A solid waste transfer risk identification method, system, device and medium

Abstract

This application relates to a method, system, equipment, and medium for identifying risks during solid waste transportation. The method includes: acquiring tracking data of solid waste during transportation and constructing a transportation process model based on the tracking data; identifying transportation risks using a pre-set risk indicator system based on the transportation process model; generating a risk accumulation equation using a risk transmission model based on the transportation risks; calculating the risk evolution trend using a pre-trained machine learning model based on the risk accumulation equation; and mapping the transportation process model, transportation risks, and risk evolution trend to a geographic information system to generate a transportation risk monitoring view as the risk identification result. This method enables dynamic identification, quantitative assessment, and visual monitoring of risks throughout the entire solid waste transportation chain, improving the accuracy and timeliness of risk management.

Description

Technical Field

[0001] This invention belongs to the field of solid waste disposal, and in particular relates to a method, system, equipment and medium for identifying risks during solid waste transportation. Background Technology

[0002] With the rapid development of technologies in the field of solid waste disposal, real-time tracking technology based on the Internet of Things (IoT) has emerged. This technology can automatically collect data such as location and status during the transportation process through sensors and GPS devices, realizing dynamic monitoring of the transport vehicle. The advantages of this technology are high efficiency in data acquisition and wide coverage. Current methods for identifying risks in solid waste transportation mainly rely on manual inspections and basic information systems. Risk management typically employs simple statistical reports or threshold alarm mechanisms, such as manually recording the time and location of transportation nodes and combining historical experience for post-event analysis of abnormal events. However, this approach lacks systematic modeling of risks across the entire supply chain.

[0003] The current approach has the following problems: low accuracy in risk identification, as it relies on decentralized manual judgment and is difficult to integrate multi-source data (such as solid waste composition, carrier state, etc.), leading to missed risk factors; poor timeliness, as it cannot dynamically capture the risk transmission path and the response is delayed; and insufficient visualization, as risk information is presented in the form of tables or text and lacks spatial mapping, which restricts the efficiency of real-time decision-making in risk monitoring. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, system, equipment, and medium for identifying risks in solid waste transportation that can solve the above problems.

[0005] Firstly, this application provides a method for identifying risks associated with the transfer of solid waste, including:

[0006] Acquire tracking data of solid waste during the transfer process, and construct a transfer process model based on the tracking data;

[0007] Based on the transshipment process model, a pre-set risk indicator system is used to identify transshipment risks;

[0008] Based on the risk of transshipment, a risk transmission model is used to generate a risk accumulation equation.

[0009] Based on the risk accumulation equation, a pre-trained machine learning model is used to calculate the risk evolution trend;

[0010] The transshipment process model, transshipment risks, and risk evolution trends are mapped to a geographic information system to generate a transshipment risk monitoring view as a risk identification result.

[0011] In one embodiment, the tracking data includes basic solid waste data, transfer process node data, transport vehicle status data, and handover and disposal data.

[0012] Based on tracking data, a transshipment process model is constructed, including:

[0013] Based on solid waste basic data, material property parameters are extracted and integrated to obtain solid waste component attribute identifiers;

[0014] Based on the node data of the transfer process, the spatiotemporal information of the nodes is extracted, and the transfer process path network is constructed based on the spatiotemporal information of the nodes.

[0015] Based on the transport vehicle status data, the characteristics of vehicle status change are extracted, and a vehicle monitoring model is constructed based on the characteristics of vehicle status change.

[0016] Based on the handover and disposal data, the authorization and authentication information and disposal status identifiers are extracted. Based on the authorization and authentication information and disposal status identifiers, correlation analysis is performed to generate a multi-party collaborative disposal verification table.

[0017] Based on solid waste component attribute identification, transfer process path network, carrier monitoring model and multi-party collaborative disposal verification table, a Bayesian network is used to establish probabilistic dependencies and obtain the transfer process model.

[0018] In one embodiment, the risk indicator system includes solid waste component risk level indicators, solid waste diffusion path indicators, carrier state risk indicators, and multi-party collaborative risk indicators.

[0019] Based on the transshipment process model, a pre-set risk indicator system is used to identify transshipment risks, including:

[0020] Based on the solid waste component attribute identification, a risk assessment matrix is ​​constructed using the analytic hierarchy process (AHP) according to the component risk level index.

[0021] Based on the transfer process path network, a node state transition probability matrix is ​​constructed according to the solid waste diffusion path index.

[0022] Based on the carrier monitoring model, the grey relational analysis method is used to calculate the state change threshold according to the carrier state risk index, and a carrier risk feature vector is generated based on the state change threshold.

[0023] Based on the multi-party collaborative handling verification table, the risk weight of the responsible party is determined according to the multi-party collaborative risk indicators, and a collaborative handling failure probability tree is constructed according to the risk weight of the responsible party.

[0024] Based on probabilistic dependencies, the risk assessment matrix, node state transition probability matrix, carrier risk feature vector, and collaborative disposal failure probability tree are quantitatively correlated to obtain a joint probability model.

[0025] Based on the joint probability model, the risk values ​​of the risk factors corresponding to each risk indicator are calculated to obtain the transfer risk.

[0026] In one embodiment, the risk transmission model includes a risk factor topology, a risk system dynamics rule base, and a historical risk transmission dataset;

[0027] Based on transshipment risk, a risk accumulation equation is generated using a risk transmission model, including:

[0028] Based on the risk factor topology and combined with the risk values ​​of the risk factors, a risk transmission path network is constructed.

[0029] Based on the risk system dynamics rule base, the risk transmission path network is mapped into a system of nonlinear differential equations;

[0030] Based on the historical risk transmission dataset, a system of nonlinear differential equations was fitted to obtain the equation coefficient matrix;

[0031] A family of spatiotemporal risk attenuation functions is constructed by spatially overlaying the coefficient matrix with preset geographic raster data.

[0032] By integrating the nonlinear differential equation system and the spatiotemporal risk decay function family, a risk accumulation equation is obtained.

[0033] In one embodiment, the machine learning model is a long short-term memory network model, and the training set of the long short-term memory network model is a historical risk transmission dataset;

[0034] Based on the risk accumulation equation, a pre-trained machine learning model is used to calculate the risk evolution trend, including:

[0035] Based on the risk accumulation equation, the spatiotemporal joint probability distribution features are extracted using the finite difference method;

[0036] Based on the spatiotemporal joint probability distribution characteristics and geographic raster data, a spatiotemporal adaptation feature set is obtained by dimensional alignment through coordinate system transformation.

[0037] A multi-head attention mechanism is used to calculate the risk transmission weight of each spatiotemporal node in the spatiotemporal adaptation feature set, and the spatiotemporal adaptation feature set is weighted and fused based on the risk transmission weight to obtain a weighted risk transmission feature matrix;

[0038] The weighted risk transmission feature matrix is ​​input into the long short-term memory network model to generate the probability distribution of risk propagation paths.

[0039] Based on the probability distribution of risk propagation paths, the confidence interval is calculated using the Monte Carlo simulation method;

[0040] By integrating the probability distribution and confidence interval of risk propagation paths, the risk evolution trend can be obtained.

[0041] In one embodiment, the coefficient matrix is ​​spatially overlaid with preset geographic raster data to construct a family of spatiotemporal risk attenuation functions, calculated as follows:

[0042]

[0043] in, Coordinates in a geographic raster The risk decay intensity at point and time t, where A is the coefficient matrix. coordinates The basic attribute quantization value of the geographic raster. This is the Hadamard product operator for matrices. This serves as the baseline coefficient for risk time decay. This is the risk decay index. Here are the initial coordinates of the risk source, and r is the geographic raster resolution. t represents the risk space decay index, and t represents the risk transmission time variable.

[0044] In one embodiment, the transshipment process model, transshipment risks, and risk evolution trends are mapped to a geographic information system to generate a transshipment risk monitoring view as a risk identification result, including:

[0045] Based on the spatiotemporal information of nodes in the transit process model, a geographic raster index is encoded.

[0046] Based on the geographic raster index and combined with the risk factor topology of transit risk, the risk value is mapped to the corresponding geographic raster cell to obtain the risk distribution raster map;

[0047] Based on the risk distribution raster map, and combined with the spatiotemporal decay function family of the risk accumulation equation, the spatial weight matrix of risk diffusion is calculated.

[0048] Based on the spatial weight matrix and combined with the probability distribution of risk propagation paths in the risk evolution trend, a risk propagation heat map is constructed.

[0049] Based on the authorization authentication information and disposal status identifier in the multi-party collaborative disposal verification form, a responsible entity label is generated, which includes the jurisdiction, jurisdiction, and solid waste disposal status.

[0050] By associating the solid waste component attribute identifiers with risk level indicators using color, a solid waste risk level color code layer is obtained.

[0051] By overlaying the risk propagation heat map, the responsible entity label, and the solid waste risk level color code layer in time and space, a transfer risk monitoring view is generated.

[0052] Secondly, this application also provides a solid waste transfer risk identification system, including:

[0053] The transfer model building module is used to acquire tracking data of solid waste during the transfer process and build a transfer process model based on the tracking data.

[0054] The transshipment risk identification module is used to identify transshipment risks based on the transshipment process model and a preset risk indicator system.

[0055] The risk accumulation equation module is used to generate risk accumulation equations based on transit risk using a risk transmission model.

[0056] The risk evolution trend module is used to calculate the risk evolution trend based on the risk accumulation equation and a pre-trained machine learning model.

[0057] The risk monitoring view module is used to map the transshipment process model, transshipment risks, and risk evolution trends to the geographic information system, generating a transshipment risk monitoring view as a risk identification result.

[0058] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method for identifying risks in solid waste transportation.

[0059] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method for identifying risks in solid waste transfer.

[0060] The aforementioned method, system, equipment, and medium for identifying risks during solid waste transfer acquire tracking data of solid waste during transfer and construct a transfer process model. It integrates multi-source information such as solid waste composition and carrier status, overcoming the integration difficulties caused by fragmented manual judgment and improving the accuracy of risk identification. Based on the transfer process model, it uses a pre-set risk indicator system to identify transfer risks, and enhances the comprehensiveness of the assessment by systematically evaluating factors such as solid waste composition and diffusion paths. Based on the identified transfer risks, it uses a risk transmission model to generate a risk accumulation equation, dynamically models the risk transmission path, and achieves real-time capture of risk changes. Based on the risk accumulation equation, it uses a pre-trained machine learning model to calculate the risk evolution trend, predict the probability of future risk propagation, and support early warning and intervention. Finally, it maps the transfer process model, transfer risks, and risk evolution trends to a geographic information system to generate a transfer risk monitoring view, presenting the overall risk picture in a spatial visualization manner and improving the efficiency of monitoring and decision-making. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart of a solid waste transfer risk identification method according to the present invention;

[0063] Figure 2 This is a structural diagram of a solid waste transfer risk identification system according to the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0065] In one embodiment, such as Figure 1 As shown, a method for identifying risks during solid waste transportation is provided. This embodiment illustrates the application of this method to a terminal, but it is understood that the method can also be applied to a server, or to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In the implementation environment, the hardware architecture mainly consists of terminal devices (such as mobile handheld terminals, vehicle-mounted IoT sensors, etc.), a server, and a Geographic Information System (GIS) platform. The terminal devices collect multi-source tracking data during the solid waste transportation process in real time via wireless networks (such as 5G or LoRa), including basic solid waste attributes, carrier GPS location, and status sensor readings, and upload them to the server. After receiving the data, the server uses built-in transportation model construction modules and risk identification modules to perform dynamic calculations based on Bayesian networks and machine learning models, generating risk accumulation equations and evolution trends. The results are then integrated into a visual view through the GIS platform by integrating spatial data. Application scenarios include: when there is a need for risk monitoring of the entire solid waste transfer chain, such as the need for real-time early warning of leakage or delay risks during cross-regional transportation, the terminal and server achieve the following interaction: the terminal collects and transmits data, the server executes risk identification algorithms, and feeds back the risk heat map and information on the responsible party to the terminal display screen, enabling operators to intervene in a timely manner and improve the accuracy and timeliness of management.

[0066] In this embodiment, the method includes the following steps:

[0067] S01, acquire tracking data of solid waste during the transfer process, and build a transfer process model based on the tracking data.

[0068] The tracking data comprises real-time solid waste transfer information collected via IoT sensors, GPS positioning devices, RFID tags, and other means. This includes the physicochemical properties of the solid waste, the spatiotemporal coordinates of transfer nodes, the operational status of the transport vehicles, and electronic records of handover and disposal. This data is transmitted to the processing unit via a communication network for cleaning, normalization, and feature extraction to construct a transfer process model. During the model's construction, data mining algorithms can be used to extract key parameters (such as material composition identification and path topology), and probabilistic graphical models (such as Bayesian networks) or system dynamics methods can be employed to establish dependencies between multiple factors, generating a digital twin model capable of simulating the transfer process. This model not only covers the solid waste's movement path and carrier behavior but also incorporates environmental factors and human-manipulated variables to support subsequent quantitative analysis of risk identification, providing a data foundation for risk management.

[0069] S02, based on the transshipment process model, uses a preset risk indicator system to identify transshipment risks.

[0070] The preset risk indicator system is a set of evaluation standards covering multiple dimensions such as solid waste composition, diffusion path, carrier state, and multi-party collaboration. It can include quantitative parameters such as risk level weights and state change thresholds, and its settings can be adaptively adjusted based on historical data or expert knowledge. In implementation, key features (such as node state probabilities and carrier behavior patterns) can be extracted based on the transfer process model. This risk indicator system can then be used for multi-factor correlation analysis, such as constructing a risk assessment matrix using the analytic hierarchy process, calculating feature vectors using the grey relational analysis method, or integrating various indicators using a joint probability model to output risk values, thereby identifying various possible risks during the transfer process.

[0071] S03, based on transshipment risk, uses a risk transmission model to generate a risk accumulation equation.

[0072] The risk transmission model is a framework for simulating the propagation mechanism of risk within a system. It includes a network model with a topological structure, system dynamics equations, or data-driven machine learning methods. Its components, such as risk factor association rules and historical transmission patterns, characterize the diffusion behavior of risk over time and space. The risk accumulation equation is a mathematical expression used to quantify the superposition of risks at specific locations and times. In implementation, a transmission path network can be constructed based on the factor dependencies in the transported risk, mapped to dynamic equations (such as a system of nonlinear differential equations), and parameters can be fitted using historical datasets to calibrate the model. Geospatial information (such as raster data) can be integrated, and a family of decay functions can be introduced to simulate the attenuation effect of risk with distance and time, generating a comprehensive accumulation equation. This achieves the computability and predictability of risk accumulation.

[0073] S04, based on the risk accumulation equation, uses a pre-trained machine learning model to calculate the risk evolution trend.

[0074] The pre-trained machine learning model is a computational model pre-trained based on historical datasets (such as risk transmission records). It can include Long Short-Term Memory (LSTM) networks, graph neural networks, or attention mechanism models, used to learn risk evolution patterns from the data. The risk evolution trend refers to the outputs such as the probability distribution and confidence interval of risk propagation, used to predict future risk changes. In implementation, spatiotemporal features can be extracted based on the risk accumulation equation, and after dimensional alignment, they can be input into the machine learning model. Its sequence processing capabilities can be used to generate the probability distribution of risk propagation paths, and Monte Carlo simulation can be combined to calculate the uncertainty range, integrating them into the evolution trend.

[0075] S05 maps the transshipment process model, transshipment risks, and risk evolution trends to a geographic information system to generate a transshipment risk monitoring view as the risk identification result.

[0076] Geographic Information System (GIS) is a software platform used to store, analyze, and display geographic reference data, and can include open-source or commercial systems such as ArcGIS. The transit risk monitoring view is a comprehensive visualization output, such as a heatmap or raster layer, used to dynamically present the spatial distribution of risks. In implementation, the geographic raster index is encoded based on the spatiotemporal information of nodes in the transit process model, mapping risk values ​​to corresponding raster cells to generate a risk distribution map. The spatial weight matrix is ​​calculated using the decay function of the risk accumulation equation, and a heatmap is constructed using the probability distribution of risk evolution trends. By integrating the responsible entity labeling and risk level color-coded layers, the final view is generated through spatiotemporal overlay, overcoming the limitations of traditional text-based reports, achieving dynamic spatial monitoring of risks across the entire chain, and improving management efficiency and decision-making accuracy.

[0077] In one embodiment, the tracking data includes basic solid waste data, transfer process node data, transport vehicle status data, and handover and disposal data.

[0078] Based on tracking data, a transshipment process model is constructed, including:

[0079] S11. Based on the basic data of solid waste, extract the material property parameters and integrate the material property parameters to obtain the solid waste component attribute identifier;

[0080] S12, Based on the data of the transfer process nodes, extract the spatiotemporal information of the nodes, and construct the transfer process path network based on the spatiotemporal information of the nodes;

[0081] S13. Based on the transport vehicle status data, extract the characteristics of vehicle status changes, and construct a vehicle monitoring model based on the characteristics of vehicle status changes.

[0082] S14. Based on the handover and disposal data, extract the authorization authentication information and disposal status identifier, and perform correlation analysis based on the authorization authentication information and disposal status identifier to generate a multi-party collaborative disposal verification table.

[0083] S15. Based on the solid waste component attribute identification, transfer process path network, carrier monitoring model and multi-party collaborative disposal verification table, a Bayesian network is used to establish probabilistic dependencies to obtain the transfer process model.

[0084] For example, the tracking data includes four core data categories: solid waste basic data, transfer process node data, transport vehicle status data, and handover and disposal data. Layered modeling operations can be performed based on this multi-source tracking data. Physical and chemical properties of solid waste are extracted from the solid waste basic data. After data cleaning and normalization, these parameters are integrated and coded to form unique solid waste component attribute identifiers, completing the digital representation of solid waste components. Geographic coordinates, timestamps, and other spatiotemporal information of each transfer node are extracted from the transfer process node data. Topological modeling is performed based on the flow relationships between nodes to construct a transfer process path network capable of reconstructing the spatial trajectory and time sequence of solid waste transfer. Carrier status change characteristics such as carrier operating speed, load capacity, and equipment operating parameters are extracted from the transport vehicle status data. Through time series analysis and feature modeling, a carrier monitoring model capable of capturing real-time changes in carrier operating status is constructed. Authorization and authentication information for each operational step is extracted from the handover and disposal data. The system identifies the status of each stage of solid waste disposal, verifies and matches the correlation between two types of information, and generates a multi-party collaborative disposal verification table that records the operating permissions, disposal links, and corresponding statuses of each participant, clarifying the rights, responsibilities, and operating statuses of each entity involved in collaborative disposal. Using solid waste component attribute identifiers, transfer process path networks, carrier monitoring models, and the multi-party collaborative disposal verification table as core inputs, a Bayesian network algorithm is employed to mine and model the inherent probabilistic dependencies between various elements, generating a transfer process model that integrates solid waste attributes, transfer paths, carrier status, and collaborative disposal information across all dimensions. This achieves a digital and systematic mapping of the entire solid waste transfer process. This model reflects the correlation and operational status of each element in the transfer process, providing a reliable data foundation for subsequent risk identification.

[0085] In one embodiment, the risk indicator system includes solid waste component risk level indicators, solid waste diffusion path indicators, carrier state risk indicators, and multi-party collaborative risk indicators.

[0086] Based on the transshipment process model, a pre-set risk indicator system is used to identify transshipment risks, including:

[0087] S21. Based on the solid waste component attribute identification, a risk assessment matrix is ​​constructed using the analytic hierarchy process according to the component risk level index.

[0088] S22, based on the transfer process path network, constructs a node state transition probability matrix according to solid waste diffusion path indicators;

[0089] S23. Based on the carrier monitoring model, the grey relational analysis method is used to calculate the state change threshold according to the carrier state risk index, and a carrier risk feature vector is generated based on the state change threshold.

[0090] S24. Based on the multi-party collaborative handling verification table, determine the risk weight of the responsible party according to the multi-party collaborative risk indicators, and construct a collaborative handling failure probability tree according to the risk weight of the responsible party.

[0091] S25. Based on probabilistic dependencies, the risk assessment matrix, node state transition probability matrix, carrier risk feature vector, and collaborative disposal failure probability tree are quantitatively correlated to obtain a joint probability model.

[0092] S26. Based on the joint probability model, calculate the risk value of the risk factor corresponding to each risk indicator to obtain the transfer risk.

[0093] Specifically, identifying transfer risks can be implemented step-by-step based on a pre-defined risk indicator system encompassing four categories: solid waste composition, diffusion path, carrier state, and multi-party collaboration. This involves: constructing dedicated quantitative models for different dimensions of risk characteristics, and achieving integrated risk calculation across all dimensions through probabilistic correlation. Based on solid waste component attribute identification and combined with component risk level indicators, an analytic hierarchy process (AHP) can be used to construct an evaluation matrix characterizing solid waste component risk through indicator stratification, weight assignment, and consistency checks. According to the node flow relationships in the transfer process network, solid waste diffusion path indicators are matched, and the probability values ​​of state transitions at each node are statistically analyzed to construct a node state transition probability matrix. Based on real-time status data from the carrier monitoring model and referring to carrier state risk indicators, grey relational analysis can be used to calculate the correlation between the carrier state sequence and the risk sequence, determine the state mutation threshold, and extract risk features to generate a carrier risk feature vector. Combining the rights and responsibilities and status information in the multi-party collaborative disposal verification table, the responsible entities are divided into levels according to the multi-party collaborative risk indicators and assigned corresponding risk weights. The failure transmission relationships among the entities are then analyzed based on these weights to construct a collaborative disposal failure probability tree. Based on the probabilistic dependencies established by the Bayesian network in the transfer process model, the risk assessment matrix, node state transition probability matrix, carrier risk feature vector, and collaborative disposal failure probability tree are dimensionally matched and quantitatively correlated. The inherent correlation features of risk factors in each dimension are integrated to obtain a joint probability model. Through this model, the risk factors corresponding to each risk indicator are probabilistically calculated, and the specific risk values ​​of each factor are output. After integration, the transfer risk of the solid waste transfer process can be obtained.

[0094] In one embodiment, the risk transmission model includes a risk factor topology, a risk system dynamics rule base, and a historical risk transmission dataset;

[0095] Based on transshipment risk, a risk accumulation equation is generated using a risk transmission model, including:

[0096] S31, based on the risk factor topology and combined with the risk value of the risk factor, constructs a risk transmission path network;

[0097] S32, based on the risk system dynamics rule base, maps the risk transmission path network into a system of nonlinear differential equations;

[0098] S33, based on the historical risk transmission dataset, fit a system of nonlinear differential equations to obtain the equation coefficient matrix;

[0099] S34, spatially overlay the coefficient matrix with the preset geographic raster data to construct a family of spatiotemporal risk attenuation functions;

[0100] S35 integrates the nonlinear differential equation system and the spatiotemporal risk decay function family to obtain the risk accumulation equation.

[0101] For example, a risk transmission model comprising a risk factor topology, a risk system dynamics rule base, and a historical risk transmission dataset can be adopted. Based on identified transit risks, a risk accumulation equation can be constructed to achieve spatiotemporal quantitative modeling of risk transmission: using the risk factor topology as the basic framework, and combining the specific risk values ​​corresponding to each risk factor, the transmission correlations and hierarchical dependencies between factors are analyzed to construct a risk transmission path network that reflects the actual propagation patterns of risk; according to the pre-set risk transmission dynamics laws and quantification criteria in the risk system dynamics rule base, the topological relationships of the risk transmission path network are transformed into a set of nonlinear differential equations that can characterize the dynamic changes of risk, achieving a mathematical mapping of the network model; using the historical risk transmission dataset as a fitting sample, the parameters of the nonlinear differential equations are solved and calibrated through a numerical fitting algorithm to obtain the equation coefficient matrix. This coefficient matrix is ​​then spatially overlaid with pre-set geographic raster data, and combined with basic geographic spatial attributes, a family of spatiotemporal risk decay functions reflecting the time and spatial decay patterns of risk is constructed. By integrating the nonlinear differential equation system with the spatiotemporal risk attenuation function family, and incorporating the dynamic transmission characteristics and spatiotemporal attenuation characteristics of risk into the same mathematical system, a risk accumulation equation that can quantify the degree of risk accumulation under different spatiotemporal dimensions can be obtained.

[0102] In one embodiment, the machine learning model is a long short-term memory network model, and the training set of the long short-term memory network model is a historical risk transmission dataset;

[0103] Based on the risk accumulation equation, a pre-trained machine learning model is used to calculate the risk evolution trend, including:

[0104] S41, based on the risk accumulation equation, uses the finite difference method to extract the spatiotemporal joint probability distribution characteristics;

[0105] S42, based on spatiotemporal joint probability distribution characteristics and geographic raster data, uses coordinate system transformation to perform dimensional alignment to obtain a spatiotemporal adaptation feature set;

[0106] S43, a multi-head attention mechanism is used to calculate the risk transmission weight of each spatiotemporal node in the spatiotemporal adaptation feature set, and the spatiotemporal adaptation feature set is weighted and fused based on the risk transmission weight to obtain a weighted risk transmission feature matrix;

[0107] S44, input the weighted risk transmission feature matrix into the long short-term memory network model to generate the probability distribution of risk propagation paths;

[0108] S45. Based on the probability distribution of risk propagation paths, the confidence interval is calculated using the Monte Carlo simulation method.

[0109] S46 integrates the probability distribution and confidence interval of risk propagation paths to obtain the risk evolution trend.

[0110] Specifically, a pre-trained Long Short-Term Memory (LSTM) network model is used as the computational model. This model is trained on a historical risk transmission dataset and quantifies the risk evolution trend based on the risk accumulation equation. Based on the mathematical characteristics of the risk accumulation equation, the equation can be discretized using the finite difference method to extract spatiotemporal joint probability distribution features that reflect the spatiotemporal distribution patterns of risk. These features are then subjected to a coordinate system transformation with pre-defined geographic raster data to achieve precise spatiotemporal alignment. After eliminating dimensional bias, the spatiotemporal adaptation feature set is obtained. A multi-head attention mechanism is used to calculate the correlation between spatiotemporal nodes in the spatiotemporal adaptation feature set, obtaining the risk transmission weights corresponding to each node. Based on these weights, the spatiotemporal adaptation feature set is weighted and fused to generate a weighted risk transmission feature matrix that highlights key risk nodes. This matrix is ​​then input into the pre-trained LSM network model, utilizing the model's temporal feature learning capability to simulate the risk propagation process and generate a risk propagation path probability distribution. Based on this probability distribution, Monte Carlo simulation can be used to perform multiple random sampling calculations to obtain the confidence interval of risk propagation. By integrating the probability distribution of risk propagation paths with the confidence interval, a risk evolution trend that combines the laws of risk propagation with the range of probability fluctuations can be formed, thereby achieving accurate prediction of the future development of risks.

[0111] In one embodiment, S51, the coefficient matrix is ​​spatially overlaid with preset geographic raster data to construct a family of spatiotemporal risk attenuation functions, the calculation formula of which is as follows:

[0112]

[0113] in, Coordinates in a geographic raster The risk decay intensity at point and time t, where A is the coefficient matrix. coordinates The basic attribute quantization value of the geographic raster. This is the Hadamard product operator for matrices. This serves as the baseline coefficient for risk time decay. This is the risk decay index. Here are the initial coordinates of the risk source, and r is the geographic raster resolution. t represents the risk space decay index, and t represents the risk transmission time variable.

[0114] For example, this calculation formula is the core quantitative formula for constructing a family of spatiotemporal risk attenuation functions by spatially overlaying the coefficient matrix with preset geographic raster data. Its purpose is to calculate any coordinate in the geographic raster. Risk decay intensity at any point and at any time t This provides a specific numerical basis for the spatiotemporal risk attenuation function family, enabling a mathematical representation of the attenuation law of risk in the time and space dimensions. The formula as a whole has a quantified relationship with each variable. First, the coefficient matrix A of the equation is connected to the coordinates using the matrix Hadamard product operator. Quantized values ​​of basic attributes of geographic raster Element-wise multiplication yields the basic quantitative value of risk decay intensity. An exponential function term then comprehensively reflects the dual decay effect of time and space, where λ serves as the baseline coefficient for risk time decay, α as the risk time decay exponent, and together with the time variable t, determine the rate of risk decay over time. As the initial coordinates of the risk source, combined with the raster resolution r, the risk space decay index β, and the coordinates of the point to be calculated... This study quantifies the degree of risk attenuation with increasing spatial distance using a spatial distance formula. The exponential function as a whole decreases with increasing t and increasing spatial distance between the point to be calculated and the risk source. The risk attenuation intensity at each spatiotemporal point calculated by this formula constitutes a family of spatiotemporal risk attenuation functions. This family of functions is integrated with a system of nonlinear differential equations to form a complete risk accumulation equation. Subsequently, the spatial weight matrix of risk diffusion can be calculated based on this family of functions in conjunction with a risk distribution raster map, providing spatiotemporal attenuation quantification support for the construction of risk propagation heatmaps and the accurate calculation of risk evolution trends.

[0115] In one embodiment, the transshipment process model, transshipment risks, and risk evolution trends are mapped to a geographic information system to generate a transshipment risk monitoring view as a risk identification result, including:

[0116] S61, based on the spatiotemporal information of nodes in the transit process model, encodes a geographic raster index;

[0117] S62, based on the geographic raster index and combined with the risk factor topology of transit risk, maps the risk value to the corresponding geographic raster cell to obtain the risk distribution raster map;

[0118] S63, based on the risk distribution grid map and combined with the spatiotemporal decay function family of the risk accumulation equation, calculate the spatial weight matrix of risk diffusion;

[0119] S64, based on the spatial weight matrix and combined with the probability distribution of risk propagation paths in the risk evolution trend, constructs a risk propagation heat map;

[0120] S65, based on the authorization authentication information and disposal status identifier in the multi-party collaborative disposal verification table, generates a responsible entity label, which includes the jurisdiction area, jurisdiction authority and solid waste disposal status;

[0121] S66, associate solid waste component attribute identifiers with risk level indicators by color to obtain solid waste risk level color code layer;

[0122] S67, by spatiotemporally overlaying the risk propagation heat map, the labeling of responsible entities, and the color-coded layer of solid waste risk level, generates a transfer risk monitoring view.

[0123] Specifically, a transportation risk monitoring view can be generated through multi-step spatialization processing and layer fusion, serving as the final risk identification result: Based on the spatiotemporal information of nodes in the transportation process model, geographic raster indexing is performed to achieve a one-to-one match between transportation nodes and geographic raster units. Based on this geographic raster index, combined with the risk factor topology of transportation risks, the specific risk values ​​of each risk factor are mapped to the corresponding geographic raster units according to spatial association rules. A risk distribution raster map is obtained through quantized rendering of the raster data. Based on the risk distribution raster map, and combined with the spatiotemporal decay function family of the risk accumulation equation, the risk diffusion impact weight of each raster unit is calculated using a spatial interpolation algorithm, resulting in a spatial weight matrix for risk diffusion. Based on this matrix and combined with the probability distribution of risk propagation paths in the risk evolution trend, kernel density analysis is used to construct a risk propagation heatmap that intuitively reflects the spatiotemporal propagation trend of risks. Authorization information and disposal status identifiers are extracted from the multi-party collaborative disposal verification table. Responsible entity labels containing jurisdictional area, jurisdictional authority, and solid waste disposal status are generated according to geographic spatial affiliation. Simultaneously, solid waste component attribute identifiers and risk level indicators can be correlated with gradient colors to generate a visual solid waste risk level color-coded layer. In the geographic information system, the risk propagation heat map, the labeling of responsible entities, and the color-coded layer of solid waste risk level are spatiotemporally registered and overlaid to achieve an integrated presentation of multi-dimensional risk information and generate a transfer risk monitoring view.

[0124] The aforementioned method for identifying risks in solid waste transportation integrates multi-source data, including solid waste composition and carrier state, by acquiring solid waste transportation tracking data and constructing a transportation process model. This overcomes the integration difficulties of traditional, fragmented manual judgment, avoids missed risk factors, and improves the accuracy of risk identification. Based on the transportation process model, a pre-set risk indicator system is used to identify transportation risks. By systematically evaluating risk factors across all dimensions of the entire chain, precise quantitative judgment of risks is achieved, compensating for the one-sidedness of traditional risk identification methods. Based on transportation risks, a risk transmission model is used to generate a risk accumulation equation, dynamically modeling the transmission path and accumulation law of risks. This enables real-time capture of dynamic changes in risks, solving the problems of traditional methods. This addresses the issue of delayed response and improves the timeliness of risk management. Based on the risk accumulation equation, a pre-trained machine learning model is used to calculate risk evolution trends, predicting future risk propagation and development, enabling early warning and proactive intervention, and strengthening the timeliness of risk control. Mapping the transfer process model, transfer risks, and risk evolution trends to a geographic information system generates monitoring views, achieving spatial visualization of risk information. This breaks free from the limitations of traditional tables and text formats, integrates spatial correlation information of risks across the entire chain, significantly improves the real-time decision-making efficiency of risk monitoring, and comprehensively solves the technical problems of low accuracy, poor timeliness, and insufficient visualization in traditional solid waste transfer risk identification.

[0125] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0126] Based on the same inventive concept, this application also provides a solid waste transfer risk identification system for implementing the aforementioned solid waste transfer risk identification method. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations of one or more embodiments of the solid waste transfer risk identification system provided below can be found in the limitations of the solid waste transfer risk identification method described above, and will not be repeated here.

[0127] In one exemplary embodiment, such as Figure 2 As shown, a solid waste transfer risk identification system is provided, including:

[0128] The transfer model construction module 101 is used to acquire tracking data of solid waste during the transfer process and to construct a transfer process model based on the tracking data.

[0129] The transshipment risk identification module 102 is used to identify transshipment risks based on the transshipment process model and a preset risk indicator system.

[0130] Risk accumulation equation module 103 is used to generate risk accumulation equations based on transfer risk using a risk transmission model;

[0131] The risk evolution trend module 104 is used to calculate the risk evolution trend based on the risk accumulation equation and a pre-trained machine learning model.

[0132] The risk monitoring view module 105 is used to map the transshipment process model, transshipment risks and risk evolution trends to the geographic information system to generate a transshipment risk monitoring view as a risk identification result.

[0133] In one embodiment, the tracking data in the transfer model construction module 101 includes solid waste basic data, transfer process node data, transport vehicle status data, and handover and disposal data.

[0134] The transport model building module 101 is also used for:

[0135] Based on solid waste basic data, material property parameters are extracted and integrated to obtain solid waste component attribute identifiers;

[0136] Based on the node data of the transfer process, the spatiotemporal information of the nodes is extracted, and the transfer process path network is constructed based on the spatiotemporal information of the nodes.

[0137] Based on the transport vehicle status data, the characteristics of vehicle status change are extracted, and a vehicle monitoring model is constructed based on the characteristics of vehicle status change.

[0138] Based on the handover and disposal data, the authorization and authentication information and disposal status identifiers are extracted. Based on the authorization and authentication information and disposal status identifiers, correlation analysis is performed to generate a multi-party collaborative disposal verification table.

[0139] Based on solid waste component attribute identification, transfer process path network, carrier monitoring model and multi-party collaborative disposal verification table, a Bayesian network is used to establish probabilistic dependencies and obtain the transfer process model.

[0140] In one embodiment, the risk indicator system in the transfer risk identification module 102 includes solid waste component risk level indicators, solid waste diffusion path indicators, carrier state risk indicators, and multi-party collaborative risk indicators.

[0141] The transshipment risk identification module 102 is also used for:

[0142] Based on the solid waste component attribute identification, a risk assessment matrix is ​​constructed using the analytic hierarchy process (AHP) according to the component risk level index.

[0143] Based on the transfer process path network, a node state transition probability matrix is ​​constructed according to the solid waste diffusion path index.

[0144] Based on the carrier monitoring model, the grey relational analysis method is used to calculate the state change threshold according to the carrier state risk index, and a carrier risk feature vector is generated based on the state change threshold.

[0145] Based on the multi-party collaborative handling verification table, the risk weight of the responsible party is determined according to the multi-party collaborative risk indicators, and a collaborative handling failure probability tree is constructed according to the risk weight of the responsible party.

[0146] Based on probabilistic dependencies, the risk assessment matrix, node state transition probability matrix, carrier risk feature vector, and collaborative disposal failure probability tree are quantitatively correlated to obtain a joint probability model.

[0147] Based on the joint probability model, the risk values ​​of the risk factors corresponding to each risk indicator are calculated to obtain the transfer risk.

[0148] In one embodiment, the risk transmission model in the risk accumulation equation module 103 includes a risk factor topology, a risk system dynamics rule base, and a historical risk transmission dataset.

[0149] Risk accumulation equation module 103 is also used for:

[0150] Based on the risk factor topology and combined with the risk values ​​of the risk factors, a risk transmission path network is constructed.

[0151] Based on the risk system dynamics rule base, the risk transmission path network is mapped into a system of nonlinear differential equations;

[0152] Based on the historical risk transmission dataset, a system of nonlinear differential equations was fitted to obtain the equation coefficient matrix;

[0153] A family of spatiotemporal risk attenuation functions is constructed by spatially overlaying the coefficient matrix with preset geographic raster data.

[0154] By integrating the nonlinear differential equation system and the spatiotemporal risk decay function family, a risk accumulation equation is obtained.

[0155] In one embodiment, the machine learning model in the risk evolution trend module 104 is a long short-term memory network model, and the training set of the long short-term memory network model is a historical risk transmission dataset.

[0156] Risk evolution trend module 104 is also used for:

[0157] Based on the risk accumulation equation, the spatiotemporal joint probability distribution features are extracted using the finite difference method;

[0158] Based on the spatiotemporal joint probability distribution characteristics and geographic raster data, a spatiotemporal adaptation feature set is obtained by dimensional alignment through coordinate system transformation.

[0159] A multi-head attention mechanism is used to calculate the risk transmission weight of each spatiotemporal node in the spatiotemporal adaptation feature set, and the spatiotemporal adaptation feature set is weighted and fused based on the risk transmission weight to obtain a weighted risk transmission feature matrix;

[0160] The weighted risk transmission feature matrix is ​​input into the long short-term memory network model to generate the probability distribution of risk propagation paths.

[0161] Based on the probability distribution of risk propagation paths, the confidence interval is calculated using the Monte Carlo simulation method;

[0162] By integrating the probability distribution and confidence interval of risk propagation paths, the risk evolution trend can be obtained.

[0163] In one embodiment, the risk accumulation equation module 103 performs spatial overlay operations on the coefficient matrix and preset geographic raster data to construct a family of spatiotemporal risk attenuation functions, with the following calculation formula:

[0164]

[0165] in, Coordinates in a geographic raster The risk decay intensity at point and time t, where A is the coefficient matrix. coordinates The basic attribute quantization value of the geographic raster. This is the Hadamard product operator for matrices. This serves as the baseline coefficient for risk time decay. This is the risk decay index. Here are the initial coordinates of the risk source, and r is the geographic raster resolution. t represents the risk space decay index, and t represents the risk transmission time variable.

[0166] In one embodiment, the risk monitoring view module 105 is further configured to:

[0167] Based on the spatiotemporal information of nodes in the transit process model, a geographic raster index is encoded.

[0168] Based on the geographic raster index and combined with the risk factor topology of transit risk, the risk value is mapped to the corresponding geographic raster cell to obtain the risk distribution raster map;

[0169] Based on the risk distribution raster map, and combined with the spatiotemporal decay function family of the risk accumulation equation, the spatial weight matrix of risk diffusion is calculated.

[0170] Based on the spatial weight matrix and combined with the probability distribution of risk propagation paths in the risk evolution trend, a risk propagation heat map is constructed.

[0171] Based on the authorization authentication information and disposal status identifier in the multi-party collaborative disposal verification form, a responsible entity label is generated, which includes the jurisdiction, jurisdiction, and solid waste disposal status.

[0172] By associating the solid waste component attribute identifiers with risk level indicators using color, a solid waste risk level color code layer is obtained.

[0173] By overlaying the risk propagation heat map, the responsible entity label, and the solid waste risk level color code layer in time and space, a transfer risk monitoring view is generated.

[0174] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the solid waste transfer risk identification method as described above.

[0175] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0176] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0177] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for identifying risks in solid waste transfer, characterized in that, The method includes: Acquire tracking data of solid waste during the transfer process, and construct a transfer process model based on the tracking data; Based on the aforementioned transshipment process model, a preset risk indicator system is used to identify transshipment risks; Based on the aforementioned transshipment risks, a risk accumulation equation is generated using a risk transmission model. Based on the aforementioned risk accumulation equation, a pre-trained machine learning model is used to calculate the risk evolution trend; The transshipment process model, the transshipment risks, and the risk evolution trends are mapped to a geographic information system to generate a transshipment risk monitoring view as the risk identification result.

2. The method according to claim 1, characterized in that, The tracking data includes basic solid waste data, transfer process node data, transport vehicle status data, and handover and disposal data. The step of constructing a transit process model based on the tracking data includes: Based on the solid waste basic data, material property parameters are extracted and integrated to obtain solid waste component attribute identifiers. Based on the data of the transfer process nodes, spatiotemporal information of the nodes is extracted, and a transfer process path network is constructed based on the spatiotemporal information of the nodes. Based on the transport vehicle status data, the carrier status change characteristics are extracted, and a carrier monitoring model is constructed based on the carrier status change characteristics. Based on the handover and disposal data, the authorization authentication information and disposal status identifier are extracted, and based on the authorization authentication information and disposal status identifier, a correlation analysis is performed to generate a multi-party collaborative disposal verification table. Based on the solid waste component attribute identifier, the transfer process path network, the carrier monitoring model, and the multi-party collaborative disposal verification table, a Bayesian network is used to establish probabilistic dependencies to obtain the transfer process model.

3. The method according to claim 2, characterized in that, The risk indicator system includes solid waste component risk level indicators, solid waste diffusion path indicators, carrier state risk indicators, and multi-party collaborative risk indicators. The method of identifying transportation risks based on the aforementioned transportation process model using a preset risk indicator system includes: Based on the solid waste component attribute identifiers, a risk assessment matrix is ​​constructed using the analytic hierarchy process (AHP) according to the component risk level indicators. Based on the aforementioned transfer process path network, and according to the solid waste diffusion path index, a node state transition probability matrix is ​​constructed. Based on the carrier monitoring model, according to the carrier state risk index, the grey relational analysis method is used to calculate the state change threshold, and the carrier risk feature vector is generated according to the state change threshold. Based on the multi-party collaborative handling verification table, the risk weight of the responsible party is determined according to the multi-party collaborative risk index, and a collaborative handling failure probability tree is constructed according to the risk weight of the responsible party. Based on the probabilistic dependencies, the risk assessment matrix, the node state transition probability matrix, the carrier risk feature vector, and the collaborative disposal failure probability tree are quantitatively correlated to obtain a joint probability model. Based on the joint probability model, the risk value of the risk factor corresponding to each risk indicator is calculated to obtain the transfer risk.

4. The method according to claim 3, characterized in that, The risk transmission model includes a risk factor topology, a risk system dynamics rule base, and a historical risk transmission dataset. The step of generating a risk accumulation equation based on the transshipment risk using a risk transmission model includes: Based on the risk factor topology and the risk value of the risk factor, a risk transmission path network is constructed. Based on the risk system dynamics rule base, the risk transmission path network is mapped into a system of nonlinear differential equations; Based on the historical risk transmission dataset, the nonlinear differential equation system is fitted to obtain the equation coefficient matrix; The coefficient matrix is ​​spatially superimposed with preset geographic raster data to construct a family of spatiotemporal risk attenuation functions. By integrating the nonlinear differential equations and the spatiotemporal risk decay function family, the risk accumulation equation is obtained.

5. The method according to claim 4, characterized in that, The machine learning model is a long short-term memory network model, and the training set of the long short-term memory network model is the historical risk transmission dataset; The calculation of risk evolution trends using a pre-trained machine learning model based on the risk accumulation equation includes: Based on the aforementioned risk accumulation equation, the spatiotemporal joint probability distribution features are extracted using the finite difference method; Based on the spatiotemporal joint probability distribution features and the geographic raster data, a spatiotemporal adaptation feature set is obtained by dimensional alignment through coordinate system transformation. A multi-head attention mechanism is used to calculate the risk transmission weight of each spatiotemporal node in the spatiotemporal adaptation feature set, and the spatiotemporal adaptation feature set is weighted and fused based on the risk transmission weight to obtain a weighted risk transmission feature matrix. The weighted risk transmission feature matrix is ​​input into the long short-term memory network model to generate a risk propagation path probability distribution. Based on the probability distribution of the risk propagation path, the confidence interval is calculated using the Monte Carlo simulation method; By integrating the probability distribution of the risk propagation path and the confidence interval, the risk evolution trend is obtained.

6. The method according to claim 4, characterized in that, The coefficient matrix is ​​spatially overlaid with preset geographic raster data to construct a family of spatiotemporal risk attenuation functions, and the calculation formula is as follows: in, Coordinates in a geographic raster The risk decay intensity at point and time t, where A is the coefficient matrix. coordinates The basic attribute quantization value of the geographic raster. This is the Hadamard product operator for matrices. This serves as the baseline coefficient for risk time decay. This is the risk decay index. Here are the initial coordinates of the risk source, and r is the geographic raster resolution. t represents the risk space decay index, and t represents the risk transmission time variable.

7. The method according to claim 4, characterized in that, The step of mapping the transshipment process model, the transshipment risk, and the risk evolution trend to a geographic information system to generate a transshipment risk monitoring view as a risk identification result includes: Based on the spatiotemporal information of nodes in the aforementioned transit process model, a geographic raster index is encoded. Based on the geographic raster index and combined with the risk factor topology of the transit risk, the risk value is mapped to the corresponding geographic raster cell to obtain a risk distribution raster map. Based on the risk distribution raster map, and combined with the spatiotemporal decay function family of the risk accumulation equation, the spatial weight matrix of risk diffusion is calculated. Based on the spatial weight matrix, and combined with the probability distribution of risk propagation paths in the risk evolution trend, a risk propagation heat map is constructed. Based on the authorization authentication information and disposal status identifier in the multi-party collaborative disposal verification table, a responsible entity label is generated, which includes the jurisdiction area, jurisdiction authority and solid waste disposal status. By associating the solid waste component attribute identifiers with the risk level indicators using color, a solid waste risk level color code layer is obtained. The risk propagation heat map, the responsible entity label, and the solid waste risk level color code layer are spatiotemporally overlaid to generate the transfer risk monitoring view.

8. A solid waste transfer risk identification system, characterized in that, The system includes: The transfer model construction module is used to acquire tracking data of solid waste during the transfer process and to construct a transfer process model based on the tracking data. The transshipment risk identification module is used to identify transshipment risks based on the transshipment process model and a preset risk indicator system. The risk accumulation equation module is used to generate a risk accumulation equation based on the transfer risk using a risk transmission model. The risk evolution trend module is used to calculate the risk evolution trend based on the risk accumulation equation using a pre-trained machine learning model. The risk monitoring view module is used to map the transshipment process model, the transshipment risks, and the risk evolution trends to a geographic information system to generate a transshipment risk monitoring view as a risk identification result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images