature is located within the distinguishing feature region, the actual process feature can provide exclusive information that distinguishes the candidate process from the permitted process. When the second matching degree is greater than the first matching degree and the difference in matching degrees is greater than the preset judgment threshold, the electronic device can determine that the candidate process has a higher interpretability of the actual process feature.

[0054] In the above implementation process, actual process features can be matched with both licensed and candidate processes simultaneously. Distinguishing feature regions can provide exclusive judgment conditions that differentiate candidate processes from licensed processes. Since the determination of process inconsistency simultaneously satisfies three conditions—distinguishing feature regions, a second matching degree greater than a first matching degree, and a matching degree difference greater than a preset judgment threshold—electronic devices can reduce the situation where process deviations are determined solely based on a single similarity or a single threshold, thereby improving the accuracy of process inconsistency judgment.

[0055] Optionally, in some embodiments, the consistency verification method may further include steps S201 to S203. Steps S201 to S203 are used to select another type of observation data to continue obtaining observation evidence when the second observation data cannot form a process judgment result that meets the preset judgment threshold.

[0056] Step S201: When the difference between the second matching degree and the first matching degree does not reach the preset judgment threshold, the candidate observation data type with the highest distinction priority is determined as another observation data type from the candidate observation data types other than the observed data type. When the difference between the second matching degree and the first matching degree does not reach the preset judgment threshold, it can be indicated that the second observation data is not yet sufficient to reliably distinguish between the permitted process and the candidate process. The electronic device can select another observation data type from the candidate observation data types that have not been determined as observation data types. When there are no candidate observation data types that have not been determined as observation data types, the electronic device can output a result to be confirmed or generate a prompt message for manual review. The other observation data type can correspond to different feature dimensions than the observation data type, thereby providing process identification information different from the second observation data.

[0057] Step S202: Obtain third observation data corresponding to the other observation data type; the electronic device can send a data acquisition request to the field observation device corresponding to the other observation data type, or it can obtain the third observation data from the data stream corresponding to the other observation data type. The third observation data can be obtained using a different observation method than the second observation data.

[0058] Step S203: Based on the third observation data, re-determine whether the actual process implemented for the target operation is consistent with the permitted process. The electronic device can process the third observation data based on the first reference feature space, the second reference feature space, and the distinguishing feature region corresponding to the other observation data type, according to the actual process feature extraction and process matching rules corresponding to steps S1051 to S1053, and can form a new process consistency judgment result based on the third observation data.

[0059] In the above implementation process, when the difference between the second matching degree and the first matching degree is insufficient, the electronic device may not directly output the process inconsistency result. Instead, it can select another observation data type with higher distinction priority from the remaining candidate observation data types and obtain third observation data for further judgment. Since the third observation data can introduce process identification information different from the second observation data, the electronic device can reduce misjudgments caused by insufficient quality of a single observation data or overlapping process features.

[0060] Optionally, in some embodiments, the step of re-determining whether the actual process implemented for the target operation is consistent with the permitted process based on the third observation data, i.e., step S203, may include steps S2031 to S2034. Steps S2031 to S2034 are used to fuse the process deviation evidence formed by the second and third observation data. The process deviation evidence value can represent the degree to which a single set of observation data supports the actual process deviating from the permitted process.

[0061] Step S2031: Determine the corresponding licensed process matching degree and candidate process matching degree based on the second observation data and the third observation data respectively; the electronic device can determine the licensed process matching degree corresponding to the second observation data based on the second observation data. Matching degree with candidate processes Alternatively, the degree of matching of the licensed process corresponding to the third observation data can be determined. Matching degree with candidate processes The permitted process matching degree and the candidate process matching degree can be determined using the same or equivalent calculation method as in step S1052. The permitted process matching degree and the candidate process matching degree corresponding to the second observation data and the third observation data can be determined according to the same matching degree mapping rule, so that the process deviation evidence values ​​formed by different observation data types are comparable.

[0062] Step S2032: Based on the difference between the candidate process matching degree and the permitted process matching degree corresponding to each observation data, determine the corresponding process deviation evidence value; the electronic equipment can proceed according to... Determine the process deviation evidence value corresponding to the second observation data. Alternatively, you can follow Determine the process deviation evidence value corresponding to the third observation data. When the evidence value for process deviation is positive and large, it indicates that the corresponding observation data supports the candidate process rather than the permitted process.

[0063] Step S2033: Based on the acquisition quality coefficients corresponding to the second and third observation data, the corresponding process deviation evidence values ​​are weighted and fused; the electronic equipment can be fused according to the following formula: ; The above formula can be used to determine the evidence value of process deviation after weighted fusion. ,in, and These represent the acquisition quality coefficients corresponding to the second and third observation data, respectively. and The acquisition quality coefficients of the observation data types corresponding to the second and third observation data can be used respectively, or they can be determined according to the actual acquisition status of the second and third observation data in the manner corresponding to step S10341. + When the value is zero, the electronic device can output the result to be confirmed.

[0064] Step S2034: When the weighted and fused process deviation evidence value reaches a preset deviation threshold, it is determined that the actual process is inconsistent with the permitted process. The preset deviation threshold can be set according to the safety risk level of different process categories or historical verification results. When the weighted and fused process deviation evidence value does not reach the preset deviation threshold, the electronic device can determine that the process deviation evidence is insufficient.

[0065] In the above implementation process, the second and third observation data can respectively form process deviation evidence values, and the acquisition quality coefficient can reflect the reliability of the second and third observation data. Since the process deviation evidence values ​​can be weighted and fused according to the corresponding acquisition quality coefficient, electronic equipment can reduce the impact of low-quality observation data on process consistency judgment, thereby improving the stability of joint judgment of multiple sets of observation data.

[0066] Optionally, in some embodiments, the first observation data and the second observation data may be associated with job sources respectively; when there are multiple job sources in the job area where the target job is located, the electronic device may first determine whether the first observation data and the second observation data correspond to the same job source, and then use the second observation data to determine the consistency between the actual process and the licensed process.

[0067] The location of the work source indicates its spatial position within the work area where the target work is located. The overlap of the observation areas indicates the degree of spatial overlap between the observation areas of the first and second observation data. The correlation of acquisition times indicates whether the acquisition times corresponding to the first and second observation data fall within the same effective time window of the target work, or whether the time difference between the acquisition times meets preset time conditions. The observation direction indicates the observation orientation of the on-site observation equipment relative to the work source. The similarity of work object features indicates the degree of similarity between tool outlines, workpiece outlines, operator movement characteristics, or equipment identifiers identified from the first and second observation data.

[0068] Step S301: Based on at least two of the following factors when considering the first and second observation data: the location of the operation source, the overlap of the observation area, the correlation of the acquisition time, the observation direction, and the similarity of the operation object features, determine the operation source correlation score. The electronic device can calculate the normalized correlation sub-scores corresponding to the available correlation items respectively, and can determine the operation source correlation score according to the following formula. : ; in, , , , and These represent the correlation sub-scores corresponding to the location of the operation source, the overlap of the observation area, the correlation of the acquisition time, the observation direction, and the similarity of the characteristics of the operation object, respectively. to Weights can be non-negative; when all related items are available. When some related items are unavailable, the electronic device can normalize the weights corresponding to the available related items.

[0069] Step S302: When the job source association score reaches a preset association threshold, it is determined that the first observation data and the second observation data are associated with the same job source. The preset association threshold can be set according to the number of job sources in the work area, the spatial distance, and the positioning accuracy of the observation equipment. When the job source association score does not reach the preset association threshold, the electronic equipment may not use the second observation data for the process consistency judgment of the target job.

[0070] Step S303: After determining that the first observation data and the second observation data are associated with the same job source, the second observation data is used to determine whether the actual process implemented for the target job is consistent with the permitted process. The electronic device can write the job source association result and the second observation data together into the process consistency verification result.

[0071] In the above implementation process, the electronic device can determine whether the first observation data and the second observation data belong to the same operation source by associating at least two operation source items. Since the second observation data is only used for process consistency judgment when the operation source association score reaches a preset association threshold, the electronic device can reduce the situation where observation data generated by adjacent operation sources, adjacent equipment, or adjacent personnel are incorrectly used for the target operation, thereby improving the reliability of process consistency verification in multi-operation source scenarios.

[0072] Based on the same inventive concept, the present invention also provides an electronic device, which includes a memory and a processor. The memory stores program instructions, and when the processor reads and runs the program instructions, it executes the steps in any of the above implementation methods.

[0073] The electronic device can be a variety of digital computers, such as laptops, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframes, etc. The electronic device can also be a variety of mobile devices, such as cellular phones, smartphones, wearable devices, and other similar computing devices.

[0074] Figure 2 An exemplary structural diagram of the electronic device is disclosed. The electronic device includes one or more processors 1101, a memory 1102, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected via different buses 1105 and can be mounted on a common motherboard or otherwise installed as needed. The electronic device may further include an input device 1103 and an output device 1104. The processor 1101, memory 1102, input device 1103, and output device 1104 can be connected via bus 1105 or other means; the diagram shows an example of connection via bus 1105.

[0075] Based on the same inventive concept, embodiments of the present invention also provide a computer-readable storage medium storing computer program instructions, which, when read and executed by a processor, perform the steps in any of the above implementations.

[0076] The above description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A consistency verification method based on work permit information and multi-source field observations, characterized in that, The method includes: Determine the permitted process information based on the electronic work permit corresponding to the target operation; Candidate processes are determined based on the first observation data of the work area where the target operation is located; Based on the licensed process information and the candidate processes, an observation data type is determined; the observation data type is used to distinguish between the licensed process and the candidate processes. Obtain the second observation data corresponding to the observation data type; Based on the second observation data, it is determined whether the actual process implemented for the target operation is consistent with the licensed process.

2. The method according to claim 1, characterized in that, The step of determining the observation data type based on the licensed process information and the candidate processes includes: For each candidate observation data type, the licensing process and the candidate process are respectively determined in the first reference feature space and the second reference feature space under the feature dimension corresponding to the candidate observation data type. The portion of the second reference feature space that does not overlap with the first reference feature space is determined as the distinguishing feature region of the corresponding candidate observation data type; Based on the distinguishing feature regions corresponding to each of the candidate observation data types, the candidate process distinguishability of each of the candidate observation data types is determined; The data type of the observation is determined based on the candidate process distinguishability.

3. The method according to claim 2, characterized in that, The step of determining the candidate process distinguishability of each candidate observation data type based on the distinguishing feature regions corresponding to each candidate observation data type includes: For each of the candidate observation data types, determine the total number of candidate process reference feature samples in the corresponding second reference feature space; Determine the number of candidate process reference feature samples that fall into the distinguishable feature regions corresponding to each of the candidate observation data types; Based on the number of candidate process reference feature samples and the total number of candidate process reference feature samples, the candidate process discrimination of the corresponding candidate observation data type is determined.

4. The method according to claim 2, characterized in that, The step of determining the observation data type based on the candidate process distinguishability includes: Based on the first observation data and the acquisition device status information corresponding to each of the candidate observation data types, the acquisition quality coefficient corresponding to each of the candidate observation data types is determined; Based on the candidate process discrimination degree and acquisition quality coefficient corresponding to each candidate observation data type, determine the discrimination priority corresponding to each candidate observation data type; The candidate observation data type with the highest priority will be identified as the observation data type.

5. The method according to claim 2, characterized in that, The step of obtaining the second observation data corresponding to the observation data type includes: Based on the feature dimensions and value range of the distinguishing feature regions corresponding to the observed data type, the acquisition parameters used to obtain the second observation data are determined; The second observation data is obtained according to the acquisition parameters, so that the second observation data contains the observation information corresponding to the feature dimension and value range; The acquisition parameters include at least one of the following: acquisition frequency band, spectral window, temporal resolution, exposure parameters, acquisition position, acquisition angle, and acquisition duration.

6. The method according to claim 4, characterized in that, The step of determining whether the actual process implemented for the target operation is consistent with the permitted process based on the second observation data includes: Based on the second observation data, the actual process characteristics are determined; Determine the first degree of matching between the actual process features and the first reference feature space, and the second degree of matching between the actual process features and the second reference feature space, respectively. When the actual process feature is located within the distinguishing feature region, the second matching degree is greater than the first matching degree, and the difference between the second matching degree and the first matching degree is greater than a preset judgment threshold, it is determined that the actual process corresponds to the candidate process, and it is determined that the actual process is inconsistent with the permitted process.

7. The method according to claim 6, characterized in that, The method further includes: When the difference between the second matching degree and the first matching degree does not reach the preset judgment threshold, the candidate observation data type with the highest distinction priority is determined as another observation data type from the candidate observation data types other than the observation data type. Obtain third observation data corresponding to the other observation data type; Based on the third observation data, it is re-determined whether the actual process implemented for the target operation is consistent with the permitted process.

8. The method according to any one of claims 1 to 6, characterized in that, The first observation data and the second observation data are each associated with a job source; The job source correlation score is determined based on at least two of the following: job source location, observation area overlap, acquisition time correlation, observation direction, and job object feature similarity, corresponding to the first and second observation data. When the job source association score reaches a preset association threshold, it is determined that the first observation data and the second observation data are associated with the same job source, and the second observation data is used to determine whether the actual process implemented by the target job is consistent with the permitted process.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing program instructions, and the processor executing the steps of the method according to any one of claims 1-8 when running the program instructions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, perform the steps of the method according to any one of claims 1-8.

Images