CN-122022483-A - Expressway construction safety risk assessment system based on knowledge graph

CN122022483ACN 122022483 ACN122022483 ACN 122022483ACN-122022483-A

Abstract

The invention discloses a highway construction safety risk assessment system based on a knowledge graph, which comprises a multi-source construction data acquisition module, a risk state mapping module, a construction safety knowledge graph module, a half-order risk state space construction module, an reachable risk state analysis module, a counterfactual half-order risk space reconstruction module, a half-order structure fracture recognition module and a instability state judgment module, wherein the multi-source construction data acquisition module is used for forming a construction safety original data set, the risk state mapping module is used for mapping and generating a construction safety risk state set, the construction safety knowledge graph module is used for constructing a knowledge graph of the highway construction safety field, the half-order risk state space construction module is used for constructing a construction safety half-order risk state space, the reachable risk state analysis module is used for determining a reachable construction safety risk state subspace, the counterfactual half-order risk space reconstruction module is used for reconstructing the counterfactual half-order risk state space, the half-order structure fracture recognition module is used for recognizing a half-order structure fracture phenomenon, and the instability state judgment module is used for outputting a risk judgment result. The construction safety risk assessment and judgment method based on the knowledge graph and the semi-ordered state structure achieves construction safety risk assessment and judgment, and has the advantages of accuracy and interpretability.

Inventors

LI JUNLIN
ZHAO JIAWEI
WANG SHENGSHENG
CHEN LIJIN
WANG CHU
WEI XING
RUAN JIANFENG
TANG XUAN
Wan Wendong
GENG ZHIGANG
LIU CHAO

Assignees

中国铁建大桥工程局集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260203

Claims (9)

1. The expressway construction safety risk assessment system based on the knowledge graph is characterized by comprising the following steps of: The multi-source construction data acquisition module is used for acquiring multi-source construction data related to safety risks in the expressway construction process, and recording corresponding time identifiers and spatial position identifiers to form a construction safety original data set; The risk state mapping module is used for mapping the construction safety original data to form a construction safety risk state set; The construction safety knowledge graph module is used for constructing a knowledge graph of the expressway construction safety field; The half-order risk state space construction module is used for determining an reachable construction safety risk state subspace in the construction safety half-order risk state space under the current construction condition; The reachable risk state analysis module is used for executing state reachability analysis based on the construction event sequence and determining a reachable construction safety risk state subspace; The inverse fact half-order risk space reconstruction module is used for executing inverse fact stripping processing in the sub-space of the reachable construction safety risk state and reconstructing an inverse fact half-order risk state space; the half-sequence structure fracture identification module is used for identifying whether the construction safety half-sequence risk state space is subjected to structure fracture under the counter-facts condition; and the destabilization state judging module is used for judging that the construction safety risk enters a destabilization state which is unavoidable in the structure and outputting a risk judging result when the half-sequence structure fracture phenomenon exists and no reachable path which is retracted to a safety minimum risk state exists.
2. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein the risk state mapping module comprises: Performing time alignment processing on multi-source construction data according to uniform time identifiers based on a construction safety original data set, wherein the multi-source construction data comprises construction procedure data, traffic organization data, equipment configuration data, personnel operation data and environmental condition data, and obtaining construction procedure states, traffic organization states, equipment operation states, personnel operation states and environmental condition states which are in one-to-one correspondence with the time identifiers; Under the same time mark, performing association processing on the construction process state, the traffic organization state, the equipment running state, the personnel operation state and the environmental condition state based on the space position mark, and performing state mapping processing from the construction state to the risk state according to the state corresponding relation between the states, and mapping various construction states into corresponding components in a construction safety risk state vector respectively; Combining the components corresponding to the same time mark and the same space position mark to generate a construction safety risk state vector corresponding to the time mark and the space position mark; And collecting construction safety risk state vectors corresponding to the different time identifiers and the different space position identifiers to form a construction safety risk state set.
3. The knowledge-based highway construction safety risk assessment system according to claim 1, wherein the construction safety knowledge-graph module comprises: Based on the construction safety original data set, performing entity identification processing on multi-source construction data related in the expressway construction process, determining construction element entities, performing entity identification processing on data related to construction safety risks, determining construction risk entities, and generating corresponding entity identifiers for each construction element entity and each construction risk entity; Constructing a process constraint relation between construction element entities according to the precedence relation and the dependency relation of the construction process in the construction process, and associating the process constraint relation with the corresponding construction element entity; Constructing a space constraint relation between the construction element entity and the construction risk entity according to the corresponding relation between the construction element entity and the construction risk entity on the spatial position identifier, and associating the space constraint relation with the corresponding construction element entity and construction risk entity; constructing a time constraint relation between the construction element entity and the construction risk entity according to the corresponding relation between the construction element entity and the construction risk entity on the time identifier, and associating the time constraint relation with the corresponding construction element entity and construction risk entity; Constructing a safety constraint relation between a construction element entity and a construction risk entity according to safety control conditions formed in the construction safety management process, and associating the safety constraint relation with the corresponding construction element entity and construction risk entity; and uniformly organizing the construction element entity, the construction risk entity, the procedure constraint relation, the space constraint relation, the time constraint relation and the safety constraint relation to form a knowledge graph of the expressway construction safety field.
4. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein the semi-ordered risk state space construction module comprises: based on the construction safety risk state set, selecting any two construction safety risk states as risk state pairs to be judged, and acquiring construction safety risk state vectors corresponding to the construction safety risk states in the risk state pairs to be judged; Aiming at the risk state pair to be judged, extracting a procedure constraint relation, a space constraint relation, a time constraint relation and a safety constraint relation corresponding to the construction safety risk state based on the expressway construction safety field knowledge graph, and organizing the constraint relation into a constraint combination; Under constraint relation limitation, judging whether one construction safety risk state is under constraint combination or not according to the satisfied coverage relation of the two construction safety risk states in the constraint combination, and forming constraint inclusion for the other construction safety risk state; Determining that a one-way dominant relationship exists between two construction safety risk states and recording the one-way dominant relationship as a half-sequence relationship between the construction safety risk states under the condition that the constraint of one construction safety risk state is determined to cover the other construction safety risk state and no constraint relationship exists to enable the coverage relationship to be reversed; And repeatedly executing domination judgment processing on all construction safety risk state pairs in the construction safety risk state set, collecting the generated half-order relations, and constructing a construction safety half-order risk state space based on the half-order relations.
5. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein the reachable risk state analysis module comprises: Under the current construction condition, according to the unified time mark, carrying out time sequence arrangement on construction safety risk state vectors corresponding to each construction safety risk state in the construction safety risk state set, and identifying the process of changing the construction safety risk state vector under the adjacent time mark as a construction event to form a construction event sequence arranged in time sequence; Selecting a construction safety risk state corresponding to the construction event sequence starting time mark from the construction safety half-sequence risk state space as a starting construction safety risk state; For each construction event in a construction event sequence, judging whether construction safety risk state change caused by the construction event keeps the half-order relation undamaged or not according to unidirectional dominant constraint defined by the half-order relation in a construction safety half-order risk state space, and only keeping the construction safety risk state change which keeps the half-order relation to be established; Based on construction safety risk state change which keeps the establishment of a half-order relation, gradually converging construction safety risk states which can be reached by the initial construction safety risk state through a construction event sequence to form a state arrival set; and determining the state arrival set as an reachable construction security risk state subspace under the current construction condition.
6. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein said inverse fact semi-ordered risk space reconstruction module comprises: Taking the subspace of the reachable construction security risk state as an initial state range of the counterfactual analysis, and determining construction factors or construction events as the counterfactual analysis objects in the initial state range; Positioning a construction element entity or a construction risk entity corresponding to a counterfactual analysis object based on a highway construction safety field knowledge graph, and extracting a procedure constraint relation, a space constraint relation, a time constraint relation and a safety constraint relation which are directly related to the construction element entity or the construction risk entity to form a constraint relation set to be stripped; Executing the counterfactual stripping processing on the constraint relation set to be stripped on the premise of keeping the rest construction conditions except the counterfactual analysis object unchanged, thereby generating the counterfactual construction conditions; Under the counterfactual construction condition, taking the subspace of the reachable construction safety risk state as a unique allowed state set, and re-judging the dominant relationship between the construction safety risk states according to the constraint combination after the constraint stripping to generate a semi-ordered relationship set consistent with the counterfactual construction condition; And replacing the half-order relation in the original construction safety half-order risk state space by the half-order relation set, and reorganizing the structural relation between the construction safety risk states based on the replaced half-order relation set to construct a counterfactual half-order risk state space.
7. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein the semi-ordered structure fracture identification module comprises: Taking the reachable construction safety risk state subspace as the unique action range of the semi-ordered structure judgment, and identifying a minimum risk state set which is not governed by any other construction safety risk state under the constraint of a semi-ordered relation based on the construction safety semi-ordered risk state space in the action range; In the limited range of the sub-space of the reachable construction safety risk state, based on the inverse fact half-order risk state space, judging whether the sub-space can still form a lower limit constraint relation with at least one minimum risk state under the constraint of the half-order relation for each construction safety risk state in the sub-space of the reachable construction safety risk state; When it is determined that at least one construction safety risk state can no longer form a lower bound relation with any extremely small risk state in a counterfactual half-order risk state space, determining that the construction safety risk state has a half-order relation under the counterfactual condition to govern connectivity failure; And determining that the construction safety half-sequence risk state space has a half-sequence structure fracture phenomenon under the counter-fact condition when at least one construction safety risk state with half-sequence relation dominant connectivity failure is judged to exist in the reachable construction safety risk state subspace.
8. The knowledge-graph-based highway construction safety risk assessment system according to claim 1, wherein the instability status determination module comprises: Taking the subspace of the reachable construction safety risk state as the only action range of the unavoidable destabilization judgment of the structure, and receiving the half-sequence structure fracture phenomenon determined by the preamble step as the structure judgment premise in the action range; within the limit range of the subspace of the reachable construction safety risk state, judging whether the structural rollback possibility still exists or not under the constraint of a half-order relation based on the inverse fact half-order risk state space aiming at each construction safety risk state; When the construction safety risk state is judged to have no structural rollback possibility in the inverse fact half-order risk state space, determining that the construction safety risk state cannot be rolled back to a safety minimum risk state through a half-order relation under the inverse fact condition; In the subspace of the reachable construction safety risk states, judging that each construction safety risk state meets the condition that the construction safety risk state cannot be returned to a safety minimum risk state through a half-sequence relation, determining that the construction safety risk enters a structurally unavoidable unstable state when the phenomenon of the breakage of the half-sequence structure is confirmed, and outputting a corresponding risk judging result.
9. The knowledge-graph-based highway construction safety risk assessment system according to claim 8, wherein the determination of the safety minimum risk state comprises: determining a candidate minimum risk state set for the minimum risk states identified therein based on the established half-order relationship in the constructed construction safety half-order risk state space; for each minimum risk state in the candidate minimum risk state set, calling a determined constraint combination corresponding to the minimum risk state, and judging whether the constraint combination simultaneously meets a construction safety establishment condition under the current construction condition; When the constraint combination is judged to meet the construction safety establishment condition, determining the corresponding minimum risk state as a safety minimum risk state; When any condition which does not meet the construction safety establishment condition exists in the constraint combination, the corresponding minimum risk state is eliminated as a safety minimum risk state; and taking the determined at least one safety minimum risk state as a safety reference state based on the semi-ordered relation rollback judgment and the structural unavoidable instability judgment.

Description

Expressway construction safety risk assessment system based on knowledge graph Technical Field The invention relates to the technical field of highway engineering construction safety management, in particular to a highway construction safety risk assessment system based on a knowledge graph. Background In the expressway construction process, construction procedures are complex, the operation environment changes dynamically, and construction safety risks are affected by various factors such as procedure arrangement, traffic organization, equipment operation, personnel operation and environmental conditions. The existing construction safety risk assessment technology is usually based on an empirical rule, a statistical analysis or a risk index quantification method, scores, grades or early warns construction safety risks, part of technology is introduced into an informatization system or an intelligent algorithm, summary analysis is carried out on multi-source construction data, and the whole is still in the core of the numerical expression of the risk degree. However, as a result of the evaluation of the multiple risk sizes in the prior art, it is difficult to characterize the structural evolution of the construction safety risk under multiple engineering constraints, and particularly difficult to determine whether the construction safety risk has entered a structurally unavoidable unstable state under a specific combination of construction conditions. The existing method generally lacks the system modeling capability of complex constraint relation among construction elements, and is also difficult to analyze the reversibility of a risk structure under the condition that the construction conditions are supposed to change, so that the key stage of the transition from controllable to unavoidable construction safety risks cannot be effectively identified. Therefore, how to provide a highway construction safety risk assessment system based on a knowledge graph is a problem to be solved by those skilled in the art. Disclosure of Invention The invention aims to provide a highway construction safety risk assessment system based on a knowledge graph, which is based on the knowledge graph and a semi-ordered state structure to realize construction safety risk assessment and judgment and has the advantages of accuracy and interpretability. According to the embodiment of the invention, the expressway construction safety risk assessment system based on the knowledge graph comprises the following steps: The multi-source construction data acquisition module is used for acquiring multi-source construction data related to safety risks in the expressway construction process, and recording corresponding time identifiers and spatial position identifiers to form a construction safety original data set; The risk state mapping module is used for mapping the construction safety original data to form a construction safety risk state set; The construction safety knowledge graph module is used for constructing a knowledge graph of the expressway construction safety field; The half-order risk state space construction module is used for determining an reachable construction safety risk state subspace in the construction safety half-order risk state space under the current construction condition; The reachable risk state analysis module is used for executing state reachability analysis based on the construction event sequence and determining a reachable construction safety risk state subspace; The inverse fact half-order risk space reconstruction module is used for executing inverse fact stripping processing in the sub-space of the reachable construction safety risk state and reconstructing an inverse fact half-order risk state space; the half-sequence structure fracture identification module is used for identifying whether the construction safety half-sequence risk state space is subjected to structure fracture under the counter-facts condition; and the destabilization state judging module is used for judging that the construction safety risk enters a destabilization state which is unavoidable in the structure and outputting a risk judging result when the half-sequence structure fracture phenomenon exists and no reachable path which is retracted to a safety minimum risk state exists. Optionally, the risk state mapping module includes: Performing time alignment processing on multi-source construction data according to uniform time identifiers based on a construction safety original data set, wherein the multi-source construction data comprises construction procedure data, traffic organization data, equipment configuration data, personnel operation data and environmental condition data, and obtaining construction procedure states, traffic organization states, equipment operation states, personnel operation states and environmental condition states which are in one-to-one correspondence with the time identifiers; Under the same time mark,