A logistics monitoring and simulation system based on digital twinning

Abstract

The application discloses a logistics monitoring and simulation system based on digital twinning, and belongs to the technical field of logistics, and the system comprises a data acquisition module, a digital twinning modeling module, a logistics simulation deduction module, a running monitoring module, an optimization decision module, a data interface module, a distributed storage module, an abnormal alarm module and a three-dimensional visualization module. With the aid of the digital twinning technology, the logistics network model can be quickly established in a virtual environment, the optimal path design scheme is found through data analysis and simulation experiments, and the purpose of saving transportation cost is achieved.

Description

Technical Field

[0001] This invention belongs to the field of logistics technology, specifically relating to a logistics monitoring and simulation system based on digital twins. Background Technology

[0002] Currently, with the rapid economic development and the gradual improvement of the logistics industry, e-commerce has been integrated into everyone's life. Every day, tens of thousands of logistics trucks are transporting goods across the country. Modern logistics involves multiple links such as warehousing, transportation, distribution, and supply chain collaboration. Traditional management methods are difficult to cope with large-scale and dynamic demands.

[0003] Chinese Patent CN202010594366.1 discloses an intelligent logistics management system and its logistics management method. The intelligent logistics management system includes an order receiving module, an order identification module, a central control module, an order monitoring module, an order processing module, an order scheduling module, an order cancellation module, an order classification module, an order collection module, an order recording module, an order feedback module, an order printing module, a goods retrieval and placement module, a goods confirmation module, a goods scanning module, a goods marking module, a goods transportation module, a goods delivery module, and an order completion module. This invention enables real-time monitoring of order information, simplifies the order processing process, has a high degree of intelligence, and ensures orderly logistics management.

[0004] Chinese Patent CN202410732051.7 discloses a smart logistics management system and its logistics management method, comprising: an identification code generation module, which acquires goods entering the logistics system, creates identification codes, pairs the identification codes with the goods, and prints the identification codes onto the goods; a logistics information collection module; an information tracking module; an access verification module; an anomaly detection module, which performs goods quality anomaly detection in the logistics system; an identification code repair module, which performs integrity checks on the identification codes on the goods; and an identification code repair module, which acquires the identification codes of characteristic goods and reprints the identification codes onto the characteristic goods. By setting up the anomaly detection module, the identification code repair module, and the information tracking module, the identification codes of the goods are repaired, and the goods can be tracked through the identification codes throughout the entire logistics transmission process.

[0005] However, the aforementioned patents still have the following drawbacks:

[0006] 1. There is a lack of dynamic tracking of the transportation process; the status of goods in transit relies on manual telephone confirmation. Vehicle positioning accuracy is only at the kilometer level, making it impossible to identify specific road congestion or abnormal parking.

[0007] 2. Relying on preset rules to handle anomalies requires a lot of time to manually develop detour plans when encountering sudden road conditions, resulting in transportation delays.

[0008] 3. Relying on traditional navigation to determine transportation routes makes it impossible to dynamically adjust transportation routes according to actual conditions, resulting in higher transportation costs. Summary of the Invention

[0009] The purpose of this invention is to provide a logistics monitoring and simulation system based on digital twins. By using digital twin technology, a logistics network model can be quickly established in a virtual environment. Through data analysis and simulation experiments, the optimal route design scheme can be found to save transportation costs.

[0010] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0011] A logistics monitoring and simulation system based on digital twins includes a data acquisition module, a digital twin modeling module, a logistics simulation and deduction module, an operation monitoring module, an optimization decision-making module, a data interface module, a distributed storage module, an anomaly alarm module, and a three-dimensional visualization module.

[0012] The data acquisition module is used to collect real-time data on the location, speed, and cargo status of transport vehicles, and simultaneously integrate road conditions, weather, and driver behavior information. Through multi-source data fusion, it provides real-time early warning for logistics monitoring and supports the logistics simulation module in deriving transportation strategies.

[0013] The digital twin modeling module is used to integrate vehicle positioning, cargo status and real-time road condition data to construct a virtual transportation scenario that is synchronized with the physical world in real time.

[0014] The logistics simulation and deduction module is based on digital twin models and real-time data to simulate dynamic scenarios in transportation, predict the impact of different routes on timeliness, energy consumption and cargo safety, generate multiple emergency plans, and optimize resource allocation by combining historical transportation data to verify the feasibility of new routes.

[0015] The operation monitoring module analyzes abnormal risks based on the data of the location, speed, cargo status, road conditions, weather, and driver behavior collected by the data acquisition module. It triggers the abnormal alarm module to issue warnings in a timely manner and links with the logistics simulation and deduction module to generate emergency plans, verify the feasibility of temporary routes, and revise the transportation plan in real time.

[0016] The optimization decision-making module integrates data from the digital twin modeling module, logistics simulation and deduction module, and operation monitoring module to balance cost, timeliness and risk, automatically generate multiple route solutions, verify the feasibility of the solutions through simulation, and find the optimal route solution.

[0017] The data interface module is used to connect the data acquisition module, digital twin modeling module, logistics simulation and simulation module and external data sources. It receives road condition information in real time, cleans and converts the data format synchronously, distributes it to the operation monitoring module to trigger the abnormal alarm module in a timely manner, and inputs road condition information parameters into the logistics simulation and simulation module to support emergency plan simulation.

[0018] The distributed storage module is used to build a multi-node collaborative data lake to achieve hierarchical storage and fast retrieval of heterogeneous data;

[0019] The anomaly alarm module, through real-time interaction with the operation monitoring module and in conjunction with a preset rule base, performs anomaly detection and issues timely warning signals.

[0020] The 3D visualization module is used to construct a global monitoring interface that maps virtual and real data, dynamically presenting the 3D spatiotemporal status of transport vehicles, goods, and the environment.

[0021] Further specifying, the data acquisition module includes:

[0022] The device access module is used to identify the sensor group of the transport vehicle, which includes a temperature and humidity sensor, a vibration sensor, and an on-board camera, and to establish a device database; it also connects to a GPS positioning device to analyze the latitude, longitude, speed, and timestamp of the transport vehicle.

[0023] The protocol adaptation module is used to unify heterogeneous data into the Apache Avro format;

[0024] The data cleaning and repair module is used to remove abnormal values ​​from the temperature and humidity sensors, generate interpolated trajectories using the Kalman filter algorithm for periods of GPS signal loss, denoise the raw data from the vibration sensors, and extract effective vibration energy features.

[0025] The spatiotemporal alignment and standardization module is used to calibrate the clocks of all devices, ensuring that sensor group data is consistent with GPS timestamps; it also converts temperature control data to °C and vibration intensity to m / s².

[0026] The multi-source heterogeneous data fusion module is used to integrate real-time rainfall data from meteorological APIs into transport vehicle data and correct temperature and humidity sensor readings; dynamically adjust the sampling frequency of sensor groups according to waybill priority; and build a unified data model based on the ISO 19848 standard to map fields from different sources.

[0027] Furthermore, the logistics simulation module includes:

[0028] The road network dynamic modeling module is used to integrate maps and real-time traffic data to generate a hierarchical road network model covering highways, national roads, and provincial roads, and to mark physical constraints such as slope, bridge load-bearing capacity, and tunnel height restrictions; it also integrates external data such as weather warnings and traffic accidents in real time to generate three-dimensional visualized restricted areas.

[0029] The vehicle energy consumption simulation module calculates energy consumption by combining vehicle speed, load, route distance, road conditions, and cumulative running time.

[0030] The real-time strategy verification module is used to receive detour routes recommended by the optimization module and to deduce the balance point between increased fuel consumption and time loss.

[0031] The logistics simulation and deduction module in this invention has the following advantages:

[0032] 1. Traditional management models, based on periodic plan adjustments, struggle to cope with dynamic disturbances such as sudden road conditions and environmental changes. The logistics simulation module, relying on real-time data streams from the operation monitoring module (such as vehicle location, weather warnings, and traffic control), enables the reconstruction of transportation scenarios and multi-strategy simulations. By pre-setting constraints such as cost, timeliness, and safety, the module can quickly generate and select the globally optimal solution, transforming the passive response of traditional manual scheduling into a systematic proactive decision-making process, significantly improving the scientific rigor of the solutions.

[0033] 2. By dynamically extrapolating potential risks such as mechanical failures, path disruptions, and sudden environmental changes, it accurately predicts the risk level and loss range of different response strategies, helping managers to make optimal decisions with controllable risks in complex scenarios and avoid the accumulation of trial and error costs driven by experience.

[0034] 3. Traditional management models focus on single-point optimization (such as minimizing vehicle routes) while neglecting cross-regional and multi-entity resource coordination. The simulation module, through a distributed computing architecture, integrates the real-time status and task requirements of vehicles, warehouses, and manpower to achieve dynamic matching and intelligent scheduling of global resources. Its breakthrough lies in breaking down resource silos and coordinating the generation of combined instructions such as route planning, cargo distribution, and relay transportation. This reduces empty-running rates and redundant inventory while systematically improving the utilization rate of resources across the entire supply chain.

[0035] Further specifying, the digital twin modeling module includes,

[0036] The geographic information input module receives data from the data acquisition module to obtain satellite maps and road network vector data of the transportation route (including road grade and bridge height limit); it imports the coordinates of logistics nodes (warehouses, sorting centers) and marks the range of dense areas (such as freight yards and loading and unloading areas within a radius of 500m).

[0037] The dynamic object recognition module is used to receive data from the vehicle's GPS and extract the location and speed of the transport vehicle in real time.

[0038] The regional density calculation module, based on the DBSCAN clustering algorithm, calculates the number of objects per unit area with logistics nodes as the center. It generates a LOD3 level accuracy model in densely populated areas of logistics nodes, with a density of >20 objects / 100㎡ (such as loading and unloading platforms). In open areas, a LOD1 level accuracy model is used, with a density of <5 objects / 100㎡ (such as highway sections).

[0039] The adaptive mesh generation and LOD grading module uses a quadtree algorithm to divide the basic mesh and converts the coordinate system to UTM partition projection; it labels the mesh level according to the region density: in dense areas, it constructs millimeter-level features for stationary or low-speed objects; when a vehicle enters a highway, it triggers the LOD1 mode.

[0040] The real-time rendering optimization and data synchronization module is used to enable the subdivision shader and load 4K PBR materials in the LOD3 region; in the LOD1 region, it uses instantiated rendering to batch process the same model; when the vehicle accelerates from the warehouse (v=0.5m / s) to the park road (v=3m / s): it uses geometric gradient technology to linearly reduce the number of model faces from 20,000 to 8,000; when the speed threshold is triggered, the LOD3 collider is retained, and the accuracy of the visual model is reduced simultaneously; the LOD state is synchronized to the logistics simulation and deduction module through Kafka.

[0041] The digital twin modeling module in this invention has the following advantages:

[0042] 1. By integrating with the operation monitoring module, the system can synchronize multi-source data such as vehicle status, cargo information, and environmental parameters (such as weather and road conditions) in real time to construct a dynamic virtual mirror of the physical world. This enables the model to accurately reflect the real-time status changes of the logistics system and provides a high-fidelity foundation for subsequent analysis and decision-making.

[0043] 2. By combining real-time models with simulation engines, it supports predictive simulations of scenarios such as mechanical failures and path blockages. It can simulate the standard evolution process of known risks and explore the chain effects of extreme scenarios through generative algorithms, significantly improving the system's risk prediction capabilities.

[0044] 3. By mapping physical entities such as vehicles and personnel in the logistics network into interactive virtual objects, it supports the rapid verification of multi-objective collaborative optimization algorithms. Simultaneously, by comparing the execution deviations between the physical system and the virtual model in real time, it can automatically trigger model parameter calibration and strategy iteration, forming a closed-loop control system and continuously improving decision-making accuracy.

[0045] Further specifying, the optimization decision module includes,

[0046] The real-time strategy generation module is used to generate decision trees that include route changes, resource allocation, and risk response based on historical data from simulations; it also receives real-time traffic information from the logistics simulation module and generates detour plans.

[0047] The resource dynamic scheduling module dynamically matches temporary orders based on the real-time location and load of vehicles en route (e.g., vehicles with 30% remaining load are given priority); dispatches nearby rescue vehicles when a vehicle breaks down; and changes priorities in real time according to the shipper's needs.

[0048] The dynamic route planning module generates the optimal route by comprehensively considering the cost of toll roads, the risk value of mountain road sections, and the distribution of gas stations; when the GPS detects that the vehicle has deviated from the predetermined route, it replans the route based on real-time weather conditions.

[0049] The transportation cost analysis module calculates transportation costs based on total energy consumption, fuel prices, environmental factors, and timeliness factors.

[0050] The decision verification and feedback module is used to send the generated scheduling plan to the logistics simulation and deduction module, compare the transportation costs of the new plan with the original plan, and select the optimal plan.

[0051] The optimization decision module in this invention has the following advantages:

[0052] 1. By accessing multi-source data streams (such as vehicle status, road conditions, and order changes) from the real-time operation monitoring module, a dynamic decision engine is built. This engine can generate and iterate alternative solutions for emergencies (such as route blockages and urgent orders), breaking through the response speed bottleneck of traditional manual decision-making and realizing the transformation from passive response to proactive intervention.

[0053] 2. By coordinating the real-time status and constraints of resources such as vehicles and manpower, and using multi-objective optimization algorithms to simultaneously calculate core indicators such as cost, timeliness, risk, and energy consumption, the optimal solution is generated, achieving a dual leap in resource utilization and service quality.

[0054] 3. By continuously analyzing the deviation between the effects of historical decisions and actual implementation, we can independently correct model parameters and optimize weights to ensure that the decision-making logic always iterates in sync with real business scenarios.

[0055] Further specifying, the 3D visualization module includes,

[0056] The high-precision 3D scene construction module integrates satellite remote sensing data and lidar point clouds to generate a 3D road network model that includes terrain elevation, bridges and tunnels, and service area buildings; it can also load detailed models of specific areas as needed, including toll station lane distribution and gas station locations.

[0057] The real-time dynamic data fusion module is used to integrate vehicle GPS trajectory, weather radar data, and roadside unit traffic statistics into a three-dimensional scene to form a dynamic heat map; and to apply a red outline warning to speeding vehicles.

[0058] The interactive decision-making sandbox module supports dragging and dropping to modify virtual roadblocks (such as simulating the location of a landslide) and simulates the feasibility of vehicle detour routes in real time.

[0059] The multi-level visualization module displays the distribution of trunk transportation traffic in the view, and clicking on a single vehicle model allows you to see through to view the temperature and humidity curves inside the cargo box and the driver fatigue index.

[0060] The simulation process replay module is used to load historical weather data and replay vehicle trajectory deviations; it displays the actual transportation route and the simulated route side by side, and marks key divergence points.

[0061] The advantages of this invention are mainly reflected in the following aspects:

[0062] 1. The data acquisition module enables real-time capture and cleaning of data across all domains, including vehicles, goods, and the environment, overcoming the latency and error bottlenecks of traditional manual data entry. Combined with the high-concurrency processing capabilities of the distributed storage module, it ensures high-speed data access and consistency, providing a stable data foundation for subsequent modeling and decision-making.

[0063] 2. The digital twin modeling module constructs a real-time virtual mirror of the logistics system, accurately mapping the state changes of the physical world. The logistics simulation and deduction module pre-determines the potential effects of strategies such as path planning and resource allocation based on this mirror, while the optimization decision-making module generates the optimal solution. The three form a closed-loop link of "perception-deduction-decision," realizing a paradigm upgrade from static rule execution to dynamic strategy adaptation.

[0064] 3. The operation monitoring module tracks key parameters such as vehicle location and equipment status in real time, while the anomaly alarm module predicts risks and triggers emergency mechanisms. Combined with the interactive interface of the 3D visualization module, managers can gain a comprehensive understanding of the entire logistics network's operational status and coordinate warehousing, transportation, and human resources simultaneously.

[0065] 4. The data interface module integrates with external systems such as suppliers and carriers through open protocols, eliminating information silos. The distributed storage module supports multi-level data access control, ensuring the security of sensitive information. The anomaly alarm module standardizes and categorizes risk events and automatically pushes them to the responsible parties, building a transparent governance system based on multi-party collaboration. Attached Figure Description

[0066] The present invention can be further illustrated by the non-limiting embodiments given in the accompanying drawings;

[0067] Figure 1This is a schematic diagram of the structure of an embodiment of the present invention;

[0068] Figure 2 This is a schematic diagram of the data acquisition module in an embodiment of the present invention;

[0069] Figure 3 This is a schematic diagram of the structure of the digital twin modeling module in an embodiment of the present invention;

[0070] Figure 4 This is a schematic diagram of the logistics simulation and deduction module in an embodiment of the present invention;

[0071] Figure 5 This is a schematic diagram of the structure of the optimization decision module in an embodiment of the present invention;

[0072] Figure 6 This is a schematic diagram of the structure of the three-dimensional visualization module in an embodiment of the present invention;

[0073] Figure 7 This is a flowchart illustrating the operation of the monitoring module in this embodiment of the invention.

[0074] Figure 8 This is a flowchart illustrating the operation of the abnormal alarm module in this embodiment of the invention. Detailed Implementation

[0075] To enable those skilled in the art to better understand the present invention, the technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0076] like Figures 1-8 As shown, the present invention provides a logistics monitoring and simulation system based on digital twins, including a data acquisition module, a digital twin modeling module, a logistics simulation and deduction module, an operation monitoring module, an optimization decision-making module, a data interface module, a distributed storage module, an anomaly alarm module, and a three-dimensional visualization module.

[0077] The data acquisition module is used to collect real-time data on the location, speed, and cargo status of transport vehicles, and simultaneously integrates road conditions, weather, and driver behavior information. Through multi-source data fusion, it provides real-time early warning for logistics monitoring and supports the logistics simulation module in deriving transportation strategies.

[0078] This data acquisition module includes:

[0079] The device access module is used to identify the sensor group of the transport vehicle, which includes a temperature and humidity sensor, a vibration sensor, and an on-board camera, and to establish a device database; it also connects to a GPS positioning device to analyze the latitude, longitude, speed, and timestamp of the transport vehicle.

[0080] The protocol adaptation module is used to unify heterogeneous data into the Apache Avro format;

[0081] The data cleaning and repair module is used to remove abnormal values ​​from the temperature and humidity sensors, generate interpolated trajectories using the Kalman filter algorithm for periods of GPS signal loss, denoise the raw data from the vibration sensors, and extract effective vibration energy features.

[0082] The spatiotemporal alignment and standardization module is used to calibrate the clocks of all devices, ensuring that sensor group data is consistent with GPS timestamps; it also converts temperature control data to °C and vibration intensity to m / s².

[0083] The multi-source heterogeneous data fusion module is used to integrate real-time rainfall data from meteorological APIs into transport vehicle data and correct temperature and humidity sensor readings; dynamically adjust the sampling frequency of sensor groups according to waybill priority; and build a unified data model based on the ISO 19848 standard to map fields from different sources.

[0084] The formula for calculating the data fusion weight of the multi-source heterogeneous data fusion module is as follows: ,in, For the first The fusion weight of each data source For the first Reliability rating of each data source For the first Timeliness rating of each data source Assigning weights to road classifications This is a real-time congestion index. For the reliability score of the j-th data source, The timeliness score for the j-th data source.

[0085] In the above data fusion weight calculation formula The calculation process is as follows:

[0086] Referring to ISO 18972:2021 "Data Quality Standard for Sensors in Logistics", a 20% error rate is the threshold for equipment scrapping, assigned a value of 0.8; a 5% error rate reaches the high-precision standard, assigned a value of 1.2.

[0087] If the monthly equipment uptime is ≤80% (below the industry standard for stable operation), a value of 0.8 is forcibly assigned; if the monthly equipment uptime is ≥95% (the threshold for stable operation), a value of 1.2 is forcibly assigned.

[0088] In the data cleaning process, the criteria for determining the best data source require that low-reliability data sources be subject to mandatory weight reduction (e.g., a value of 0.8 is assigned when the missing rate is ≥10%; a value of 1.2 is assigned when the missing rate is ≤2%).

[0089] therefore The range of values ​​for is (0.8≤ ≤1.2).

[0090] Assuming a certain data metric is: historical error rate 23%, monthly equipment uptime 75%, and data missing rate 12%, then what is the data source reliability score affected by the error rate? Since the result is below 0.8, it is forcibly assigned a value of 0.8. Data source reliability score affected by monthly equipment uptime. Since the result is below 0.8, it is forcibly assigned a value of 0.8. Data source reliability score affected by missing rate. Since the result is lower than 0.8, it is forcibly assigned the value 0.8. Finally, the most stringent value of the three is taken. =0.8.

[0091] The calculation process is as follows:

[0092] Assuming time delay =3 seconds, update frequency =0.2Hz, transmission success rate =96%, weights are determined using the Delphi method: time delay weights =0.5, Update frequency weight =0.3, Transmission success rate weight =0.2.

[0093] Normalize the original data:

[0094] Set maximum allowable delay =10 seconds, then the time delay score .

[0095] Set ideal frequency =1Hz, then update frequency score .

[0096] If the transmission success rate score is directly expressed as a percentage value, then... .

[0097] Calculated according to the weighted formula .

[0098] Road grade weight Reflecting the inherent traffic capacity of road infrastructure, it is determined comprehensively by road type, design standards, and historical data. The calculation process is as follows:

[0099] According to national standards (such as the "Technical Standards for Highway Engineering"), roads are divided into four categories and assigned benchmarks. Values, as shown in Table 1:

[0100]

[0101] Based on actual road conditions, the benchmark The value is dynamically adjusted:

[0102] Hypothesis correction factor This indicates traffic control policies (such as restricted hours for trucks). If the current time is 7:00-9:00 (trucks are prohibited during the morning rush hour). The value increases by 0.2, and the benchmark value for a main road in a certain urban area... =0.5, which falls within the restricted hours for trucks: =0.5+ =0.5 + 0.2 = 0.7, indicating that the road's traffic efficiency is only 30% of the ideal state. These should be avoided as a priority.

[0103] Real-time congestion index Reflecting the real-time level of road congestion, it is dynamically calculated based on multi-source traffic data, with a value range of [0,1], where 0 represents smooth traffic and 1 represents complete congestion. The calculation process is as follows:

[0104] ,

[0105] in The vehicle speed ratio is calculated as the real-time average vehicle speed divided by the design speed. Traffic saturation is calculated as the current traffic volume divided by the road's maximum capacity (e.g., if the maximum capacity is 2000 vehicles / hour and the actual traffic volume is 1600 vehicles). =0.8); The duration of the accident's impact, i.e., the cumulative delay caused by the accident; The threshold for the maximum impact of the accident; This is a weighting coefficient, based on the experience of urban traffic management departments.

[0106] Suppose a traffic accident occurs on a main road in a certain city, and the data is as follows: Weighting coefficients: , =30 minutes, design speed 40km / h, real-time average vehicle speed 12km / h, traffic flow saturation =0.9, the accident has caused a delay of 25 minutes, then , =0.3, substitute into the formula: ≈0.25.

[0107] The digital twin modeling module is used to integrate vehicle positioning, cargo status, and real-time road condition data to construct a virtual transportation scenario that is synchronized with the physical world in real time. The digital twin modeling module further includes:

[0108] The geographic information input module receives data from the data acquisition module to obtain satellite maps and road network vector data of the transportation route (including road grades and bridge height restrictions); it also imports the coordinates of logistics nodes (warehouses, sorting centers) and marks densely populated areas (such as freight yards and loading / unloading areas within a radius of 500m).

[0109] The dynamic object recognition module is used to receive data from the vehicle's GPS and extract the location and speed of the transport vehicle in real time.

[0110] The regional density calculation module, based on the DBSCAN clustering algorithm, calculates the number of objects per unit area with logistics nodes as the center. It generates a LOD3 level accuracy model in densely populated areas of logistics nodes, where the density (LOD3) is greater than 20 objects / 100㎡ (such as loading and unloading platforms); in open areas, it uses a LOD1 level accuracy model, where the density (LOD1) is less than 5 objects / 100㎡ (such as highway sections).

[0111] The adaptive mesh generation and LOD grading module uses a quadtree algorithm to divide the basic mesh and converts the coordinate system to UTM partition projection; it labels the mesh level according to the region density: in dense areas, it constructs millimeter-level features for stationary or low-speed objects; when a vehicle enters a highway, it triggers the LOD1 mode.

[0112] The real-time rendering optimization and data synchronization module is used to enable the subdivision shader and load 4K PBR materials in the LOD3 region; use instantiated rendering to batch process the same model in the LOD1 region; when the vehicle accelerates from the warehouse to the park road, the geometric gradient technique is used to linearly reduce the number of model faces from 20,000 to 8,000; when the speed threshold is triggered, the LOD3 collider is retained and the accuracy of the visual model is reduced simultaneously; and the LOD state is synchronized to the logistics simulation and inference module through Kafka.

[0113] The logistics simulation module, based on digital twin models and real-time data, simulates dynamic scenarios in transportation, predicts the impact of different routes on timeliness, energy consumption, and cargo safety, generates multiple emergency plans, and optimizes resource allocation by combining historical transportation data to verify the feasibility of new routes.

[0114] This logistics simulation module includes:

[0115] The road network dynamic modeling module is used to integrate maps and real-time traffic data to generate a hierarchical road network model covering highways, national roads, and provincial roads, and to mark physical constraints such as slope, bridge load-bearing capacity, and tunnel height restrictions; it also integrates external data such as weather warnings and traffic accidents in real time to generate three-dimensional visualized restricted areas.

[0116] The vehicle energy consumption simulation module calculates energy consumption by combining vehicle speed, load, route distance, road conditions, and cumulative running time.

[0117] The real-time strategy verification module is used to receive detour routes recommended by the optimization module and to deduce the balance point between increased fuel consumption and time loss.

[0118] The energy consumption prediction formula for the vehicle energy consumption simulation module is as follows: ,in, This is the predicted total energy consumption value. For the speed of transport vehicles, For real-time load capacity, The path distance. This is the road condition attenuation coefficient. This represents the cumulative running time.

[0119] The road condition attenuation coefficient in the above energy consumption prediction formula The calculation process is as follows:

[0120] The calculation formula is: , Assigning weights to road classifications This is a real-time congestion index. These are environmental disturbance factors. and The calculation process has been explained previously and will not be repeated here. Environmental disturbance factor This is calculated in real-time using meteorological API and road condition monitoring. For example:

[0121] Assuming that during transportation, a certain section of the road encounters a crosswind of force 8 (wind speed 18 m / s) and the snow accumulation on the road is 5 cm, then what is the environmental disturbance factor affected by wind speed? ,in The value represents wind speed, and 0.15 is the baseline value for plains areas. When the ambient temperature is -10℃, the engine thermal efficiency decreases by 5%, and the environmental disturbance factor affected by ambient temperature is... ,in Ambient temperature. Combined meteorological disturbances. Coefficient of friction on snow-covered roads coefficient of friction of dry road surface The environmental disturbance factors affected by snow accumulation Environmental disturbance factors .

[0122] The operation monitoring module analyzes abnormal risks based on the data collected by the data acquisition module, including the location, speed, cargo status, road conditions, weather, and driver behavior data of the transport vehicles. It then triggers an alarm module to issue a warning in a timely manner and links with the logistics simulation module to generate emergency plans, verify the feasibility of temporary routes, and revise the transportation plan in real time.

[0123] The operation steps of this monitoring module are as follows:

[0124] S101. Receive data and obtain real-time information from the data acquisition module, including vehicle location, speed, cargo temperature and humidity, road conditions (such as congestion, construction), weather (such as rain, snow, strong wind), and driver operation (such as continuous driving time).

[0125] S102. Organize data and filter out erroneous data: for example, exclude abnormal coordinates with excessive GPS positioning deviation; supplement missing values. If vehicle speed data is lost for a certain period of time, calculate a reasonable value based on the speed at the time points before and after the loss.

[0126] S103. Identify potential risks and, based on the processed data, determine whether any anomalies exist:

[0127] S104. Activate emergency response. When a high risk is detected, trigger an alarm and transmit the risk type and severity level to the abnormal alarm module. At the same time, send a request to the logistics simulation module to generate a detour route.

[0128] S105. Verify the feasibility of the emergency plan and quickly verify the temporary plan provided by the logistics simulation module: calculate whether the detour route will cause the delivery time to exceed the limit, and confirm that the new route has no conflict risk (such as heavy trucks cannot pass through bridges with a height limit of 3 meters).

[0129] S106. Adjust the transportation plan and update it in real time after verification.

[0130] The optimization decision-making module integrates data from the digital twin modeling module, logistics simulation and extrapolation module, and operation monitoring module to balance cost, timeliness and risk, automatically generate multiple route options, verify the feasibility of the options through simulation, and find the optimal route option.

[0131] This optimization decision-making module includes:

[0132] The real-time strategy generation module is used to generate decision trees that include route changes, resource allocation, and risk response based on historical data from simulations; it also receives real-time traffic information from the logistics simulation module and generates detour plans.

[0133] The resource dynamic scheduling module dynamically matches temporary orders based on the real-time location and load of vehicles en route (e.g., vehicles with 30% remaining load are given priority); dispatches nearby rescue vehicles when a vehicle breaks down; and changes priorities in real time according to the shipper's needs.

[0134] The dynamic route planning module generates the optimal route by comprehensively considering the cost of toll roads, the risk value of mountain road sections, and the distribution of gas stations; when the GPS detects that the vehicle has deviated from the predetermined route, it replans the route based on real-time weather conditions.

[0135] The transportation cost analysis module calculates transportation costs based on total energy consumption, fuel prices, environmental factors, and timeliness factors.

[0136] The decision verification and feedback module is used to send the generated scheduling plan to the logistics simulation and deduction module, compare the transportation costs of the new plan with the original plan, and select the optimal plan.

[0137] The transportation cost calculation formula in the above transportation cost analysis module is as follows: ,in This is the predicted total energy consumption value. For fuel prices, As environmental disturbance factors, This is the time-loss coefficient. To simulate and predict delay time.

[0138] Among them, the total energy consumption forecast value and environmental disturbance factors The calculation process has been explained previously and will not be repeated here. Time-loss coefficient The calculation needs to be combined with the cargo priority and the risk of transportation delay. For example, suppose a refrigerated truck is transporting fresh goods (total freight cost of 10,000 yuan). A multi-vehicle rear-end collision is detected on a certain section of the road, and the congestion is expected to last for 1.5 hours. On another section of the road, there is heavy fog and the speed limit is reduced to 40 km / h, resulting in a delay of 1 hour.

[0139] Goods type priority: Fresh goods are prioritized according to industry standards. =200 yuan / hour;

[0140] Contractual penalty percentage: The contract stipulates that a penalty of 0.5% of the freight fee will be imposed for each hour of delay, i.e., 10000 × 0.5% = 50 yuan / hour;

[0141] Initial time loss coefficient: .

[0142] Fresh produce needs to be maintained at -18℃, which increases the risk of downtime by 30% and fuel consumption by 15% in low-temperature environments. .

[0143] The data interface module is used to connect the data acquisition module, digital twin modeling module, logistics simulation and simulation module and external data sources. It receives road condition information in real time, cleans and converts the data format synchronously, distributes it to the operation monitoring module to trigger the abnormal alarm module in a timely manner, and inputs road condition information parameters into the logistics simulation and simulation module to support emergency plan simulation.

[0144] The distributed storage module is used to build a multi-node collaborative data lake to achieve hierarchical storage and fast retrieval of heterogeneous data.

[0145] The anomaly alarm module interacts with the operation monitoring module in real time and uses a preset rule base to determine anomalies and issue early warning signals in a timely manner.

[0146] The working steps of this abnormal alarm module are as follows:

[0147] S201, Data Access: Receives structured data packets preprocessed by the operation monitoring module, including dynamic data of transport vehicles, environmental characteristic data, driver status, and cargo safety parameters.

[0148] S202. Warning type determination: When the lateral acceleration is greater than 0.4g, a vehicle sideslip warning is triggered; when the slope of the decrease in eye movement frequency is greater than 15% / minute, a fatigue driving warning is triggered; when the road surface temperature is less than or equal to 0℃ and the humidity is greater than or equal to 80%, a road icing warning is triggered; when the temperature fluctuation in the passenger compartment exceeds ±2℃ / 15 minutes, a temperature control warning is triggered.

[0149] S203. Warning signal classification: When the hazard is lethal, the warning level is red level 1; when the hazard is severe, the warning level is orange level 2; when the hazard is moderate, the warning level is yellow level 3.

[0150] S204. Warning signal distribution: Red Level 1 alarm: directly activates the ESP electronic stability system to force speed reduction; Orange Level 2 alarm: project warning icons through the HUD; for the dispatch center, push alarm details and handling suggestions through a dedicated protocol; for the driver, choose between a buzzer alarm or voice broadcast.

[0151] The 3D visualization module is used to build a global monitoring interface that maps virtual and real data, dynamically presenting the 3D spatiotemporal status of transport vehicles, goods, and the environment.

[0152] This 3D visualization module includes:

[0153] The high-precision 3D scene construction module integrates satellite remote sensing data and lidar point clouds to generate a 3D road network model that includes terrain elevation, bridges and tunnels, and service area buildings; it can also load detailed models of specific areas as needed, including toll station lane distribution and gas station locations.

[0154] The real-time dynamic data fusion module is used to integrate vehicle GPS trajectory, weather radar data, and roadside unit traffic statistics into a three-dimensional scene to form a dynamic heat map; and to apply a red outline warning to speeding vehicles.

[0155] The interactive decision-making sandbox module supports dragging and dropping to modify virtual roadblocks (such as simulating the location of a landslide) and simulates the feasibility of vehicle detour routes in real time.

[0156] The multi-level visualization module displays the distribution of trunk transportation traffic in the view, and clicking on a single vehicle model allows you to see through to view the temperature and humidity curves inside the cargo box and the driver fatigue index.

[0157] The simulation process replay module is used to load historical weather data and replay vehicle trajectory deviations; it displays the actual transportation route and the simulated route side by side, and marks key divergence points.

[0158] The working principle of this invention is as follows: The data acquisition module acquires real-time data such as vehicle location and cargo status through GPS, sensors, and other devices. After standardization by the data interface module, the data is stored in the distributed storage module. The digital twin modeling module calls the stored data to construct a virtual transportation scenario, generating dynamic digital twins of vehicles and road networks. The operation monitoring module analyzes the real-time data stream and triggers a tiered warning from the anomaly alarm module when an anomaly is detected. Simultaneously, it sends an emergency simulation request to the logistics simulation module. The logistics simulation module simulates detour routes and resource allocation plans based on the twin model, submits the feasibility results to the optimization decision module, and generates the final execution instructions based on constraints such as cost and timeliness, updating the digital twin model. The 3D visualization module synchronously renders the entire process dynamically, intuitively displaying vehicle trajectories, risk points, and the effectiveness of emergency plans. All modules achieve cross-platform data interoperability through the data interface module, forming a closed loop of "perception-analysis-decision-execution".

[0159] The above provides a detailed description of a logistics monitoring and simulation system based on digital twins provided by this invention. The specific embodiments are described only to aid in understanding the method and core ideas of this invention. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the scope of protection of the claims of this invention.

Claims

1. A logistics monitoring and simulation system based on digital twins, characterized in that: It includes a data acquisition module, a digital twin modeling module, a logistics simulation and deduction module, an operation monitoring module, an optimization decision-making module, a data interface module, a distributed storage module, an anomaly alarm module, and a 3D visualization module; The data acquisition module is used to collect real-time data on the location, speed, and cargo status of transport vehicles, and simultaneously integrate road conditions, weather, and driver behavior information. Through multi-source data fusion, it provides real-time early warning for logistics monitoring and supports the logistics simulation module in deriving transportation strategies. The digital twin modeling module is used to integrate vehicle positioning, cargo status and real-time road condition data to construct a virtual transportation scenario that is synchronized with the physical world in real time. The logistics simulation and deduction module is based on digital twin models and real-time data to simulate dynamic scenarios in transportation, predict the impact of different routes on timeliness, energy consumption and cargo safety, generate multiple emergency plans, and optimize resource allocation by combining historical transportation data to verify the feasibility of new routes. The operation monitoring module analyzes abnormal risks based on the data of the location, speed, cargo status, road conditions, weather, and driver behavior collected by the data acquisition module. It triggers the abnormal alarm module to issue warnings in a timely manner and links with the logistics simulation and deduction module to generate emergency plans, verify the feasibility of temporary routes, and revise the transportation plan in real time. The optimization decision-making module integrates data from the digital twin modeling module, logistics simulation and deduction module, and operation monitoring module to balance cost, timeliness and risk, automatically generate multiple route solutions, verify the feasibility of the solutions through simulation, and find the optimal route solution. The data interface module is used to connect the data acquisition module, digital twin modeling module, logistics simulation and simulation module and external data sources. It receives road condition information in real time, cleans and converts the data format synchronously, distributes it to the operation monitoring module to trigger the abnormal alarm module in a timely manner, and inputs road condition information parameters into the logistics simulation and simulation module to support emergency plan simulation. The distributed storage module is used to build a multi-node collaborative data lake to achieve hierarchical storage and fast retrieval of heterogeneous data; The anomaly alarm module, through real-time interaction with the operation monitoring module and in conjunction with a preset rule base, performs anomaly detection and issues timely warning signals. The three-dimensional visualization module is used to construct a global monitoring interface that maps virtual and real data, dynamically presenting the three-dimensional spatiotemporal status of transport vehicles, goods, and the environment. The data acquisition module further includes, The device access module is used to identify the sensor group of the transport vehicle, which includes a temperature and humidity sensor, a vibration sensor, and an on-board camera, and to establish a device database. Connect to a GPS positioning device to analyze the latitude, longitude, speed, and timestamp of the transport vehicle; The protocol adaptation module is used to unify heterogeneous data into the Apache Avro format; The data cleaning and repair module is used to remove abnormal values ​​from the temperature and humidity sensors, generate interpolated trajectories using the Kalman filter algorithm for periods of GPS signal loss, denoise the raw data from the vibration sensors, and extract effective vibration energy features. The spatiotemporal alignment and standardization module is used to calibrate the clocks of all devices, ensuring that sensor group data is consistent with GPS timestamps; it also converts temperature control data to °C and vibration intensity to m / s². The multi-source heterogeneous data fusion module is used to integrate real-time rainfall data from meteorological APIs into transport vehicle data and correct temperature and humidity sensor readings; dynamically adjust the sampling frequency of sensor groups according to waybill priority; and build a unified data model based on the ISO 19848 standard to map fields from different sources.

2. The logistics monitoring and simulation system based on digital twins according to claim 1, characterized in that: The logistics simulation module further includes, The road network dynamic modeling module is used to integrate maps and real-time traffic data to generate a hierarchical road network model covering highways, national roads, and provincial roads, and to mark physical constraints such as slope, bridge load-bearing capacity, and tunnel height limits; it also integrates external data such as weather warnings and traffic accidents in real time to generate three-dimensional visualized restricted areas. The vehicle energy consumption simulation module calculates energy consumption by combining vehicle speed, load, route distance, road conditions, and cumulative running time. The real-time strategy verification module is used to receive detour routes recommended by the optimization module and to deduce the balance point between increased fuel consumption and time loss.

3. The logistics monitoring and simulation system based on digital twins according to claim 2, characterized in that: The digital twin modeling module further includes, The geographic information input module receives data from the data acquisition module to obtain satellite maps and road network vector data of the transportation route; it also imports the coordinates of logistics nodes and marks the range of densely populated areas. The dynamic object recognition module is used to receive data from the vehicle's GPS and extract the location and speed of the transport vehicle in real time. The regional density calculation module, based on the DBSCAN clustering algorithm, calculates the number of objects per unit area with logistics nodes as the center. It generates a LOD3 level accuracy model in dense logistics node areas and uses a LOD1 level accuracy model in open areas. The adaptive mesh generation and LOD grading module uses a quadtree algorithm to divide the basic mesh and converts the coordinate system to UTM partition projection; it labels the mesh level according to the region density: in dense areas, it constructs millimeter-level features for stationary or low-speed objects; when a vehicle enters a highway, it triggers the LOD1 mode. The real-time rendering optimization and data synchronization module is used to enable the subdivision shader and load 4K PBR materials in the LOD3 region; in the LOD1 region, it uses instantiated rendering to process the same model in batches; when the vehicle accelerates from the warehouse to the park road: it uses geometric gradient technology to linearly reduce the number of model faces from 20,000 to 8,000; when the speed threshold is triggered, it retains the LOD3 collider and simultaneously reduces the accuracy of the visual model. Synchronize the LOD status to the logistics simulation module via Kafka.

4. A logistics monitoring and simulation system based on digital twins according to claim 3, characterized in that: The optimization decision module further includes, The real-time strategy generation module is used to generate decision trees that include path changes, resource allocation, and risk response based on historical data from simulations. Receive real-time traffic information pushed by the logistics simulation and detour module and generate detour plans; The resource dynamic scheduling module dynamically matches temporary orders based on the real-time location and load of vehicles en route; dispatches nearby rescue vehicles when a vehicle breaks down; and changes priorities in real time according to the needs of the cargo owner. The dynamic route planning module generates the optimal route by comprehensively considering the cost of toll roads, the risk value of mountain road sections, and the distribution of gas stations; when the GPS detects that the vehicle has deviated from the predetermined route, it replans the route based on real-time weather conditions. The transportation cost analysis module calculates transportation costs based on total energy consumption, fuel prices, environmental factors, and timeliness factors. The decision verification and feedback module is used to send the generated scheduling plan to the logistics simulation and deduction module, compare the transportation costs of the new plan with the original plan, and select the optimal plan.

5. A logistics monitoring and simulation system based on digital twins according to claim 4, characterized in that, The 3D visualization module further includes, A high-precision 3D scene construction module is used to integrate satellite remote sensing data and lidar point clouds to generate a 3D road network model that includes terrain elevation, bridges, tunnels, and service area buildings; and to load detailed models of specific areas as needed, including toll station lane distribution and gas station locations. The real-time dynamic data fusion module is used to integrate vehicle GPS trajectory, weather radar data and roadside unit traffic statistics into a three-dimensional scene to form a dynamic heat map; Apply a red outline warning to speeding vehicles; The interactive decision-making sandbox module supports dragging and dropping to modify virtual roadblocks and simulates the feasibility of vehicle detour routes in real time. The multi-level visualization module displays the distribution of trunk transportation traffic in the view, and clicking on a single vehicle model allows you to see through to view the temperature and humidity curves inside the cargo box and the driver fatigue index. The simulation process replay module is used to load historical weather data and replay vehicle trajectory deviations; it displays the actual transportation route and the simulated route side by side, and marks key divergence points.

Images