A dynamic early warning system for subway section safety management and control

Abstract

The application relates to the technical field of subway safety management and control, and discloses a dynamic early warning system for subway field section safety management and control, which comprises a monitoring module, an identification module, a boundary evolution module, a violation prediction module and an early warning module; wherein: the monitoring module is used for collecting field data of each management and control area of a subway field section; the identification module is used for identifying hard violation behaviors and soft violation behaviors of each management and control area; the boundary evolution module is used for learning a discrimination boundary of the soft violation behaviors; the violation prediction module is used for constructing a violation behavior chain of each management and control area and predicting a safety management and control risk of each management and control area based on the violation behavior chain; and the early warning module performs supplementary identification on the safety management and control risk based on the violation behavior chain and performs risk early warning based on the safety management and control risk; the application realizes intelligent identification and accurate early warning of hidden violations of a subway field section and improves the safety management and control level of the subway field section.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of subway safety management and control, in particular to a dynamic early warning system for subway yard section safety management and control. BACKGROUND

[0002] In the prior art, the safety management and control of the subway yard section mainly relies on traditional manual inspection, video monitoring, access control system and other basic means. These traditional inspection methods rely on subjective judgment of human beings, and the hidden dangers are not fully covered, and there is a lack of fine-grained dynamic identification of the behavior trajectory and rule-breaking tendency of the personnel inside the yard section.

[0003] Most of the existing safety monitoring systems rely on hard rule-breaking judgment standards, such as not wearing a safety helmet, not wearing a prescribed uniform, illegal crossing, etc. For soft rule-breaking behaviors that are difficult to define and have fuzzy boundaries, such as personnel frequently approaching restricted areas, repeatedly entering and exiting key areas, etc., the existing systems lack effective dynamic judgment mechanisms, resulting in long-term supervision of implicit rule-breaking behaviors in the blind area, and safety hazards are difficult to be discovered in time.

[0004] The existing rule-breaking judgment generally adopts a static threshold setting method, which cannot be flexibly adjusted in combination with the actual situation of different areas and different types of work tasks. The on-site risk level and personnel behavior characteristics corresponding to different work tasks differ significantly, and the use of a fixed rule-breaking judgment standard uniformly leads to a high misjudgment rate or a high omission rate. In addition, the existing subway yard section safety management system is not perfect in terms of rule-breaking risk warning mechanism, and usually only a simple alarm prompt is given after the rule-breaking behavior has occurred or the rule-breaking behavior has constituted a fact, lacking comprehensive analysis and prediction means for rule-breaking trends and rule-breaking intentions, and unable to realize early perception and active warning, affecting the overall safety management and control level.

[0005] A subway construction safety risk intelligent evaluation system is disclosed in Chinese Patent No. CN119294834B, which includes a field data acquisition module, a device data import module, a construction data analysis module, a construction risk assessment module, a risk sorting and display module, a database, and a warning execution and feedback terminal. The technical solution sets up key areas, and acquires real-time construction site data, imports cumulative construction time length of key areas and real-time equipment monitoring data, solves the deficiency in data acquisition range and depth, and through risk assessment of personnel, equipment and structure, forms an independent and complete risk index analysis system. By constructing a risk indication map in each key area, construction management personnel can conduct comprehensive risk monitoring and management, and the effectiveness of the risk control and feedback link is improved.

[0006] The patent application with publication number CN116882938A discloses a subway protection area engineering activity fine management and control method, including the following steps: obtaining subway protection area data and determining control parameters and safety evaluation indexes; constructing and training a neural network model based on the control parameters and safety evaluation indexes; drawing a four-dimensional cloud chart of the control parameters and safety evaluation index relationship based on the neural network model; based on the four-dimensional cloud chart, obtaining the determined items in the control parameters of the protection area engineering activity to be managed and controlled, and determining the management and control standard or safety of the protection area engineering activity. Compared with the prior art, the technical solution fine determines the subway protection area limit for different engineering activities, and through visual means, the technical personnel can intuitively, quickly and quantitatively judge the influence of the engineering activity on the safety of the subway structure.

[0007] The above technical solutions all have the problem raised in the background art: lack of comprehensive analysis and prediction means for violation trends and violation intentions.

[0008] The information disclosed in this background section is only intended to increase the understanding of the overall background of the present application and should not be considered as acknowledging or implying in any form that this information constitutes prior art known to those of ordinary skill in the art. SUMMARY

[0009] The technical problem to be solved by the present application is to overcome the defects of the prior art and provide a dynamic early warning system for subway field section safety management and control, to realize intelligent identification and accurate early warning of hidden violations of subway field sections, and to improve the safety management and control level of subway field sections.

[0010] To solve the above technical problems, the present application provides the following technical solutions:

[0011] The present application provides a dynamic early warning system for subway field section safety management and control, comprising a monitoring module, an identification module, a boundary evolution module, a violation prediction module and an early warning module; wherein:

[0012] The monitoring module is used to collect field data of each management and control area of the subway field section;

[0013] The identification module identifies hard violation behaviors and soft violation behaviors of each management and control area based on the field data;

[0014] The boundary evolution module learns the discrimination boundary of the soft violation behavior based on historical soft violation records;

[0015] The violation prediction module constructs a violation behavior chain of each management and control area based on the hard violation behavior, the soft violation behavior and the discrimination boundary, and predicts the safety management and control risk of each management and control area based on the violation behavior chain;

[0016] The pre-warning module supplements identification of safety control risks based on the violation chain, and performs risk pre-warning based on the safety control risks.

[0017] As a preferred scheme of the dynamic pre-warning system for metro field section safety control, the field data comprises forbidden area boundary range, personnel positioning data, and monitoring picture data.

[0018] The identification module comprises a first identification unit and a second identification unit.

[0019] The first identification unit is configured to identify hard violation behaviors of each control area; the hard violation behaviors at least comprise not wearing a safety helmet and not wearing a specified workwear.

[0020] The first identification unit is configured with an image recognition algorithm; the first identification unit processes the monitoring picture data based on the image recognition algorithm to identify the hard violation behaviors.

[0021] The second identification unit identifies soft violation behaviors of each control area based on the forbidden area boundary range and the personnel positioning data, and records violation indexes of each soft violation behavior; the soft violation behaviors at least comprise forbidden area approaching behavior and frequent access behavior; the violation index corresponding to the forbidden area approaching behavior is a continuous time length during which the forbidden area approaching behavior is detected; and the violation index corresponding to the frequent access behavior is a number of times of detecting the frequent access to the specified area per unit time.

[0022] As a preferred scheme of the dynamic pre-warning system for metro field section safety control, the boundary evolution module comprises a database unit; the database unit is configured with historical soft violation records of each control area; any historical soft violation record comprises a violation index of a soft violation behavior and a soft violation label; and the soft violation label comprises explicit soft violation and implicit soft violation.

[0023] The explicit soft violation indicates that the violation index of the soft violation behavior is greater than or equal to a corresponding discrimination boundary; and the implicit soft violation indicates that the violation index of the soft violation behavior is less than the corresponding discrimination boundary.

[0024] Any historical soft violation data further comprises a work task type corresponding to the soft violation behavior.

[0025] As a preferred scheme of the dynamic pre-warning system for metro field section safety control, the boundary evolution module further comprises a dynamic boundary unit; the dynamic boundary unit is configured with a boundary evolution strategy for learning discrimination boundaries of the soft violation behaviors; and the boundary evolution strategy specifically comprises:

[0026] obtaining a current job task type of each control area; filtering, from the historical soft violation records, a historical soft violation record of the same job task type as the current one for each control area as a reference soft violation record of the corresponding control area;

[0027] For any control area, classifying the reference soft violation records based on the type of soft violation behavior to obtain reference violation records of each soft violation behavior;

[0028] For any soft violation behavior, further classifying the reference violation records based on the soft violation label; counting all violation indicators corresponding to the explicit soft violation in the reference violation records of the soft violation behavior, and composing a first boundary range of the corresponding soft violation indicator; counting all violation indicators corresponding to the implicit soft violation in the reference violation records of the soft violation behavior, and composing a second boundary range of the corresponding soft violation indicator;

[0029] Determining a discrimination boundary of the soft violation behavior based on the first boundary range and the second boundary range.

[0030] As a preferred scheme of the dynamic early warning system for subway field section safety control described in the present application, wherein: the boundary evolution strategy further comprises determining a discrimination boundary of any soft violation behavior based on the first boundary range and the second boundary range, specifically as follows:

[0031] Setting a false judgment weight value and a missed judgment weight value; the false judgment weight value and the missed judgment weight value are both greater than 0 and less than 1, and the sum is 1;

[0032] When the first boundary range and the second boundary range do not exist intersection, determining the discrimination boundary according to the false judgment weight value and the missed judgment weight value; specifically including: if the false judgment weight value is greater than the missed judgment weight value, the discrimination boundary is the maximum value of the violation indicator in the second boundary range; otherwise, the discrimination boundary is the minimum value of the violation indicator in the first boundary range.

[0033] As a preferred scheme of the dynamic early warning system for subway field section safety control described in the present application, wherein: the boundary evolution strategy further comprises: if the first boundary range and the second boundary range exist intersection, obtaining the values of all violation indicators in the intersection as alternative discrimination boundaries; any alternative discrimination boundary is greater than or equal to the minimum value of the violation indicator in the first boundary range, and less than or equal to the maximum value of the violation indicator in the second boundary range;

[0034] Calculating the false judgment rate and the missed judgment rate of any alternative discrimination boundary as a discrimination boundary, and weighting and summing the false judgment rate and the missed judgment rate to obtain the confidence degree of the corresponding alternative discrimination boundary, wherein the weight of the false judgment rate is the false judgment weight value, and the weight of the missed judgment rate is the missed judgment weight value;

[0035] Select the alternative discrimination boundary with the highest confidence as the discrimination boundary.

[0036] As a preferred scheme of the dynamic early warning system for safety control of metro field section, the violation prediction module comprises a behavior modeling unit and a violation matching unit.

[0037] The behavior modeling unit is configured with a violation modeling strategy for constructing a violation behavior chain of any control area.

[0038] The arrangement order of each hard violation behavior and soft violation behavior is specified, and the violation behavior chain is initialized.

[0039] The corresponding element in the violation behavior chain is assigned a value based on the identification result of the hard violation behavior, wherein if there is any hard violation behavior in the control area, the corresponding element is assigned a value of 1, otherwise, the corresponding element is assigned a value of 0.

[0040] The corresponding element in the violation behavior chain is assigned a value based on the identification result of the soft violation behavior, the violation index, and the corresponding discrimination boundary, specifically as follows: if no hard violation behavior is identified in the control area, the corresponding element is assigned a value of 0; if any hard violation behavior is identified in the control area, the soft violation label is determined based on the violation index and the corresponding discrimination boundary, wherein if the soft violation label is an explicit soft violation, the corresponding element is assigned a value of 1, otherwise, the corresponding element is assigned a value of r, r is a positive number less than 1.

[0041] As a preferred scheme of the dynamic early warning system for safety control of metro field section, the violation matching unit is configured with a reference violation chain for each control area.

[0042] The violation matching unit is also configured with a risk prediction strategy for predicting the safety control risk of any control area.

[0043] The similarity between the violation behavior chain and each reference violation chain is calculated in sequence.

[0044] The field data is continuously collected and the violation behavior chain is updated in real time; when the value of at least one element in the violation behavior chain changes, the similarity between the violation behavior chain and each reference violation chain is recalculated.

[0045] The safety control risk is predicted based on the similarity between the violation behavior chain and the reference violation chain.

[0046] As a preferred scheme of the dynamic early warning system for metro field section safety control, the safety control risk is predicted based on the similarity between the illegal behavior chain and the reference illegal behavior chain, and the safety control risk is specifically predicted by:

[0047] The illegal behavior matching unit is also configured with a first similarity threshold and a second similarity threshold.

[0048] If the similarity between the illegal behavior chain and any reference illegal behavior chain is greater than the first similarity threshold, there is a safety control risk, and a first risk alarm is sent to the early warning module.

[0049] If the similarity between the illegal behavior chain and any reference illegal behavior chain is greater than the second similarity threshold, and the similarity between the illegal behavior chain and the corresponding reference illegal behavior chain continues to increase in the last continuous at least m times of similarity calculation, there is a safety control risk, and a first risk alarm is sent to the early warning module; m is a positive integer.

[0050] As a preferred scheme of the dynamic early warning system for metro field section safety control, the early warning module includes a risk identification unit and a risk early warning unit.

[0051] The risk identification unit is configured with a risk identification strategy for supplementary identification of the safety control risk; the risk identification strategy specifically includes: weighted summation of each element in the illegal behavior chain to obtain a risk value of the corresponding control area.

[0052] The risk identification unit is also configured with a risk value threshold, and if the risk value of any control area is greater than the risk value threshold, there is a safety control risk, and a second risk alarm is sent to the risk early warning unit.

[0053] The risk early warning unit performs risk early warning based on the safety control risk, specifically including: if at least one of the first risk alarm and the second risk alarm is received, risk early warning is performed.

[0054] Compared with the prior art, the application has the following beneficial effects:

[0055] The application improves the identification range and accuracy of the implicit illegal behavior of the metro field section by extracting and discriminating the characteristic parameters of the implicit illegal behavior. By statistically analyzing the historical illegal data and non-illegal data, the implicit illegal discrimination boundary corresponding to each region and each behavior type is automatically determined to ensure that the system discrimination standard is objective and can be dynamically optimized.

[0056] The application realizes intelligent identification, development trend prediction and accurate early warning of the implicit illegal behavior of the metro field section through dynamic determination of the discrimination boundary driven by historical data, task linkage and illegal behavior chain correlation analysis, and comprehensively improves the safety risk control level. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative labor. Among them:

[0058] Figure 1 A structural schematic diagram of a dynamic early warning system for subway field section safety management and control provided by the present application;

[0059] Figure 2 A functional schematic diagram of a dynamic early warning system for subway field section safety management and control provided by the present application. DETAILED DESCRIPTION

[0060] The technical solutions of the present application will be described in detail below by means of the drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, but not limitations of the technical solutions of the present application. The technical features in the embodiments of the present application and the embodiments can be combined with each other without conflict.

[0061] The present embodiment introduces a dynamic early warning system for subway field section safety management and control, which refers to Figure 1 The system comprises a monitoring module, an identification module, a boundary evolution module, a violation prediction module and an early warning module. The functions of the above modules are shown in Figure 2 , wherein:

[0062] The monitoring module is used to collect field data of each management and control area of the subway field section.

[0063] The monitoring module comprises a camera unit, a positioning unit and a map unit.

[0064] The field data comprises forbidden area boundary range, personnel positioning data and monitoring picture data. The camera unit is used to collect the monitoring picture data. The positioning unit is used to collect the personnel positioning data. The map unit is used to obtain an electronic map of the subway field section and mark the forbidden area boundary range of each management and control area.

[0065] The identification module identifies hard violation behaviors and soft violation behaviors of each management and control area based on the field data.

[0066] The identification module comprises a first identification unit and a second identification unit.

[0067] The first identification unit is used to identify the hard violation behaviors of each management and control area. The hard violation behaviors at least include not wearing a safety helmet and not wearing a specified workwear.

[0068] The first identification unit is configured with an image recognition algorithm; the first identification unit processes the monitoring picture data based on the image recognition algorithm to identify the hard violation behavior;

[0069] The second identification unit identifies the soft violation behavior of each management area based on the restricted area boundary range and the personnel positioning data, and records the violation index of each soft violation behavior; the soft violation behavior at least includes a restricted area approaching behavior and a frequent access behavior; wherein the violation index corresponding to the restricted area approaching behavior is the duration of the detected restricted area approaching behavior; and the violation index corresponding to the frequent access behavior is the number of times of detecting the frequent access to the specified area within a unit time.

[0070] The preferred way of identifying the soft violation behavior in this embodiment is as follows: the restricted area approaching behavior is identified according to the restricted area boundary range and the personnel positioning data; for example, when the distance between the positioning of the personnel and the boundary of the restricted area is less than a preset detection threshold, the restricted area approaching behavior is identified; the boundary range of the specified area such as the construction area is determined according to the electronic map, and the number of times of accessing the specified area within a unit time such as 15 minutes is identified in combination with the positioning of the personnel; if the number of times is greater than a preset access number threshold, the frequent access behavior is identified.

[0071] The boundary evolution module learns the discrimination boundary of the soft violation behavior based on the historical soft violation records;

[0072] The boundary evolution module includes a database unit and a dynamic boundary unit;

[0073] The database unit is configured with historical soft violation records of each management area; any historical soft violation record includes a violation index of a soft violation behavior and a soft violation label; the soft violation label includes an explicit soft violation and an implicit soft violation;

[0074] The explicit soft violation means that the violation index of the soft violation behavior is greater than or equal to the corresponding discrimination boundary; and the implicit soft violation means that the violation index of the soft violation behavior is less than the corresponding discrimination boundary. In this embodiment, the soft violation behavior reflects a potential violation intention, and the explicit soft violation means that the violation intention is obvious, and the probability of factual violation is relatively large; on the contrary, the implicit soft violation means that the violation intention is not obvious, and the probability of factual violation is relatively small.

[0075] Any historical soft violation data further includes an operation task type corresponding to the soft violation behavior. The preferred operation task type in this embodiment includes equipment maintenance operation, temporary fault repair, line construction operation, etc. Each operation task type corresponds to different violation identification standards, for example, for temporary fault repair, the violation index of the restricted area approaching behavior related to the equipment is relaxed, and for line construction operation, the violation index of the related frequent access behavior is tightened.

[0076] The dynamic boundary unit is configured with a boundary evolution strategy for learning the discriminant boundary of the soft violation behavior; the boundary evolution strategy specifically includes:

[0077] Obtain the current work task type of each control area; filter the historical soft violation records with the same work task type as the current one from the historical soft violation records as the reference soft violation records of the corresponding control area;

[0078] For any control area, classify the reference soft violation records based on the type of soft violation behavior to obtain the reference violation records of each soft violation behavior;

[0079] For any soft violation behavior, further classify the reference violation records based on the soft violation label; count all violation indicators corresponding to the explicit soft violation in the reference violation records of the soft violation behavior, and form the first boundary range of the corresponding soft violation indicator; count all violation indicators corresponding to the implicit soft violation in the reference violation records of the soft violation behavior, and form the second boundary range of the corresponding soft violation indicator;

[0080] Determine the discriminant boundary of the soft violation behavior based on the first boundary range and the second boundary range.

[0081] The boundary evolution strategy further includes determining the discriminant boundary of any soft violation behavior based on the first boundary range and the second boundary range, specifically as follows:

[0082] Set a false positive weight value and a false negative weight value; the false positive weight value and the false negative weight value are both greater than 0 and less than 1, and the sum is 1;

[0083] The embodiment preferably sets the false positive weight value and the false negative weight value as follows: set a weight comparison table; the weight comparison table is used to specify the false positive weight value and the false negative weight value for different work task types and different soft violation behaviors. For example, the work task type is temporary fault repair, and the soft violation behavior is the forbidden zone close behavior, then the false positive weight value is 0.3 and the false negative weight value is 0.7, that is, the smaller false positive weight value is taken, indicating that for this case, the false positive is moderately tolerated to avoid too many false positives due to complex on-site operation.

[0084] When the first boundary range and the second boundary range do not have an intersection, the discriminant boundary is determined according to the false positive weight value and the false negative weight value; specifically including: if the false positive weight value is greater than the false negative weight value, the discriminant boundary is the maximum value of the violation indicator in the second boundary range; otherwise, the discriminant boundary is the minimum value of the violation indicator in the first boundary range.

[0085] The boundary evolution strategy further comprises: if the first boundary range and the second boundary range have an intersection, obtaining values of all rule violation indicators in the intersection as candidate discriminant boundaries; any candidate discriminant boundary is greater than or equal to the minimum value of the rule violation indicator in the first boundary range and less than or equal to the maximum value of the rule violation indicator in the second boundary range;

[0086] Calculate the misjudgment rate and the missed judgment rate of any candidate discriminant boundary as a discriminant boundary, and weight-sum the misjudgment rate and the missed judgment rate to obtain the confidence degree corresponding to the candidate discriminant boundary, wherein the weight of the misjudgment rate is the misjudgment weight value, and the weight of the missed judgment rate is the missed judgment weight value;

[0087] Take the candidate discriminant boundary with the highest confidence degree as the discriminant boundary.

[0088] In the embodiment, the way of calculating the misjudgment rate and the missed judgment rate of any candidate discriminant boundary as a discriminant boundary is as follows:

[0089] Sample the first boundary range and the second boundary range n times, and extract the value of a rule violation indicator as a target rule violation indicator each time; n is a positive integer;

[0090] If the target rule violation indicator is less than the candidate discriminant boundary, the determination label of the target rule violation indicator is a latent soft rule violation; otherwise, the determination label of the target rule violation indicator is an explicit soft rule violation;

[0091] If the soft rule violation label of the target rule violation indicator is an explicit soft rule violation and is inconsistent with the determination label, the target rule violation indicator is a missed judgment sample; if the soft rule violation label of the target rule violation indicator is a latent soft rule violation and is inconsistent with the determination label, the target rule violation indicator is a misjudgment sample;

[0092] Statistically count the number of misjudgment samples and divide by n to obtain the misjudgment rate; statistically count the number of missed judgment samples and divide by n to obtain the missed judgment rate.

[0093] The current subway field section rule violation recognition mainly relies on hard judgment rules, such as whether to wear a safety helmet, whether to enter a warning area, etc. However, there are a large number of soft rule violations in the actual field, such as frequently approaching the forbidden area but not entering, wandering at the edge of the forbidden area, etc. The application realizes self-adaptive detection of rule violation behaviors according to factors such as work scene and control area position by dynamically learning the discriminant boundary of soft rule violation behaviors.

[0094] The rule violation prediction module constructs a rule violation behavior chain of each control area based on the hard rule violation behaviors, the soft rule violation behaviors and the discriminant boundary, and predicts the safety control risk of each control area based on the rule violation behavior chain;

[0095] The rule violation prediction module comprises a behavior modeling unit and a rule violation matching unit;

[0096] The behavior modeling unit is configured with a violation modeling strategy for constructing a violation behavior chain of any control area; the violation modeling strategy specifically includes:

[0097] The arrangement order of each hard violation behavior and soft violation behavior is specified, and the violation behavior chain is initialized; any element in the violation behavior chain corresponds to a hard violation behavior or a soft violation behavior;

[0098] The corresponding element in the violation behavior chain is valued based on the identification result of the hard violation behavior, wherein if there is any hard violation behavior in the control area, the corresponding element is valued as 1, otherwise, the corresponding element is valued as 0;

[0099] The corresponding element in the violation behavior chain is valued based on the identification result of the soft violation behavior, the violation index and the corresponding discrimination boundary, specifically as follows: if no hard violation behavior is identified in the control area, the corresponding element is valued as 0; if any hard violation behavior is identified in the control area, the soft violation label is determined based on the violation index and the corresponding discrimination boundary, wherein if the soft violation label is an explicit soft violation, the corresponding element is valued as 1, otherwise, the corresponding element is valued as r, r is a positive number less than 1. In the embodiment, the specific value of r is set based on the difference between the discrimination boundary and the violation index, and the greater the difference, the smaller r is.

[0100] The violation matching unit is configured with a reference violation chain of each control area; any reference violation chain is constructed based on the violation modeling strategy; in the embodiment, any reference violation chain corresponds to a safety control risk.

[0101] The violation matching unit is also configured with a risk prediction strategy for predicting the safety control risk of any control area; the risk prediction strategy specifically includes:

[0102] The similarity of the violation behavior chain and each reference violation chain is calculated in turn; in the embodiment, the reciprocal of the Euclidean distance is preferably used as the similarity;

[0103] The violation matching unit is also configured with an update cycle, and at the beginning of each update cycle, a data collection instruction is sent to the monitoring module; the monitoring module responds to the data collection instruction to reacquire the field data;

[0104] The violation behavior chain is updated in real time based on the reacquired field data; when the value of at least one element in the violation behavior chain changes, the similarity of the violation behavior chain and each reference violation chain is recalculated;

[0105] The safety control risk is predicted based on the similarity of the violation behavior chain and the reference violation chain, specifically including:

[0106] The rule violation matching unit is further configured with a first similarity threshold and a second similarity threshold; the first similarity threshold is greater than the second similarity threshold;

[0107] If the similarity between the rule violation chain and any reference rule violation chain is greater than the first similarity threshold, a safety control risk exists, and a first risk warning is sent to the early warning module;

[0108] If the similarity between the rule violation chain and any reference rule violation chain is greater than the second similarity threshold, and the similarity between the rule violation chain and the corresponding reference rule violation chain continues to increase in the last continuous at least m times of similarity calculation, a safety control risk exists, and a first risk warning is sent to the early warning module; m is a positive integer.

[0109] The early warning module performs supplementary identification of the safety control risk based on the rule violation chain, and performs risk early warning based on the safety control risk.

[0110] The early warning module includes a risk identification unit and a risk early warning unit;

[0111] The risk identification unit is configured with a risk identification strategy for supplementary identification of the safety control risk; the risk identification strategy specifically includes: weighted summation of each element in the rule violation chain to obtain a risk value of the corresponding control area;

[0112] In this embodiment, when calculating the risk value, the weight value of each element in the rule violation chain can be set by a person skilled in the art based on actual needs.

[0113] The risk identification unit is further configured with a risk value threshold; if the risk value of any control area is greater than the risk value threshold, a safety control risk exists, and a second risk warning is sent to the risk early warning unit;

[0114] The risk early warning unit performs risk early warning based on the safety control risk, specifically including: if at least one of the first risk warning and the second risk warning is received, risk early warning is performed. In this embodiment, the risk early warning includes at least one of on-site sound and light alarm and monitoring interface pop-up warning.

[0115] The existing subway field section safety control is mainly based on single behavior analysis to make immediate alarm, such as not wearing a safety helmet, mismatched tooling, and crossing the boundary, etc., lacking overall perception of the rule violation chain and prediction of the rule violation development trend. The present application builds a rule violation chain and reference rule violation chains covering typical rule violation development paths, identifies rule violation intention evolution and rule violation trend in real time, predicts whether the rule violation behavior is deteriorating positively, and actively performs early warning intervention before a clear safety risk event is formed, thereby reducing the probability of sudden accidents.

[0116] Those skilled in the art will appreciate that embodiments of the present application can be readily used as a method, a system or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer-readable program code.

[0117] The embodiments of the present application are described above with reference to the accompanying drawings, but the present application is not limited to the above-described specific embodiments, and the above-described specific embodiments are merely illustrative, not restrictive, and those of ordinary skill in the art can make many forms under the inspiration of the present application without departing from the purpose of the present application and the scope of protection, which are all within the scope of protection of the present application.

Claims

1. A dynamic early warning system for safe control of a subway section, characterized in that: The system comprises a monitoring module, an identification module, a boundary evolution module, a violation prediction module, and a warning module. The monitoring module is configured to collect field data of each control area in a subway field section. The identification module is configured to identify hard violation behaviors and soft violation behaviors of each control area based on the field data. The hard violation behaviors include at least not wearing a safety helmet and not wearing a specified workwear. The soft violation behaviors include at least a forbidden area approaching behavior and a frequent entry and exit behavior. The boundary evolution module is configured to learn a discrimination boundary of the soft violation behaviors based on historical soft violation records. The boundary evolution module comprises a dynamic boundary unit configured with a boundary evolution strategy for learning the discrimination boundary of the soft violation behaviors. The boundary evolution strategy comprises the following steps: Obtaining a current work task type of each control area, and screening historical soft violation records of the same work task type as the current work task type from the historical soft violation records as reference soft violation records of the corresponding control area. For any control area, classifying the reference soft violation records based on the types of soft violation behaviors to obtain reference violation records of each soft violation behavior. For any soft violation behavior, further classifying the reference violation records based on soft violation labels, and counting all violation indicators corresponding to explicit soft violation in the reference violation records of the soft violation behavior to form a first boundary range of the corresponding soft violation indicator. Counting all violation indicators corresponding to implicit soft violation in the reference violation records of the soft violation behavior to form a second boundary range of the corresponding soft violation indicator. Determining the discrimination boundary of the soft violation behavior based on the first boundary range and the second boundary range. The violation prediction module is configured to construct a violation behavior chain of each control area based on the hard violation behaviors, the soft violation behaviors, and the discrimination boundary, and to predict a safety control risk of each control area based on the violation behavior chain. The violation prediction module comprises a behavior modeling unit and a violation matching unit. The behavior modeling unit is configured with a violation modeling strategy for constructing a violation behavior chain of any control area. The violation modeling strategy comprises the following steps: Specifying an arrangement order of each hard violation behavior and soft violation behavior, and initializing the violation behavior chain. Assigning a value to a corresponding element in the violation behavior chain based on an identification result of the hard violation behavior, wherein if there is any hard violation behavior in the control area, the corresponding element is assigned a value of 1, otherwise, the corresponding element is assigned a value of 0. Assigning a value to the corresponding element in the violation behavior chain based on an identification result of the soft violation behavior, a violation indicator, and a corresponding discrimination boundary, specifically as follows: if no hard violation behavior is identified in the control area, the corresponding element is assigned a value of 0; if any hard violation behavior is identified in the control area, a soft violation label is determined based on the violation indicator and the corresponding discrimination boundary, wherein if the soft violation label is an explicit soft violation, the corresponding element is assigned a value of 1, otherwise, the corresponding element is assigned a value of r, r being a positive number less than 1. The warning module is configured to supplementally identify a safety control risk based on the violation behavior chain, and to perform risk warning based on the safety control risk.

2. The dynamic early warning system for safe control of metro field section according to claim 1, characterized in that: The field data includes a forbidden zone boundary range, personnel positioning data, and monitoring picture data; The identification module includes a first identification unit and a second identification unit; The first identification unit is configured to identify hard violation behaviors in each control area; The first identification unit is configured with an image recognition algorithm, and the first identification unit processes the monitoring picture data based on the image recognition algorithm to identify the hard violation behaviors; The second identification unit identifies soft violation behaviors in each control area based on the forbidden zone boundary range and the personnel positioning data, and records violation indicators of each soft violation behavior; wherein a forbidden zone approaching behavior corresponds to a violation indicator that is a duration of the detected forbidden zone approaching behavior; A frequent access behavior corresponds to a violation indicator that is a number of times of detecting the frequent access to the specified area per unit time.

3. The dynamic early warning system for subway section safety management according to claim 2, characterized in that: The boundary evolution module includes a database unit; the database unit is configured with historical soft violation records of each control area; any historical soft violation record includes a violation indicator of a soft violation behavior and a soft violation label; the soft violation label includes an explicit soft violation and an implicit soft violation; The explicit soft violation indicates that the violation indicator of the soft violation behavior is greater than or equal to the corresponding discrimination boundary; The implicit soft violation indicates that the violation indicator of the soft violation behavior is less than the corresponding discrimination boundary; Any historical soft violation data further includes a work task type corresponding to the soft violation behavior.

4. The dynamic early warning system for subway section safety management according to claim 3, characterized in that: The boundary evolution strategy further includes determining a discrimination boundary of any soft violation behavior based on the first boundary range and the second boundary range, and specifically as follows: Set a false judgment weight value and a missed judgment weight value; the false judgment weight value and the missed judgment weight value are both greater than 0 and less than 1, and the sum is 1; When the first boundary range and the second boundary range do not have an intersection, the discrimination boundary is determined according to the false judgment weight value and the missed judgment weight value; specifically including: if the false judgment weight value is greater than the missed judgment weight value, the discrimination boundary is the maximum value of the violation indicator in the second boundary range; otherwise, the discrimination boundary is the minimum value of the violation indicator in the first boundary range.

5. The dynamic early warning system for subway field section safety management according to claim 4, characterized in that: The boundary evolution strategy further includes: if the first boundary range and the second boundary range have an intersection, obtaining values of all violation indicators in the intersection as candidate discrimination boundaries; any candidate discrimination boundary is greater than or equal to the minimum value of the violation indicator in the first boundary range, and less than or equal to the maximum value of the violation indicator in the second boundary range; Calculate the false judgment rate and the missed judgment rate of any candidate discrimination boundary as the discrimination boundary, and weight-sum the false judgment rate and the missed judgment rate to obtain the confidence degree corresponding to the candidate discrimination boundary, wherein the weight of the false judgment rate is the false judgment weight value, and the weight of the missed judgment rate is the missed judgment weight value; Take the candidate discrimination boundary with the highest confidence degree as the discrimination boundary.

6. The dynamic early warning system for subway section safety management according to claim 5, characterized in that: The violation matching unit is configured with a reference violation chain of each control area; any reference violation chain is constructed based on the violation modeling strategy; The violation matching unit is further configured with a risk prediction strategy for predicting the safety control risk of any control area; the risk prediction strategy specifically includes: Calculate the similarity between the violation behavior chain and each reference violation chain in turn; continuously collecting the field data and updating the violation chain in real time; triggering a re-computation of the similarity between the violation chain and each reference violation chain when a value of at least one element in the violation chain changes; predicting the safety management risk based on the similarity between the violation chain and the reference violation chain.

7. The dynamic early warning system for subway field section safety management according to claim 6, characterized in that: The method for predicting the safety management risk based on the similarity between the violation chain and the reference violation chain specifically comprises: The violation matching unit is further configured with a first similarity threshold and a second similarity threshold; if the similarity between the violation chain and any reference violation chain is greater than the first similarity threshold, a safety management risk exists, and a first risk warning is sent to the early warning module; if the similarity between the violation chain and any reference violation chain is greater than the second similarity threshold, and the similarity between the violation chain and the corresponding reference violation chain continuously increases in the last m consecutive similarity computations, a safety management risk exists, and a first risk warning is sent to the early warning module; m is a positive integer.

8. The dynamic early warning system for subway field section safety management according to claim 7, characterized in that: The early warning module comprises a risk identification unit and a risk early warning unit; The risk identification unit is configured with a risk identification strategy for supplementary identification of the safety management risk. The risk identification strategy specifically comprises: weighted summation of each element in the violation chain to obtain a risk value of the corresponding management area; The risk identification unit is further configured with a risk value threshold; if the risk value of any management area is greater than the risk value threshold, a safety management risk exists, and a second risk warning is sent to the risk early warning unit; The risk early warning unit performs risk early warning based on the safety management risk; specifically, if at least one of the first risk warning and the second risk warning is received, risk early warning is performed.

Images