Multi-dimensional logistics information supervision and early warning system and method based on artificial intelligence

Abstract

The invention provides a multi-dimensional logistics information supervision and early warning system and method based on artificial intelligence, and relates to the technical field of logistics management. Logistics transportation track information in a logistics supply chain network is acquired in real time; performing compensation check on the cargo right change chain record to obtain a transportation attribute index in the logistics transportation process, and determining a trajectory interaction period when the logistics object performs trajectory tracking on the transportation line according to the transportation attribute index; performing risk feature analysis on the circulation path information to obtain logistics node risk data of the logistics object during transportation, and determining a risk situation level in a logistics transportation process according to the logistics node risk data and the logistics transportation track information; and performing early warning on the dynamic logistics supervision blind area in the logistics combined transportation process according to the track interaction period and the risk situation level. According to the invention, an accurate risk situation assessment mechanism can be constructed in a logistics information fusion supervision mode, so that the accuracy of logistics information supervision is improved.

Description

Technical Field

[0001] The present application relates to the field of logistics management technology, and more specifically, to a multi-dimensional logistics information supervision and early warning system and method based on artificial intelligence. Background Art

[0002] Logistics management refers to the entire process of organizing, coordinating, controlling and optimizing the material circulation and transportation process in the supply chain, covering the storage, transportation, distribution, information transmission and the rational allocation of related resources. With the acceleration of global trade and the rapid development of information technology, logistics management has gradually changed from traditional manual operations to modern management methods based on informatization and intelligence. The goal of logistics management is to improve supply chain efficiency, reduce operating costs, ensure that goods arrive at their destination on time and safely, and ensure the visualization, controllability and accuracy of the entire logistics process through effective supervision of each link.

[0003] However, the existing multi-dimensional logistics information supervision and early warning methods and systems based on artificial intelligence rely on static information flows and cannot obtain information in distributed logistics nodes in real time, resulting in inaccurate and in-time monitoring of cargo status in the logistics process, and failure to effectively integrate various heterogeneous data sources. In particular, there is a lag and lack of accuracy in risk identification and risk situation assessment of dynamic logistics nodes, and failure to effectively respond to emergencies or complex risk changes in the logistics chain, resulting in the failure to timely identify and warn of logistics supervision blind spots. Therefore, how to build an accurate risk situation assessment mechanism under the logistics information fusion supervision model to improve the accuracy of logistics information supervision is a problem faced by the industry. Summary of the invention

[0004] The present application provides a multi-dimensional logistics information supervision and early warning system and method based on artificial intelligence, which can build an accurate risk situation assessment mechanism under the logistics information fusion supervision mode to improve the accuracy of logistics information supervision.

[0005] In a first aspect, the present application provides a multi-dimensional logistics information supervision and early warning method based on artificial intelligence, and the supervision and early warning method comprises the following steps: Collect chain records of cargo ownership changes in the logistics supply chain network from distributed logistics nodes, and obtain logistics transportation track information in the logistics supply chain network in real time; Under the logistics information fusion supervision mode, the chain record of cargo ownership change is compensated and checked to obtain the transportation attribute index in the logistics transportation process, and the trajectory interaction cycle of the logistics object when tracking the trajectory on the transportation route is determined according to the transportation attribute index; Obtaining the flow path information during the operation of the logistics network, analyzing the risk characteristics of the flow path information, obtaining the logistics node risk data of the logistics object during transportation, and determining the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; According to the trajectory interaction cycle and the risk situation level, early warning is given to the dynamic logistics supervision blind spots in the logistics intermodal transport process.

[0006] In this embodiment, the chain record of change of ownership of goods refers to a continuous data chain that records the time, location, participants and related transaction information of the change of ownership of goods in the logistics supply chain network.

[0007] In this embodiment, the logistics transportation trajectory information represents the geographical location, timestamp, transportation tool and transportation status information that the logistics object passes through during the transportation process.

[0008] In this embodiment, determining the trajectory interaction period of the logistics object when tracking the trajectory on the transportation route according to the transportation attribute index specifically includes: Determine the spatiotemporal constraint model of the logistics object in the transportation network according to the transportation attribute index; Outputting a tracking cooperation boundary for dynamic trajectory tracking from the spatiotemporal constraint model; The tracking cooperation boundary determines the trajectory interaction period of the logistics object when tracking the trajectory on the transportation route.

[0009] In this embodiment, obtaining the flow path information during the operation of the logistics network specifically includes: Collect waybill status information between transportation nodes in real time; Determine the core circulation path in the logistics process according to the waybill status information; The core circulation path is used to determine the circulation path information during the operation of the logistics network.

[0010] In this embodiment, the risk characteristics of the circulation path information are analyzed to obtain the logistics node risk data of the logistics object during transportation, which specifically includes: Based on the circulation path information, a spatiotemporal graph association model of the dynamic behavior of logistics nodes and historical risk events is constructed; Output the risk coupling intensity of the logistics object during transportation through the spatiotemporal correlation model; Determine the risk impact characteristics of each logistics node in the transportation network through the risk coupling intensity; The logistics node risk data of the logistics object during transportation is determined according to the risk impact characteristics.

[0011] In this embodiment, determining the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information specifically includes: Determine risk association constraints of logistics during logistics transportation according to the logistics node risk data; Determine the abnormal coupling granularity of logistics in the logistics transportation process through the logistics transportation trajectory information; The risk situation level in the logistics transportation process is determined according to the risk association constraints and the abnormal coupling granularity.

[0012] In this embodiment, the early warning of the dynamic logistics supervision blind spot in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level specifically includes: Determining the amount of elastic monitoring during the logistics intermodal transport process according to the trajectory interaction cycle; Determine the blind spot risk exposure index in the dynamic logistics supervision blind spot according to the risk situation level; The elastic monitoring quantity and the blind spot risk exposure index are matched with warning rules to generate a dynamically adjustable logistics supervision blind spot warning map.

[0013] In this embodiment, under the logistics information fusion supervision mode, the chain record of the change of ownership of the goods is compensated and checked, and the transportation attribute indicators obtained during the logistics transportation process specifically include: Under the logistics information integration supervision mode, the ownership transfer path of the logistics transportation link is determined according to the chain record of cargo ownership change; Probabilistically compensate for the missing cargo ownership transfer nodes in the cargo ownership change chain record to generate a complete logistics event sequence; Performing feature aggregation on the logistics event sequence to obtain risk entropy value information in the logistics transportation process; The transport attribute index in the logistics transport process is determined by the risk entropy value information.

[0014] In a second aspect, the present application provides a multi-dimensional logistics information supervision and early warning system based on artificial intelligence, which is used to execute a multi-dimensional logistics information supervision and early warning method based on artificial intelligence, and the supervision and early warning system includes: The data acquisition module is used to collect the chain records of cargo ownership changes in the logistics supply chain network from the distributed logistics nodes, and obtain the logistics transportation track information in the logistics supply chain network in real time; An association processing module is used to perform compensation inspection on the chain record of cargo ownership change under the logistics information fusion supervision mode, obtain the transportation attribute index in the logistics transportation process, and determine the trajectory interaction cycle of the logistics object when tracking the trajectory on the transportation route according to the transportation attribute index; A risk determination module is used to obtain the flow path information during the operation of the logistics network, analyze the risk characteristics of the flow path information, obtain the logistics node risk data of the logistics object during transportation, and determine the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; The risk warning module is used to warn the dynamic logistics supervision blind spots in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level.

[0015] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects: Collect the chain records of cargo ownership changes in the logistics supply chain network from distributed logistics nodes, and obtain the logistics transportation trajectory information in the logistics supply chain network in real time; under the logistics information fusion supervision mode, perform compensation inspection on the chain records of cargo ownership changes to obtain the transportation attribute indicators in the logistics transportation process, and determine the trajectory interaction period of the logistics object during trajectory tracking on the transportation route according to the transportation attribute indicators; obtain the flow path information during the operation of the logistics network, perform risk feature analysis on the flow path information, obtain the logistics node risk data of the logistics object during transportation, and determine the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; and issue early warnings for dynamic logistics supervision blind spots in the logistics intermodal transport process based on the trajectory interaction period and the risk situation level.

[0016] It can be seen that in this application, dynamic supervision and early warning can be carried out on the multi-dimensional logistics information of each link in the logistics supply chain; among them, by real-time collection of chain records of changes in cargo ownership and logistics transportation trajectory information in distributed logistics nodes, full transparency and real-time monitoring can be achieved, information lag can be reduced, and the visualization and information flow efficiency of the logistics supply chain can be improved; by compensating and checking the chain records of changes in cargo ownership, the integrity and accuracy of logistics event data can be ensured, and the transportation attribute indicators can be further calculated, and dynamic monitoring can be carried out according to the trajectory interaction cycle to provide real-time transportation process analysis; by real-time acquisition and analysis of flow path information, combined with dynamic evaluation of logistics node risk data, it is possible to accurately identify and predict the risk situation in the logistics transportation process; by combining the trajectory interaction cycle and the risk situation level, it is possible to identify the regulatory blind spots in the logistics intermodal transport process, and timely issue warnings through the dynamic early warning mechanism, reduce the blind spot risk exposure in logistics management, improve the initiative and intelligence level of supervision, and ensure the safety and efficiency of the logistics process.

[0017] To sum up, the technical solution adopted in this application can build an accurate risk situation assessment mechanism under the logistics information fusion supervision model to improve the accuracy of logistics information supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0019] Figure 1 It is a flow chart of the multi-dimensional logistics information supervision and early warning method based on artificial intelligence provided by this application; Figure 2 It is a schematic diagram of a process for determining a transport attribute index according to the present application; Figure 3 It is a schematic diagram of the process of determining risk data of logistics nodes according to the application; Figure 4 This is a module structure diagram of the multi-dimensional logistics information supervision and early warning system based on artificial intelligence provided by this application. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0021] The embodiment of the present application provides a multi-dimensional logistics information supervision and early warning system and method based on artificial intelligence. The core of the system is to collect chain records of changes in cargo ownership in a logistics supply chain network from distributed logistics nodes, and obtain logistics transportation trajectory information in the logistics supply chain network in real time; in a logistics information fusion supervision mode, the chain records of changes in cargo ownership are compensated and checked to obtain transportation attribute indicators in the logistics transportation process, and the trajectory interaction period of the logistics object during trajectory tracking on the transportation route is determined according to the transportation attribute indicators; the flow path information during the operation of the logistics network is obtained, and the risk characteristics of the flow path information are analyzed to obtain the logistics node risk data of the logistics object during transportation, and the risk situation level in the logistics transportation process is determined according to the logistics node risk data and the logistics transportation trajectory information; and the dynamic logistics supervision blind spots in the logistics intermodal transport process are warned according to the trajectory interaction period and the risk situation level.

[0022] Embodiment 1: In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1As shown, this figure is an exemplary flow chart of the multi-dimensional logistics information supervision and early warning method based on artificial intelligence shown in this embodiment of the present application. The supervision and early warning method includes the following steps: In step S1, chain records of changes in cargo ownership in the logistics supply chain network are collected from distributed logistics nodes, and logistics transportation track information in the logistics supply chain network is obtained in real time.

[0023] In specific implementation, the collection of chain records of changes in ownership of goods in the logistics supply chain network from distributed logistics nodes can be achieved in the following way: deploy blockchain nodes at each distributed logistics node and use blockchain smart contracts to define the rules for changing ownership of goods. Whenever the ownership of goods changes, such as warehouse receipt, agent transshipment, terminal delivery, etc., the smart contract will automatically record the change information and update it synchronously on all blockchain nodes to ensure data consistency and immutability. At the same time, the IoT sensor equipment is combined to automatically identify the status of the goods, upload it to the blockchain network, and read the chain records of changes in ownership of goods from the blockchain network.

[0024] It should be noted that in this application, distributed logistics nodes refer to independently operated storage, transportation, and distribution links in the logistics supply chain; the logistics supply chain network refers to the cargo flow relationship between logistics nodes, including the connection and collaboration structure of transportation, warehousing, and distribution links; the chain record of changes in cargo ownership refers to a continuous data chain that records the time, place, participants, and related transaction information of changes in cargo ownership in the logistics supply chain network.

[0025] In addition, in the specific implementation, real-time acquisition of logistics transportation track information in the logistics supply chain network can be achieved in the following ways, namely: installing GPS terminals or Beidou positioning devices on transportation vehicles, containers or goods themselves, collecting location information in real time, and sending data to cloud servers through 5G networks or LoRa Internet of Things communication protocols. Use high-precision GIS geographic information systems to parse location information, and combine electronic waybills to record key data such as transportation time and loading and unloading locations. When goods pass through logistics transfer stations, warehouses, and distribution points, use RFID or NFC reading and writing devices to automatically collect logistics node information, use the logistics node information as logistics transportation track information, and synchronize it to the supply chain management system.

[0026] It should be noted that, in the present application, the logistics transportation trajectory information represents the geographical location, timestamp, transportation tool and transportation status information that the logistics object passes through during the transportation process.

[0027] In step S2, under the logistics information fusion supervision mode, the cargo ownership change chain record is compensated and checked to obtain the transportation attribute indicators in the logistics transportation process, and the trajectory interaction cycle of the logistics object during trajectory tracking on the transportation route is determined based on the transportation attribute indicators.

[0028] Preferably, in this embodiment, under the logistics information fusion supervision mode, the cargo ownership change chain record is checked for compensation to obtain the transportation attribute index in the logistics transportation process, and the reference Figure 2 As shown, this figure is a schematic diagram of the process of determining the transport attribute index in some embodiments of the present application. In this embodiment, the transport attribute index can be determined by the following steps: In step S21, in the logistics information fusion supervision mode, the ownership transfer path of the logistics transportation link is determined according to the chain record of the change of ownership of the goods; In step S22, the missing cargo ownership transfer nodes in the cargo ownership change chain record are probabilistically compensated to generate a complete logistics event sequence; In step S23, feature aggregation is performed on the logistics event sequence to obtain risk entropy value information in the logistics transportation process; In step S24, the transportation attribute index in the logistics transportation process is determined according to the risk entropy value information.

[0029] In the specific implementation, first, the blockchain smart contract is used to parse the chain record of cargo ownership change, and the cargo ownership transfer relationship between logistics nodes is extracted. The logistics nodes include warehouses, transshipment centers, and distribution points. The cargo ownership transfer path is constructed by timestamp sorting, and the cargo ownership transfer path is used as the ownership transfer path. The electronic waybill and RFID scanning log are combined to verify the path integrity. If there is a discontinuity in a certain link, it is recorded as a potential missing node. Then, the Markov chain model or hidden Markov model is used to calculate the most likely location and time of the missing node based on the historical cargo ownership transfer pattern. For each possible compensation point, its probability of occurrence is calculated, and the path with the highest probability is selected as the compensation result. The possible miscompensation nodes are identified by combining the anomaly detection algorithm, and cross-validated through the historical records of the supply chain management system. The results of the cross-validation are sorted in chronological order to obtain the logistics event sequence. Then, the TF-IDF weighted method is used to extract features of logistics events and analyze the importance of different types of events in the event sequence. Among them, different types of events include delays and abnormal handovers. The uncertainty of different logistics paths is calculated based on Shannon entropy, and the calculated uncertainty of different logistics paths is used as the risk entropy value information in the logistics transportation process. Finally, based on the risk entropy information, the average transportation time of goods from the starting point to the end point is calculated as the transportation timeliness; the timeliness fluctuation of the same transportation path is measured based on the standard deviation as the transportation stability; the proportion of abnormal handover events is counted as the abnormal handover rate; the proportion of high entropy logistics paths in the overall transportation chain is calculated as the proportion of high-risk paths, and these results are subjected to multivariate regression analysis. Among them, multivariate regression analysis can use linear regression algorithm, and the results of multivariate regression analysis are used as transportation attribute indicators in the logistics transportation process.

[0030] It should be noted that, in this application, the logistics information fusion supervision model refers to the use of multi-source data integration, intelligent analysis and real-time monitoring technology to achieve coordinated supervision of the entire logistics process and improve supply chain transparency and risk management capabilities; the ownership transfer path refers to the time chain of the transfer of goods ownership from one node to another in the logistics process; the cargo ownership handover node refers to the specific time and place where the goods are handed over between different owners in the logistics process; the logistics event sequence refers to the logistics transportation events arranged in chronological order, including key nodes such as cargo handover, transportation, and storage; the risk entropy value information refers to the uncertainty measurement in the logistics transportation process; the transportation attribute index represents the information that measures the important characteristics of the logistics transportation process.

[0031] In this embodiment, the following steps may be used to determine the trajectory interaction period of the logistics object when tracking the trajectory on the transportation route according to the transportation attribute index: Determine the spatiotemporal constraint model of the logistics object in the transportation network according to the transportation attribute index; Outputting a tracking cooperation boundary for dynamic trajectory tracking from the spatiotemporal constraint model; The tracking cooperation boundary determines the trajectory interaction period of the logistics object when tracking the trajectory on the transportation route.

[0032] In the specific implementation, first, the spatiotemporal behavior pattern of logistics objects in the transportation network is determined by combining the indicators of transportation timeliness, transportation stability, and abnormal handover rate, and the Kalman filter or particle filter is used to smooth the historical trajectory to eliminate abnormal deviations and obtain the steady-state trajectory distribution of logistics objects. Then, the spatiotemporal path modeling method is used to calculate the possible stop points and path distribution of logistics objects on different transportation routes based on the GIS geographic information system to form a spatiotemporal constraint model of logistics objects. Then, the movement mode of logistics objects in the transportation network is calculated, where the movement mode includes acceleration, speed change points, and stop points. The density-based clustering method is used to identify different transportation states, where different transportation states include normal driving, stagnation, and abnormal deviation; then, combined with the characteristics of the transportation route, the dynamic time warping technology is used to compare different transportation trajectories and determine the trajectory change boundary of logistics objects under similar transportation conditions; and based on historical transportation data, the Bayesian optimal interval estimation method is used to calculate the reasonable tracking time interval between different transportation nodes of logistics objects to form a tracking coordination boundary. Finally, based on the tracking coordination boundary, the reasonable trajectory interaction cycle range under different transportation modes is defined. Different transportation modes include high-speed driving, low-speed transportation, and warehousing docking. Then, the Markov decision process is used to select the optimal trajectory tracking interval according to the current transportation status. For example, if the logistics object is in a high-speed transportation state, the collection interval is increased; if it enters a high-risk area, the collection interval is shortened. The sampling frequency is dynamically adjusted in the GPS trajectory data stream using an adaptive sampling algorithm to ensure the accuracy of trajectory tracking and data transmission efficiency. The adjusted sampling frequency is used as the trajectory interaction cycle when the logistics object is tracked on the transportation route.

[0033] It should be noted that in the present application, the spatiotemporal constraint model represents the establishment of constraint conditions for trajectory tracking of logistics objects based on their temporal and spatial characteristics; the tracking collaborative boundary represents the reasonable tracking frequency range of logistics objects in different transportation stages; and the trajectory interaction cycle represents the time interval for trajectory tracking of logistics objects on the transportation route.

[0034] In step S3, the flow path information during the operation of the logistics network is obtained, and the risk characteristics of the flow path information are analyzed to obtain the logistics node risk data of the logistics object during transportation, and the risk situation level during the logistics transportation process is determined based on the logistics node risk data and the logistics transportation trajectory information.

[0035] In this embodiment, the flow path information during the operation of the logistics network can be obtained by using the following steps: Collect waybill status information between transportation nodes in real time; Determine the core circulation path in the logistics process according to the waybill status information; The core circulation path is used to determine the circulation path information during the operation of the logistics network.

[0036] In the specific implementation, first, deploy the IoT device GPS terminal at the logistics transportation nodes, such as the origin warehouse, transshipment center, and distribution station, obtain the change of the waybill status of the goods in real time, use the message queue telemetry transmission protocol to transmit the collected waybill status data to the cloud logistics management platform, combine the blockchain evidence technology to ensure the immutability of the waybill status data, and use the blockchain evidence result as the waybill status information. Then, use the graph database to store the waybill flow information, regard the logistics nodes as the vertices in the graph, and the waybill flow relationship as the edge in the graph, construct the logistics network topology structure, calculate the node flow weight, and use the weighted shortest path algorithm to identify the high-frequency and high-weight flow paths in the logistics network, and screen out the core flow paths. Finally, based on the core flow path, combined with the historical waybill data, the K-Means clustering algorithm is used to classify similar flow patterns to form the flow path information of different logistics networks, that is, the flow path information during the operation of the logistics network, and then visualize the flow path information through the geographic information system to monitor the operation status of the logistics network in real time.

[0037] It should be noted that in this application, the waybill status information refers to the cargo status data recorded by different transport nodes during the cargo circulation process; the core circulation path refers to the main circulation direction of the cargo circulation; and the circulation path information refers to the cargo circulation relationship between the various transport nodes in the logistics network.

[0038] Preferably, in this embodiment, the risk characteristics of the circulation path information are analyzed to obtain the logistics node risk data of the logistics object during transportation. Figure 3 As shown, this figure is a schematic diagram of the process of determining logistics node risk data in some embodiments of the present application. In this embodiment, determining the logistics node risk data can be achieved by using the following steps: In step S31, a spatiotemporal graph association model of the dynamic behavior of logistics nodes and historical risk events is constructed based on the circulation path information; In step S32, the risk coupling intensity of the logistics object during transportation is output through the spatiotemporal correlation model; In step S33, the risk impact characteristics of each logistics node in the transportation network are determined by the risk coupling strength; In step S34, the logistics node risk data of the logistics object during transportation is determined according to the risk impact characteristics.

[0039] In the specific implementation, first, a spatiotemporal graph of logistics node risks is constructed, with logistics nodes (warehouses, distribution centers, transportation hubs) as vertices of the graph, cargo flow paths (vehicle transportation routes, air transport routes, railway lines) as edges of the graph, and historical risk events (cargo damage, delays, theft, policy restrictions) as attribute information of the graph. A temporal graph neural network is used to learn the risk propagation model of logistics nodes. The training data sources include historical waybill data, logistics node operation data, and external environment data. Time-weighted calculation is used to calculate the temporal risk impact weight of each logistics node. The long short-term memory network is then used to analyze the changing trend of historical risk events over time. The analysis results are output as a spatiotemporal graph association model of the dynamic behavior of logistics nodes and historical risk events. Next, the spatiotemporal contact frequency between logistics objects and high-risk nodes is calculated. The trajectory similarity calculation is used to analyze the matching degree between the current transportation path and the historical high-risk event path. The distance formula is used to calculate the spatial distance between the current trajectory of the logistics object and the high-risk logistics node. The time window analysis is used to count the residence time of the logistics object in the high-risk area. The coupling strength of the logistics object under different risk factors is evaluated. The multi-factor weighted scoring model is used to comprehensively consider factors such as cargo type, transportation mode, and historical default records to quantify the risk exposure of the logistics object. The Bayesian risk reasoning is used to calculate the risk accumulation value of the logistics object in the current transportation network, and the risk accumulation value is used as the risk coupling strength of the logistics object during transportation. Then, the risk propagation of the logistics node is calculated. The random walk model is used to analyze the risk diffusion ability of the high-risk node to the surrounding nodes, and the flow weighted influence analysis is used to calculate the dynamic risk impact of the node due to the change of logistics flow. The cumulative risk degree of the logistics node is then evaluated. Combined with the kernel density estimation method, it is analyzed whether the logistics node has been a high-frequency risk occurrence site in history. The anomaly detection algorithm is used to identify abnormally high-risk logistics nodes, and the cumulative risk degree of the logistics node obtained by the evaluation is used as the risk impact feature of each logistics node in the transportation network. Finally, the analytic hierarchy process is used in combination with historical data to calculate the risk level of logistics objects on specific transportation routes. The risk index calculation model is used to assign a risk level to each logistics node according to the risk impact characteristics, and each risk level is quantified. The quantification process can be normalized and the quantification results can be used as the logistics node risk data of the logistics object during transportation.

[0040] It should be noted that in the present application, the spatiotemporal graph association model represents a logistics node risk assessment model that includes time, space and association relationships based on dynamic transportation behaviors and historical risk events in the logistics network, which is used to identify high-risk nodes in the logistics network; historical risk events represent recorded risk events that occur during logistics transportation, such as cargo damage, delays, theft or transportation accidents; risk coupling intensity represents the degree of association between the travel path of logistics objects during transportation and risk logistics nodes; risk impact characteristics represent the degree of influence of logistics nodes on the risk level of logistics objects; logistics node risk data refers to the risk impact data on logistics objects when passing through different logistics nodes.

[0041] In this embodiment, the risk situation level in the logistics transportation process can be determined according to the logistics node risk data and the logistics transportation trajectory information by using the following steps: Determine risk association constraints of logistics during logistics transportation according to the logistics node risk data; Determine the abnormal coupling granularity of logistics in the logistics transportation process through the logistics transportation trajectory information; The risk situation level in the logistics transportation process is determined according to the risk association constraints and the abnormal coupling granularity.

[0042] In the specific implementation, first, risk data is obtained from each logistics node, including historical risk events, node traffic flow, weather, public security and other influencing factors. A graph algorithm is used to construct a logistics risk network with each node in the logistics network as a vertex and the cargo flow path as an edge. Association rule mining technology is used to mine the risk correlation between different logistics nodes and identify nodes with strong risk transfer relationships. Bayesian network analysis is then used to calculate the conditional probability and influencing factors of each node in the risk transfer process, and the conditional probability and influencing factors are used as risk association constraints between nodes. Then, based on the real-time collection of transportation trajectories such as GPS and sensor data, combined with transportation methods, including roads, railways, and waterways, the actual paths of logistics objects during transportation are obtained. Spatiotemporal analysis technology is used to match trajectory data with historical risk events, and the contact frequency and residence time of logistics objects with high-risk nodes are identified. The K-means clustering algorithm is used to cluster transportation trajectories and risk nodes according to their contact intensity, and areas with different risk levels are divided. Combined with dynamic time warping technology, the temporal and spatial contact degree between the trajectory of logistics objects and high-risk nodes is calculated, and the abnormal coupling granularity is evaluated. Based on the coupling relationship between the trajectory and the risk node, the risk index model is used to assign risk weights to each trajectory point, and finally the abnormal coupling granularity, that is, the abnormal coupling granularity of logistics during logistics transportation, is obtained. Finally, the risk association between nodes, the granularity of abnormal coupling, and the frequency of occurrence of historical risk events are combined according to weights through the weighted scoring method. Fuzzy logic reasoning is used to set different thresholds for classification based on the input risk data to determine the overall risk situation in the transportation process. According to the weighted scoring results, the hierarchical analysis method is used to determine the risk situation of each logistics node, and the risk model is used to simulate different risk scenarios, including high-risk weather, emergencies, etc., to determine the situation level under different risk conditions. The final output risk situation level is the risk situation level in the logistics transportation process.

[0043] It should be noted that in this application, risk association constraints represent the mutual propagation paths and restriction relationships of risks during the transportation process; abnormal coupling granularity represents the correlation strength and granularity between the trajectory of logistics objects and high-risk nodes or high-risk events during the transportation process; and risk situation level refers to an indicator of the overall risk level in the logistics transportation process.

[0044] In step S4, a warning is issued for the dynamic logistics supervision blind spot in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level.

[0045] In this embodiment, the following steps can be used to provide early warning for dynamic logistics supervision blind spots in the logistics intermodal transport process based on the trajectory interaction cycle and the risk situation level: Determining the amount of elastic monitoring during the logistics intermodal transport process according to the trajectory interaction cycle; Determine the blind spot risk exposure index in the dynamic logistics supervision blind spot according to the risk situation level; The elastic monitoring quantity and the blind spot risk exposure index are matched with warning rules to generate a dynamically adjustable logistics supervision blind spot warning map.

[0046] In the specific implementation, first, through the spatiotemporal data of the logistics transportation path, the trajectory interaction cycle of each transportation node and path point is calculated, that is, the interaction frequency and residence time of the logistics object with a certain node. Based on the frequency-time distribution model, the trajectory interaction cycle of the logistics object is analyzed, and the time window of the key transportation node is calibrated; the adaptive monitoring method is used to dynamically adjust the distribution of monitoring points according to the change of the trajectory interaction cycle. For example, nodes with shorter trajectory interaction cycles can be set with higher monitoring frequencies, and nodes with longer trajectory interaction cycles can be appropriately reduced in monitoring frequency. Then, K-means clustering is combined to cluster different interaction cycles to generate dynamically adjusted monitoring areas, and the data in the monitoring areas are used as elastic monitoring quantities in the logistics intermodal process. Then, based on the risk situation level, combined with the transportation flow and historical risk data of the blind area, the risk exposure index of the blind area is evaluated, and the exposure-sensitivity model is used to analyze the probability and impact range of potential risk events in the blind area. For each blind area, linear regression is used to establish the relationship between the risk exposure index and its external factors, and the high-risk exposure points of the blind area are determined. The high-risk exposure points of the blind area are used as the blind area risk exposure index in the dynamic logistics supervision blind area. Finally, multi-dimensional matching algorithms are used, such as the weighted average method and matching learning algorithm, to match the elastic monitoring quantity with the blind spot risk exposure index, determine the risk exposure level and monitoring needs of each regulatory blind spot, and use a dynamic threshold model to dynamically adjust the matching rules between the blind spot risk exposure index and the elastic monitoring quantity based on real-time monitoring data. Data visualization tools are used to generate real-time updated logistics regulatory blind spot warning maps to dynamically display the risk status of each blind spot. Based on real-time data streams, the monitoring quantity and risk exposure index are updated in real time and visualized in the form of a map to timely display potential high-risk areas and their monitoring status. Combined with the automatic alarm system, when the risk exposure index of a blind spot exceeds the set threshold, an alarm is automatically triggered and the monitoring strategy is adjusted.

[0047] It should be noted that in this application, the flexible monitoring quantity refers to the dynamic monitoring indicator that adapts to different logistics and transportation conditions; the risk exposure index refers to the exposure degree of risk areas that are not effectively supervised during the logistics and transportation process; the logistics supervision blind spot warning map refers to a monitoring map that displays the risk situation and warning status of each supervision blind spot in real time.

[0048] It can be seen that in this application, dynamic supervision and early warning can be carried out on the multi-dimensional logistics information of each link in the logistics supply chain; among them, by real-time collection of chain records of changes in cargo ownership and logistics transportation trajectory information in distributed logistics nodes, full transparency and real-time monitoring can be achieved, information lag can be reduced, and the visualization and information flow efficiency of the logistics supply chain can be improved; by compensating and checking the chain records of changes in cargo ownership, the integrity and accuracy of logistics event data can be ensured, and the transportation attribute indicators can be further calculated, and dynamic monitoring can be carried out according to the trajectory interaction cycle to provide real-time transportation process analysis; by real-time acquisition and analysis of flow path information, combined with dynamic evaluation of logistics node risk data, it is possible to accurately identify and predict the risk situation in the logistics transportation process; by combining the trajectory interaction cycle and the risk situation level, it is possible to identify the regulatory blind spots in the logistics intermodal transport process, and timely issue warnings through the dynamic early warning mechanism, reduce the blind spot risk exposure in logistics management, improve the initiative and intelligence level of supervision, and ensure the safety and efficiency of the logistics process.

[0049] To sum up, the technical solution adopted in this application can build an accurate risk situation assessment mechanism under the logistics information fusion supervision model to improve the accuracy of logistics information supervision.

[0050] Embodiment 2: This application provides a multi-dimensional logistics information supervision and early warning system based on artificial intelligence, referring to Figure 4 As shown, this figure is a module structure diagram of a multi-dimensional logistics information supervision and early warning system based on artificial intelligence according to this embodiment of the present application, and the supervision and early warning system includes: The data acquisition module 100 is used to collect the chain records of the change of ownership of goods in the logistics supply chain network from the distributed logistics nodes, and obtain the logistics transportation track information in the logistics supply chain network in real time; The association processing module 200 is used to perform compensation inspection on the chain record of cargo ownership change in the logistics information fusion supervision mode, obtain the transportation attribute index in the logistics transportation process, and determine the trajectory interaction cycle of the logistics object when tracking the trajectory on the transportation route according to the transportation attribute index; The risk determination module 300 is used to obtain the flow path information during the operation of the logistics network, analyze the risk characteristics of the flow path information, obtain the logistics node risk data of the logistics object during transportation, and determine the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; The risk warning module 400 is used to warn the dynamic logistics supervision blind spots in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level.

[0051] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0052] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically-erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0053] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A multi-dimensional logistics information supervision and early warning method based on artificial intelligence, characterized in that: The regulatory early warning method comprises the following steps: Collect chain records of cargo ownership changes in the logistics supply chain network from distributed logistics nodes, and obtain logistics transportation track information in the logistics supply chain network in real time; Under the logistics information fusion supervision mode, the chain record of cargo ownership change is compensated and checked to obtain the transportation attribute index in the logistics transportation process, and the trajectory interaction cycle of the logistics object when tracking the trajectory on the transportation route is determined according to the transportation attribute index; Obtaining the flow path information during the operation of the logistics network, analyzing the risk characteristics of the flow path information, obtaining the logistics node risk data of the logistics object during transportation, and determining the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; According to the trajectory interaction cycle and the risk situation level, early warning is given to the dynamic logistics supervision blind spots in the logistics intermodal transport process.

2. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: The chain record of change of ownership of goods refers to a continuous data chain that records the time, location, participants and related transaction information of the change of ownership of goods in the logistics supply chain network.

3. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: The logistics transportation track information represents the geographical location, timestamp, transportation tool and transportation status information that the logistics object passes through during the transportation process.

4. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: Determining the trajectory interaction cycle when the logistics object is tracked on the transportation route according to the transportation attribute index specifically includes: Determine the spatiotemporal constraint model of the logistics object in the transportation network according to the transportation attribute index; Outputting a tracking cooperation boundary for dynamic trajectory tracking from the spatiotemporal constraint model; The tracking cooperation boundary determines the trajectory interaction period of the logistics object when tracking the trajectory on the transportation route.

5. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: The specific information on the flow path during the operation of the logistics network includes: Collect waybill status information between transportation nodes in real time; Determine the core circulation path in the logistics process according to the waybill status information; The core circulation path is used to determine the circulation path information during the operation of the logistics network.

6. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: The risk characteristics of the circulation path information are analyzed to obtain the logistics node risk data of the logistics object during transportation, which specifically includes: Based on the circulation path information, a spatiotemporal graph association model of the dynamic behavior of logistics nodes and historical risk events is constructed; Output the risk coupling intensity of the logistics object during transportation through the spatiotemporal correlation model; Determine the risk impact characteristics of each logistics node in the transportation network through the risk coupling intensity; The logistics node risk data of the logistics object during transportation is determined according to the risk impact characteristics.

7. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: Determining the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information specifically includes: Determine risk association constraints of logistics during logistics transportation according to the logistics node risk data; Determine the abnormal coupling granularity of logistics in the logistics transportation process through the logistics transportation trajectory information; The risk situation level in the logistics transportation process is determined according to the risk association constraints and the abnormal coupling granularity.

8. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: The early warning of the dynamic logistics supervision blind spot in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level specifically includes: Determining the amount of elastic monitoring during the logistics intermodal transport process according to the trajectory interaction cycle; Determine the blind spot risk exposure index in the dynamic logistics supervision blind spot according to the risk situation level; The elastic monitoring quantity and the blind spot risk exposure index are matched with warning rules to generate a dynamically adjustable logistics supervision blind spot warning map.

9. The multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in claim 1 is characterized in that: Under the logistics information integration supervision mode, the chain record of cargo ownership change is compensated and checked, and the transportation attribute indicators obtained during the logistics transportation process include: Under the logistics information integration supervision mode, the ownership transfer path of the logistics transportation link is determined according to the chain record of cargo ownership change; Probabilistically compensate for the missing cargo ownership transfer nodes in the cargo ownership change chain record to generate a complete logistics event sequence; Performing feature aggregation on the logistics event sequence to obtain risk entropy value information in the logistics transportation process; The transport attribute index in the logistics transport process is determined by the risk entropy value information.

10. A multi-dimensional logistics information supervision and early warning system based on artificial intelligence, used to execute a multi-dimensional logistics information supervision and early warning method based on artificial intelligence as claimed in any one of claims 1 to 9, characterized in that: The regulatory early warning system includes: The data acquisition module is used to collect the chain records of cargo ownership changes in the logistics supply chain network from the distributed logistics nodes, and obtain the logistics transportation track information in the logistics supply chain network in real time; An association processing module is used to perform compensation inspection on the chain record of cargo ownership change under the logistics information fusion supervision mode, obtain the transportation attribute index in the logistics transportation process, and determine the trajectory interaction cycle of the logistics object when tracking the trajectory on the transportation route according to the transportation attribute index; A risk determination module is used to obtain the flow path information during the operation of the logistics network, analyze the risk characteristics of the flow path information, obtain the logistics node risk data of the logistics object during transportation, and determine the risk situation level in the logistics transportation process according to the logistics node risk data and the logistics transportation trajectory information; The risk warning module is used to warn the dynamic logistics supervision blind spots in the logistics intermodal transport process according to the trajectory interaction cycle and the risk situation level.

Images