Goods logistics monitoring management system and method based on Internet of Things

Abstract

The invention relates to the technical field of logistics transportation, in particular to a cargo logistics monitoring management system and method based on the Internet of Things. The method comprises the following steps: acquiring temperature distribution data and micro-deformation data of the surface of the hazardous article container; performing non-contact inversion processing on the internal state of the container according to the temperature distribution data and the micro-deformation data based on a thermodynamic equilibrium principle to obtain dangerous article state characteristic data; acquiring gas spectrum array data through spectrum gas sensors at four corners of the carriage; and according to the dangerous article state characteristic data and the gas spectrum array data, performing three-dimensional positioning processing based on fluid mechanics diffusion science on the leakage source position to obtain leakage risk distribution data. According to the invention, the position of the gas leakage source can be accurately positioned in a three-dimensional manner through the spectrum gas sensors at the four corners of the carriage and the fluid mechanics diffusion model, and rapid tracking and risk assessment of the leakage source are realized through the gas spectrum array data.

Description

Technical Field

[0001] The present invention relates to the field of logistics and transportation technology, and in particular to a cargo logistics monitoring and management system and method based on the Internet of Things. Background Art

[0002] Cargo logistics monitoring and management refers to the real-time monitoring and management of the location, status, environmental conditions and other information of goods in the logistics links such as transportation, warehousing and distribution through various technical means and management methods to ensure the safe, timely and efficient transportation of goods. The cargo logistics monitoring and management system based on the Internet of Things uses the Internet of Things technology to achieve real-time monitoring and management of goods in the process of transportation, warehousing and distribution through sensors, communication networks and data processing platforms. The system can provide information such as the location, status, environmental conditions and other information of goods to help logistics companies improve operational efficiency, reduce operating costs and increase customer satisfaction.

[0003] However, traditional cargo logistics monitoring and management methods often have the following problems: In the transportation of dangerous goods, it is necessary to track the vehicle location, cargo status and surrounding environmental parameters throughout the entire process. The system achieves real-time monitoring through GPS positioning, gas concentration sensors and other equipment. But the technical difficulty of this scenario is: first, the explosion-proof requirements of the sensor are extremely high, which limits the available equipment models and also greatly increases the system cost. Secondly, there is the complexity of multi-parameter collaborative monitoring, which requires simultaneous monitoring of multiple indicators such as position, temperature, pressure, and gas concentration, which puts great pressure on data processing and analysis. In addition, the emergency response mechanism in the event of an abnormality also faces challenges, such as how to achieve remote intervention and control while ensuring safety. Summary of the invention

[0004] Based on this, it is necessary for the present invention to provide a cargo logistics monitoring and management system and method based on the Internet of Things to solve at least one of the above technical problems.

[0005] To achieve the above purpose, a cargo logistics monitoring and management method based on the Internet of Things includes the following steps:

[0006] Step S1: obtaining temperature distribution data and micro-deformation data on the surface of the dangerous goods container; performing non-contact inversion processing on the internal state of the container according to the temperature distribution data and the micro-deformation data based on the principle of thermodynamic equilibrium to obtain characteristic data of the state of the dangerous goods, wherein the characteristic data of the state of the dangerous goods includes pressure trend data, temperature gradient data and chemical activity data;

[0007] Step S2: acquiring gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; performing three-dimensional positioning processing on the leakage source position based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtaining leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data;

[0008] Step S3: using a preset neural network model in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data;

[0009] Step S4: performing distributed control strategy optimization processing according to the situation evolution prediction data and the response characteristic data of the vehicle-mounted actuator obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spraying system;

[0010] Step S5: Dynamically plan the emergency avoidance path according to the collaborative control instruction data and the preset navigation data to obtain emergency response strategy data.

[0011] The present invention can accurately obtain the state characteristic data of dangerous goods, including pressure trend data, temperature gradient data and chemical activity data, by acquiring the temperature distribution data and micro-deformation data on the surface of the dangerous goods container, and then performing non-contact inversion processing based on the principle of thermodynamic equilibrium. This process does not require direct contact with the inside of the dangerous goods container, avoiding the safety risks that may be caused by contact. At the same time, it can grasp the state changes of dangerous goods in the transportation process in real time and accurately, providing a solid data foundation for subsequent risk assessment and disposal, and effectively improving the safety and reliability of dangerous goods transportation. The gas spectrum array data is obtained by using the spectral gas sensors at the four corners of the carriage, and combined with the state characteristic data of dangerous goods, the position of the leakage source is three-dimensionally positioned according to the principle of fluid mechanics and diffusion, and then the leakage risk distribution data is obtained. This makes it possible to quickly and accurately determine the specific location of the leakage source, as well as key information such as the diffusion rate and concentration gradient of the leaked gas when a dangerous goods leak occurs. Compared with the traditional leak detection method, this step can locate the leakage source more quickly and accurately, and win precious time for timely and effective emergency measures, which helps to minimize the harm that may be caused by leakage accidents and ensure personnel safety and environmental pollution. With the help of the preset neural network model in the on-board edge computing unit, the leakage risk distribution data is analyzed for leakage risk prediction to obtain the situation evolution prediction data. The neural network model has powerful data processing and analysis capabilities, and can fully explore the potential laws and trends in the leakage risk distribution data, so as to accurately predict the situation evolution of dangerous goods leakage accidents. This enables managers to predict the development direction and possible severity of accidents in advance, formulate corresponding response strategies in advance, enhance the active control capabilities of dangerous goods logistics processes, improve the scientificity and effectiveness of emergency response, and further reduce the risk of accidents. According to the situation evolution prediction data and the response characteristic data of the on-board actuator, the distributed control strategy is optimized to obtain the collaborative control instruction data to regulate the ventilation system, refrigeration system and neutralizer spray system. This process realizes the intelligent and coordinated control of the vehicle's internal environmental control system. It can accurately adjust the working status of ventilation, refrigeration and neutralizer spraying systems according to the specific situation and evolution trend of dangerous goods leakage, quickly and effectively reduce the concentration of dangerous gases in the car, improve the interior environment, and create favorable conditions for the safe evacuation of personnel and emergency disposal of accidents. At the same time, it also reduces the corrosion and damage of dangerous goods to vehicle equipment, ensuring the normal operation and service life of the vehicle. Based on the coordinated control instruction data and the preset navigation data, the emergency avoidance path is dynamically planned and processed to obtain the emergency disposal strategy data. This step can comprehensively consider factors such as the current position of the vehicle, surrounding road conditions, leakage risk distribution, and coordinated control instructions, dynamically plan the optimal emergency avoidance path, and generate corresponding emergency disposal strategies.This ensures that when a dangerous goods leakage accident occurs, the vehicle can quickly and safely leave the dangerous area to avoid secondary accidents such as collisions with other vehicles and personnel. At the same time, it is also convenient for rescue forces to arrive at the accident site in time to carry out rescue work, improve the overall efficiency and success rate of emergency response, and minimize the accident losses. In summary, the cargo logistics monitoring and management method based on the Internet of Things realizes all-round monitoring, precise positioning, situation prediction, intelligent control and dynamic planning of the dangerous goods logistics process through the close cooperation and synergy of each step, effectively improving the safety, reliability and emergency response capabilities of dangerous goods logistics, and providing strong technical support and guarantee for the safe operation of the dangerous goods logistics industry.

[0012] Preferably, the present invention further provides a cargo logistics monitoring and management system based on the Internet of Things, which is used to execute the above-mentioned cargo logistics monitoring and management method based on the Internet of Things, and the cargo logistics monitoring and management system based on the Internet of Things includes:

[0013] The dangerous goods state detection module is used to obtain the temperature distribution data and micro-deformation data on the surface of the dangerous goods container; based on the principle of thermodynamic equilibrium, the internal state of the container is non-contact inverted according to the temperature distribution data and micro-deformation data to obtain the dangerous goods state characteristic data, wherein the dangerous goods state characteristic data includes pressure trend data, temperature gradient data and chemical activity data;

[0014] The leakage risk positioning module is used to obtain gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; perform three-dimensional positioning processing on the leakage source based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtain leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data;

[0015] The risk situation prediction module is used to use the neural network model preset in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data;

[0016] The collaborative control optimization module is used to optimize the distributed control strategy based on the situation evolution prediction data and the response characteristic data of the on-board actuators obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spray system;

[0017] The emergency response planning module is used to dynamically plan the emergency avoidance path according to the collaborative control command data and the preset navigation data to obtain the emergency response strategy data.

[0018] The dangerous goods status detection module in the present invention can accurately obtain the characteristic data of the dangerous goods status, including pressure trend, temperature gradient and chemical activity, by acquiring the temperature distribution data and micro-deformation data of the container surface and performing non-contact inversion processing based on the principle of thermodynamic equilibrium, which provides a solid data basis for subsequent risk assessment and disposal, effectively improves the safety and reliability of dangerous goods transportation, and does not need to directly contact dangerous goods, avoiding possible safety risks. The leakage risk positioning module uses the spectral gas sensors at the four corners of the carriage to obtain gas spectrum array data, and combines the dangerous goods status characteristic data for three-dimensional positioning processing, which can quickly and accurately determine the specific location of the leakage source and the leakage risk distribution, including key information such as leakage source coordinates, diffusion rate and concentration gradient, so as to buy precious time for timely and effective emergency measures, and help to minimize the harm that may be caused by leakage accidents. The risk situation prediction module uses the neural network model in the on-board edge computing unit to predict and analyze the leakage risk distribution data, and obtain the situation evolution prediction data, so that managers can predict the development direction and possible severity of the accident in advance, formulate corresponding response strategies in advance, enhance the active control ability of the dangerous goods logistics process, and improve the scientificity and effectiveness of emergency disposal. The collaborative control optimization module optimizes the distributed control strategy based on the situation evolution prediction data and the response characteristic data of the on-board actuators, obtains the collaborative control command data, and accurately regulates the ventilation, refrigeration and neutralizer spray systems, which can quickly and effectively reduce the concentration of dangerous gases in the car, improve the interior environment, create favorable conditions for the safe evacuation of personnel and emergency disposal of accidents, and also reduce the corrosion and damage of dangerous goods to vehicle equipment. The emergency response planning module dynamically plans the emergency avoidance path based on the collaborative control command data and preset navigation data, and obtains the emergency response strategy data to ensure that when a dangerous goods leakage accident occurs, the vehicle can quickly and safely leave the dangerous area to avoid collisions with other vehicles and personnel and other secondary accidents. At the same time, it is also convenient for rescue forces to arrive at the accident site in time to carry out rescue work, improve the overall efficiency and success rate of emergency disposal, and minimize accident losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0020] Figure 1 The figure is a flow chart of the steps of the cargo logistics monitoring and management method based on the Internet of Things of the present invention;

[0021] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0022] Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0023] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0024] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0025] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0026] To achieve this, please refer to Figures 1 to 3 The present invention provides a cargo logistics monitoring and management method based on the Internet of Things, the method comprising the following steps:

[0027] Step S1: obtaining temperature distribution data and micro-deformation data on the surface of the dangerous goods container; performing non-contact inversion processing on the internal state of the container according to the temperature distribution data and the micro-deformation data based on the principle of thermodynamic equilibrium to obtain characteristic data of the state of the dangerous goods, wherein the characteristic data of the state of the dangerous goods includes pressure trend data, temperature gradient data and chemical activity data;

[0028] In the embodiment of the present invention, an infrared thermal imager and a high-precision laser displacement sensor are arranged on the outer surface of the dangerous goods container. The infrared thermal imager collects the temperature distribution data on the surface of the container. Taking the thermal imaging device with a resolution of 640×480 pixels as an example, its temperature resolution reaches ±0.05°C, the acquisition frequency is 10Hz, and the temperature field changes are recorded. The laser displacement sensor records the slight deformation of the container surface with a resolution of 0.01mm and a sampling frequency of 1kHz. Based on the principle of thermodynamic equilibrium, a thermal expansion model and a stress-strain relationship model of the container material are established, for example, using the following formula: σ=E·∈+α·(T-T0); wherein σ is stress, E is elastic modulus, ε is deformation, α is linear expansion coefficient, T is current temperature, and T0 is reference temperature. By inputting the collected temperature distribution and micro-deformation data into the model, the internal state inversion is performed in combination with the finite element simulation method. Assume that the container material is carbon steel, the elastic modulus is 200GPa, and the linear expansion coefficient is 1.2×10 -5 / ℃, the pressure trend inside the dangerous goods is calculated to be 2.5MPa, the temperature gradient is 12℃ / m, and the chemical activity is calculated to be 0.3mol / s through the chemical reaction rate based on the Arrhenius formula.

[0029] Step S2: acquiring gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; performing three-dimensional positioning processing on the leakage source position based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtaining leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data;

[0030] In the embodiment of the present invention, high-precision spectral gas sensors are arranged at the four corners of the car, and optical fiber transmission technology is used to collect gas spectrum array data in the air. The sensor detection range is 2-14μm, and the resolution is 0.01μm. Through the spectral decomposition algorithm, such as the least squares method to fit the spectral absorption peak, the type and concentration of the leaked gas are identified. For example, if a tank of a transport vehicle leaks, the sensor detects that the spectral peak absorption intensity is 5.3. After calculation, the absorption peak corresponds to benzene, and the concentration is 0.15kg / m 3 . Input the pressure trend data, temperature gradient data and gas spectrum array data obtained in step S1 into the fluid mechanics diffusion model. The model is based on the Navier-Stokes equations and Fourier diffusion law, taking into account turbulent diffusion and steady-state diffusion characteristics, and calculates the location and diffusion rate of the leakage source. Assuming the ambient wind speed is 3m / s, the model calculates that the leakage source is located on the left side of the rear of the car (X=1.2m, Y=2.3m, Z=0.8m), the diffusion rate is 0.85m / s, and the concentration gradient is 0.06kg / m 3 / m.

[0031] Step S3: using a preset neural network model in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data;

[0032] The embodiment of the present invention inputs the leakage risk distribution data generated in step S2 into the vehicle-mounted edge computing unit, which has a built-in deep convolutional neural network model (CNN) trained with a massive leakage data set. The model includes 16 convolutional layers and 8 fully connected layers, and uses ReLU activation function and cross entropy loss function. The input data includes the coordinates of the leakage source, diffusion rate, concentration gradient, and real-time environmental parameters (such as wind speed, wind direction, and humidity). Through the multi-layer feature extraction and situation evolution simulation of the neural network, the leakage diffusion range, concentration change trend and high-risk areas within the next hour are output. For example, in the current scenario, the area of ​​the leakage risk area is 50m 2 The model predicts that the diffusion range will expand to 160m after 30 minutes. 2 , the concentration in high concentration area reaches 0.25kg / m 3 .

[0033] Step S4: performing distributed control strategy optimization processing according to the situation evolution prediction data and the response characteristic data of the vehicle-mounted actuator obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spraying system;

[0034] The embodiment of the present invention optimizes the distributed control strategy using a genetic algorithm based on the situation evolution data predicted in step S3, combined with the dynamic response characteristic data of the vehicle ventilation system, refrigeration system and neutralizer spraying system. The optimization goal is to minimize the diffusion range and concentration of the leaked gas. Taking the ventilation system as an example, the maximum flow rate of the fan is 800m 3 / h, the response delay is 2 seconds; the refrigeration system can reduce the regional temperature by 5-10℃, the neutralizer spraying system has a single spraying volume of 10L, and each spraying time is 5 seconds. The optimization algorithm sets the population size to 50, the maximum number of iterations to 100, and the final output control strategy is: the ventilation system runs at 90% power, the neutralizer spraying system sprays once every 30 seconds, the refrigeration system reduces the temperature by 7℃, and the system control response time is 1 second.

[0035] Step S5: Dynamically plan the emergency avoidance path according to the collaborative control instruction data and the preset navigation data to obtain emergency response strategy data.

[0036] The embodiment of the present invention uses the collaborative control instruction data and the geographic information data of the vehicle navigation system as input, and uses the dynamic programming method based on the Dijkstra algorithm to generate an emergency avoidance path. The input data includes the real-time leakage area coordinates, the scope of the high-risk area, the surrounding road network structure and the traffic conditions. The algorithm sets the weight function as the sum of the leakage area risk index and the road travel time. For example, in a certain transport vehicle leakage accident, there are 3 surrounding roads to choose from: the first one has the shortest distance but needs to pass through the high-risk area, with a risk index of 0.9; the second path has a longer detour distance but completely avoids the high-risk area, with a risk index of 0.3; the third path has a moderate distance but the risk index of some areas is 0.6. Finally, the second path is selected, with a total length of 14.5km and an estimated avoidance time of 18 minutes. The system replans the path every 1 minute according to the real-time situation update to ensure the safety and efficiency of the avoidance action.

[0037] The present invention can accurately obtain the state characteristic data of dangerous goods, including pressure trend data, temperature gradient data and chemical activity data, by acquiring the temperature distribution data and micro-deformation data on the surface of the dangerous goods container, and then performing non-contact inversion processing based on the principle of thermodynamic equilibrium. This process does not require direct contact with the inside of the dangerous goods container, avoiding the safety risks that may be caused by contact. At the same time, it can grasp the state changes of dangerous goods in the transportation process in real time and accurately, providing a solid data foundation for subsequent risk assessment and disposal, and effectively improving the safety and reliability of dangerous goods transportation. The gas spectrum array data is obtained by using the spectral gas sensors at the four corners of the carriage, and combined with the state characteristic data of dangerous goods, the position of the leakage source is three-dimensionally positioned according to the principle of fluid mechanics and diffusion, and then the leakage risk distribution data is obtained. This makes it possible to quickly and accurately determine the specific location of the leakage source, as well as key information such as the diffusion rate and concentration gradient of the leaked gas when a dangerous goods leak occurs. Compared with the traditional leak detection method, this step can locate the leakage source more quickly and accurately, and win precious time for timely and effective emergency measures, which helps to minimize the harm that may be caused by leakage accidents and ensure personnel safety and environmental pollution. With the help of the preset neural network model in the on-board edge computing unit, the leakage risk distribution data is analyzed for leakage risk prediction to obtain the situation evolution prediction data. The neural network model has powerful data processing and analysis capabilities, and can fully explore the potential laws and trends in the leakage risk distribution data, so as to accurately predict the situation evolution of dangerous goods leakage accidents. This enables managers to predict the development direction and possible severity of accidents in advance, formulate corresponding response strategies in advance, enhance the active control capabilities of dangerous goods logistics processes, improve the scientificity and effectiveness of emergency response, and further reduce the risk of accidents. According to the situation evolution prediction data and the response characteristic data of the on-board actuator, the distributed control strategy is optimized to obtain the collaborative control instruction data to regulate the ventilation system, refrigeration system and neutralizer spray system. This process realizes the intelligent and coordinated control of the vehicle's internal environmental control system. It can accurately adjust the working status of ventilation, refrigeration and neutralizer spraying systems according to the specific situation and evolution trend of dangerous goods leakage, quickly and effectively reduce the concentration of dangerous gases in the car, improve the interior environment, and create favorable conditions for the safe evacuation of personnel and emergency disposal of accidents. At the same time, it also reduces the corrosion and damage of dangerous goods to vehicle equipment, ensuring the normal operation and service life of the vehicle. Based on the coordinated control instruction data and the preset navigation data, the emergency avoidance path is dynamically planned and processed to obtain the emergency disposal strategy data. This step can comprehensively consider factors such as the current position of the vehicle, surrounding road conditions, leakage risk distribution, and coordinated control instructions, dynamically plan the optimal emergency avoidance path, and generate corresponding emergency disposal strategies.This ensures that when a dangerous goods leakage accident occurs, the vehicle can quickly and safely leave the dangerous area to avoid secondary accidents such as collisions with other vehicles and personnel. At the same time, it is also convenient for rescue forces to arrive at the accident site in time to carry out rescue work, improve the overall efficiency and success rate of emergency response, and minimize the accident losses. In summary, the cargo logistics monitoring and management method based on the Internet of Things realizes all-round monitoring, precise positioning, situation prediction, intelligent control and dynamic planning of the dangerous goods logistics process through the close cooperation and synergy of each step, effectively improving the safety, reliability and emergency response capabilities of dangerous goods logistics, and providing strong technical support and guarantee for the safe operation of the dangerous goods logistics industry.

[0038] Preferably, step S1 comprises the following steps:

[0039] Step S11: acquiring real-time temperature data collected by a distributed temperature sensor array on the surface of the dangerous goods container;

[0040] In the embodiment of the present invention, a distributed temperature sensor array is arranged on the surface of the dangerous goods container, for example, a high-precision NTC thermistor sensor is used, the temperature measurement accuracy of each sensor is ±0.1°C, and the arrangement density is 4 sensors per square centimeter, covering the entire container surface. The real-time temperature data is recorded at a sampling frequency of 1Hz through the on-board data acquisition module and transmitted to the processing unit for storage. For example, when the ambient temperature is 25°C, the collected data of a certain area on the surface of the container is (T1=30.2°C, T2=31.0°C, T3=29.8°C), and the acquisition module can encapsulate the data into a time series format for subsequent processing.

[0041] Step S12: performing temperature field interpolation and reconstruction processing according to the real-time temperature acquisition data, thereby obtaining temperature distribution data on the container surface;

[0042] The embodiment of the present invention inputs the real-time temperature data collected in step S11 into the two-dimensional temperature field interpolation model, for example, using the Laplace interpolation method to reconstruct the temperature field. The model is based on the assumption of temperature gradient continuity and combines the finite difference method to construct the interpolation equation. The boundary condition is the fixed temperature of the edge of the container surface. The collected temperature data (30.2°C, 31.0°C, 29.8°C) is used as the reference point, and the temperature distribution matrix of the container surface is obtained after interpolation, for example, 0.1m 2 The interpolation result in the grid is T ij The matrix has the highest temperature point of 31.5℃ and the lowest temperature point of 29.5℃, and the temperature difference distribution shows a linear gradient change.

[0043] Step S13: Acquire wavelength data collected by the fiber grating strain sensor on the surface of the dangerous goods container;

[0044] In the embodiment of the present invention, a fiber Bragg grating strain sensor is attached to the surface of a dangerous goods container. The spacing between each fiber Bragg grating is 5 cm, the sensor wavelength range is 1525-1565 nm, and the sensitivity is 1 pm / με (strain unit: microstrain). The wavelength data is collected in real time by a demodulation device (such as a spectrum analyzer) and recorded in a data processing unit. For example, when a container is affected by changes in internal pressure, the fiber wavelength changes from 1530.0 nm to 1530.8 nm, indicating that the strain generated at this position is 800 με.

[0045] Step S14: demodulating the wavelength acquisition data to obtain optical fiber strain data;

[0046] In the embodiment of the present invention, the wavelength data collected in step S13 is input into a demodulation algorithm for processing, and the Bragg grating strain demodulation formula is adopted: Among them, Δλ is the wavelength change, λ B is the Bragg wavelength, n eff is the effective refractive index of the optical fiber, E is the elastic modulus, and σ is the stress. The strain data is calculated by the formula. For example, if the wavelength change is 0.8nm, the strain obtained after demodulation is 800με. After the demodulation process is completed, the strain data is stored in a matrix form for subsequent analysis.

[0047] Step S15: performing strain field reconstruction calculation processing according to the optical fiber strain data, thereby obtaining micro-deformation data of the container surface;

[0048] In the embodiment of the present invention, the optical fiber strain data is input into the strain field reconstruction algorithm, for example, the strain field is constructed based on the radial basis function interpolation method (RBF). Each optical fiber sensor in the model is used as an interpolation point, and the three-dimensional strain field distribution of the container surface is generated according to the strain data. For example, the interpolation result shows that the maximum strain in a certain local area is 1200με, and the corresponding stress concentration point is located in the upper seam area of ​​the container. The strain field shows a typical non-uniform distribution characteristic, indicating the existence of a potential weak structure.

[0049] Step S16: Based on the principle of thermodynamic equilibrium, non-contact inversion processing is performed on the internal state of the container according to the temperature distribution data and the micro-deformation data, so as to obtain the characteristic data of the state of dangerous goods, wherein the characteristic data of the state of dangerous goods includes pressure trend data, temperature gradient data and chemical activity data.

[0050] In the embodiment of the present invention, the temperature distribution data and strain field data obtained in step S12 and step S15 are combined with the principle of thermodynamic equilibrium to perform the inversion of the internal state of the container. Finite element analysis (FEA) tools, such as ANSYS or ABAQUS, are used to construct a multi-physics field coupling model inside the container, and the material parameters of the container (such as the elastic modulus of carbon steel 200 GPa, the linear expansion coefficient 1.2×10-5 / ℃), load the temperature distribution and strain field as boundary conditions, and obtain the internal pressure trend, temperature gradient and chemical activity data through iterative solution. For example, the calculation results show that the pressure trend inside the container is 2.8MPa, the temperature gradient is 10℃ / m, and the chemical activity data show that the decomposition rate of dangerous goods is 0.4mol / s. These data can be used to evaluate the safety status of the container in real time and provide a basis for subsequent control decisions.

[0051] The implementation of the present invention brings many significant effects to the monitoring and management of dangerous goods logistics. First, by acquiring the real-time temperature acquisition data of the distributed temperature sensor array, the temperature change on the surface of the dangerous goods container can be fully and real-time grasped, providing an accurate data basis for the subsequent temperature field reconstruction and container internal state analysis. Then, the temperature field interpolation reconstruction process is performed to make the temperature data more complete and continuous in space, which helps to more accurately depict the temperature distribution on the surface of the container, thereby providing a strong basis for judging the thermal state and potential risks of dangerous goods. At the same time, the wavelength acquisition data of the fiber grating strain sensor is obtained, and the fiber strain data is demodulated to obtain the fiber strain data. This process can accurately capture the slight deformation of the container surface. Then, by performing strain field reconstruction calculation processing on the fiber strain data, the micro-deformation data of the container surface is obtained, which provides key information for analyzing the structural integrity and stress conditions of the container, helps to timely discover possible damage or deformation of the container, and take corresponding protective measures in advance. Finally, based on the principle of thermodynamic equilibrium, the internal state of the container is non-contact inverted in combination with the temperature distribution data and the micro-deformation data, and the state characteristic data of dangerous goods including pressure trend data, temperature gradient data and chemical activity data are obtained. This process achieves non-contact, all-round monitoring of the internal status of dangerous goods containers, without the need for direct contact with dangerous goods, thus avoiding possible safety risks. At the same time, it can grasp the various status changes of dangerous goods during transportation in real time and accurately, providing scientific and accurate data support for the safe transportation and emergency management of dangerous goods logistics, effectively improving the safety and reliability of dangerous goods logistics, and ensuring the safety of personnel and the environment.

[0052] Preferably, step S16 comprises the following steps:

[0053] Step S161: performing thermal stress calculation processing on the container wall according to the temperature distribution data, thereby obtaining thermal stress distribution data;

[0054] The embodiment of the present invention inputs the temperature distribution data obtained in step S12 into the thermal stress calculation model, and calculates the thermal stress distribution of the container wall based on the theory of thermoelasticity. A finite element analysis (FEA) tool, such as ANSYS, is used to construct a heat conduction and stress coupling field model by inputting the container material properties (such as the elastic modulus of carbon steel 200GPa, Poisson's ratio 0.3, thermal conductivity 50W / (m·K)) and temperature gradient data (such as the temperature distribution range of 30°C to 80°C, and the gradient is 5°C / m). The simulation results show that the thermal stress distribution caused by the temperature difference on the surface of the container wall forms a high stress concentration point in the local area, and the maximum thermal stress value reaches 50MPa, which is located at the upper and lower seams of the container. This data provides a basis for the subsequent analysis of the comprehensive stress state of the container.

[0055] Step S162: superimposing and analyzing the micro-deformation data and thermal stress distribution data of the container surface to obtain the comprehensive stress state data of the container wall;

[0056] In the embodiment of the present invention, the micro-deformation data obtained in step S15 and the thermal stress distribution data obtained in step S161 are input into the superposition analysis module. The superposition principle is adopted to superimpose the mechanical stress and thermal stress by the vector summation method to construct a comprehensive stress state data model of the container wall. The stress field superposition algorithm is written using Matlab, and the mechanical stress corresponding to the micro-deformation is set as σ m (e.g. the local maximum value is 30MPa), the thermal stress is σ t (For example, the local maximum value is 50MPa), the comprehensive stress is σ c =σ m +σ t The calculation results show that the maximum value of the comprehensive stress of the container wall is 80MPa and the minimum value is 10MPa. The comprehensive stress distribution diagram shows that the stress concentration area is mainly at the joints in the high temperature area.

[0057] Step S163: performing elastic mechanics inversion calculation processing according to the comprehensive stress state data of the container wall, thereby obtaining the internal pressure data of the container;

[0058] The embodiment of the present invention inputs the comprehensive stress state data of the container wall into the elastic mechanics inversion model and uses the inverse finite element analysis (InverseFEA) to invert the internal pressure of the container. According to the geometric parameters of the container (such as thickness 10mm, diameter 1.2m) and material properties, a container force boundary condition model is constructed. Taking the maximum comprehensive stress value of 80MPa as input, the Poisson relationship and elastic mechanics formula are used to calculate the internal pressure of the container. The internal pressure p is calculated in reverse, where D is the container diameter and t is the wall thickness. The calculation results show that the internal pressure of the container is 2.5 MPa. This data shows that the inside of the container is in a medium pressure state, which is suitable for further analysis of the thermodynamic state.

[0059] Step S164: performing thermodynamic state equation calculation processing according to the temperature distribution data and the internal pressure data of the container, thereby obtaining the temperature field data of the dangerous goods;

[0060] The embodiment of the present invention inputs the temperature distribution data obtained in step S12 and the internal pressure data in step S163 into the thermodynamic state equation model to calculate the temperature field of the dangerous goods in the container. The state equation PV = nRT (where P is pressure, V is volume, n is the amount of substance, R is the gas constant, and T is temperature) is combined with the temperature gradient equation (where q is the heat flux and k is the thermal conductivity) Solve. Input the container volume 1m 3 The specific heat capacity of dangerous goods is 4.2 J / (g·K). The temperature field data is calculated under the condition of a pressure of 2.5 MPa. For example, the temperature at the center of the container is 65°C, and the temperature near the wall is 45°C. The temperature difference distribution conforms to the linear decreasing law.

[0061] Step S165: performing thermodynamic equilibrium analysis on the temperature field data of the dangerous goods, thereby obtaining chemical activity prediction data of the dangerous goods;

[0062] The embodiment of the present invention inputs the temperature field data of the dangerous goods obtained in step S164 into the thermodynamic equilibrium model, and predicts the chemical activity by analyzing the relationship between temperature change and chemical reaction rate. (where k is the reaction rate constant, A is the frequency factor, E a =Activation energy, R is the gas constant, T is the temperature) to calculate the chemical reaction rate of dangerous goods. At a temperature of 65°C, the activation energy is 50 kJ / mol, and the calculated reaction rate constant is 1.2×10 -3 s -1 Further analysis yields predicted data on the chemical activity of hazardous materials, including decomposition rates and heat released at different temperatures.

[0063] Step S166: extracting features from the container internal pressure data and the hazardous material chemical activity prediction data, thereby obtaining hazardous material state feature data.

[0064] In the embodiment of the present invention, the internal pressure data of the container obtained in step S163 and the chemical activity prediction data in step S165 are input into the feature extraction module. The principal component analysis (PCA) is used to reduce the dimension of the multidimensional features and extract the feature quantity that best reflects the state of the dangerous goods. The input feature data includes pressure (2.5MPa), reaction rate (1.2×10 -3 s -1), temperature gradient (20℃ / m), etc., and the main components of the characteristic data of the state of dangerous goods are obtained after dimensionality reduction. For example, the pressure trend data shows that the pressure of the container gradually increases, the temperature gradient data shows that the heat flow gradually decreases from the center to the outside, and the chemical activity data predicts that the decomposition rate of dangerous goods will gradually accelerate. These characteristic data are used for real-time monitoring and risk assessment, providing a reliable basis for decision-making.

[0065] The present invention obtains thermal stress distribution data by calculating and processing the thermal stress of the container wall according to the temperature distribution data, which helps to deeply understand the stress state of the container wall under different temperature conditions, thereby evaluating the structural integrity and safety of the container. Then, the micro-deformation data and thermal stress distribution data of the container surface are superimposed and analyzed to obtain the comprehensive stress state data of the container wall. This process can more comprehensively reflect the actual stress condition of the container wall and provide a more accurate basis for the subsequent elastic mechanics inversion calculation. Further, elastic mechanics inversion calculation is performed based on the comprehensive stress state data of the container wall to obtain the internal pressure data of the container. This key data is crucial for real-time monitoring of the pressure change inside the container, and can timely discover potential pressure anomalies and prevent the container from rupturing due to overpressure. Subsequently, the thermodynamic state equation is calculated in combination with the temperature distribution data and the internal pressure data of the container to obtain the temperature field data of the dangerous goods, which helps to more accurately grasp the temperature distribution of the dangerous goods in the container and provide a basis for evaluating the thermal stability of the dangerous goods. In addition, the temperature field data of the dangerous goods is subjected to thermodynamic equilibrium analysis to obtain the chemical activity prediction data of the dangerous goods. This prediction data can warn in advance of possible chemical reactions of the dangerous goods, provide a scientific basis for taking preventive measures, and effectively reduce the risk of chemical accidents. Finally, the internal pressure data of the container and the predicted data of the chemical activity of dangerous goods are extracted to obtain the characteristic data of the state of dangerous goods. These data comprehensively reflect the various state changes of dangerous goods during transportation, and provide comprehensive and accurate information support for the refined management and emergency disposal of dangerous goods logistics. In summary, these steps realize the comprehensive monitoring and evaluation of the state of dangerous goods containers through scientific calculation and analysis methods, which not only improves the safety and reliability of dangerous goods transportation, but also provides strong data support for emergency management and decision-making.

[0066] Preferably, step S2 comprises the following steps:

[0067] Step S21: Acquire gas spectrum array data of spectrum gas sensors at four corners of the vehicle compartment;

[0068] The embodiment of the present invention arranges spectral gas sensors (such as mid-infrared laser absorption spectrometers) at the four corners of the car to collect gas spectrum array data from all corners of the car in real time. These sensors can detect absorption spectra within a specific wavelength range with high sensitivity, covering the characteristic bands of target gases (such as ammonia, methane, carbon monoxide, etc.). For example, in the case of a spectral range of 2.5μm to 4.5μm, the sensor collects 100 sets of spectral data per second with a resolution of 0.01nm. The timestamp and sensor position of each set of data are synchronously recorded by the data acquisition system to provide complete spatiotemporal distribution information for subsequent processing.

[0069] Step S22: performing baseline correction processing on the gas spectrum array data to obtain standardized spectrum data;

[0070] The embodiment of the present invention inputs the gas spectrum array data collected in step S21 into the baseline correction model to correct the spectrum baseline drift. First, a low-pass filtering algorithm is used to remove high-frequency noise, and then a polynomial fitting method (such as a third-order polynomial fitting) is used to correct the background signal of the spectrum, thereby eliminating the baseline offset caused by ambient light, equipment performance changes, etc. For example, the original spectrum baseline offset of a sensor is 0.1 to 0.5 au (absorbance unit). After correction, the spectrum baseline is restored to the standard range (near 0 a.u). The corrected standardized spectrum data can more accurately reflect the gas concentration and characteristic information.

[0071] Step S23: performing multi-component analysis through standardized spectral data and flow field modeling processing to obtain airflow field data in the vehicle compartment;

[0072] The embodiment of the present invention uses the standardized spectral data of step S22, and uses a multi-component spectral analysis algorithm (such as least squares method or principal component regression analysis) to separate the various gas components and their concentrations in the compartment. For example, in a mixed gas containing methane, ammonia and carbon dioxide, the wavelength position of the characteristic absorption peak of each component (such as methane at 3.31 μm, ammonia at 3.00 μm) and absorbance are used to calculate the concentration, and methane is measured to be 10 ppm, ammonia is 20 ppm, and carbon dioxide is 5 ppm. Subsequently, based on the gas concentration distribution data, combined with the flow field modeling method (such as CFD computational fluid dynamics modeling), the compartment geometric parameters (such as length 10m, width 2.5m, height 2.5m) and the compartment ventilation rate (such as 1m / s) are input to reconstruct the airflow field in the compartment, and the flow field velocity distribution, vortex area and gas concentration gradient distribution diagram are obtained.

[0073] Step S24: Perform three-dimensional positioning processing on the leakage source based on fluid mechanics and diffusion according to the dangerous goods status characteristic data and the airflow field data in the car to obtain leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data.

[0074] In the embodiment of the present invention, the airflow field data obtained in step S23 and the state characteristic data of the dangerous goods (such as pressure trend, chemical activity, etc.) are input into the leakage source positioning model, and three-dimensional positioning processing is performed based on the fluid mechanics diffusion theory and reverse simulation technology. Through CFD simulation software (such as Fluent), the initial leakage rate (such as 5g / s) and diffusion coefficient (such as 1×10 -5 m 2 / s), build a leakage diffusion model, and simulate the impact of different initial leakage positions on the gas concentration distribution in the car. Use the concentration gradient to invert the three-dimensional coordinates of the leakage source (such as 2.5m, 1.0m, 0.5m), and calculate the diffusion rate (such as 2m / s) and the maximum concentration gradient (such as 10ppm / m). The output leakage risk distribution data includes a visualization chart of high-risk areas to guide emergency treatment and risk control.

[0075] The present invention can comprehensively and real-time monitor the gas composition and concentration changes in the compartment by acquiring the gas spectrum array data of the spectrum gas sensors at the four corners of the compartment. The spectrum gas sensor has the advantages of high sensitivity, high selectivity, low cost, etc., and provides a new solution for gas detection and monitoring. Then, the gas spectrum array data is baseline corrected to obtain standardized spectrum data. This process can eliminate the background spectrum drift caused by the change of the measurement environment, improve the accuracy and reliability of the spectrum data, and thus provide a more accurate data basis for subsequent multi-component analysis and flow field modeling. Further, multi-component analysis is performed through standardized spectrum data, and flow field modeling is performed to obtain the airflow field data in the compartment. This process can accurately analyze the concentration distribution of various gases in the compartment, and combined with the airflow field data, fully understand the diffusion of gases in the compartment, and provide key information for the location and risk assessment of the leakage source. Finally, according to the state characteristic data of dangerous goods and the airflow field data in the compartment, the leakage source position is three-dimensionally located based on fluid mechanics diffusion to obtain leakage risk distribution data. This includes leak source coordinate data, diffusion rate data, and concentration gradient data, so that when a dangerous goods leak occurs, the specific location of the leak source, as well as the diffusion range and speed of the leaked gas, can be quickly and accurately determined, so that effective emergency measures can be taken in a timely manner to minimize the harm that may be caused by the leak accident and ensure the safety of personnel and the environment from pollution. In summary, these steps have achieved all-round monitoring and precise positioning of dangerous goods leaks through scientific calculation and analysis methods, which not only improves the safety and reliability of dangerous goods transportation, but also provides strong data support for emergency management and decision-making.

[0076] Preferably, step S23 includes the following steps:

[0077] Step S231: performing multi-component spectrum analysis processing according to the standardized spectrum data, thereby obtaining gas concentration data of each component;

[0078] The embodiment of the present invention inputs the standardized spectral data into a multi-component spectral analysis algorithm (such as the full spectrum least squares method or the non-negative matrix decomposition algorithm) to analyze the concentration of each component gas. Taking the infrared absorption spectrum as an example, the concentration data of each component is inverted using the Lambert-Beer law (absorbance = absorption coefficient × concentration × path length) according to the position of the characteristic absorption peak (such as 3.31 μm for methane and 3.00 μm for ammonia) and the absorbance intensity during analysis. Assuming that the path length is 10 cm and the measured absorbance is 0.1, the analysis shows that the methane concentration is 50 ppm and the ammonia concentration is 25 ppm. By processing the spectral data at each moment point by point, a set of time-series gas concentration data is formed, which lays the foundation for subsequent analysis.

[0079] Step S232: performing time series comparison analysis on the gas concentration data of each component, thereby obtaining concentration change trend data;

[0080] The embodiment of the present invention arranges the gas concentration data of each component obtained in step S231 in chronological order, and calculates the concentration change trend using a time series analysis method (such as a sliding average method or a linear regression method). For example, for ammonia concentration, continuous data points are extracted at a sampling interval of 1 second, and a concentration change curve over time is fitted. The results show that the concentration rise rate is 2ppm / s within 0-10 seconds and remains stable within 10-20 seconds. By comparing and analyzing the concentration trend data, possible leakage moments and related events can be identified, providing input for spatial distribution analysis.

[0081] Step S233: performing spatial interpolation processing according to the concentration change trend data of the four corners of the carriage, thereby obtaining preliminary gas concentration distribution data;

[0082] The embodiment of the present invention uses the concentration change trend data of the four corners of the car as input, and uses a spatial interpolation method (such as Kriging interpolation or inverse distance weighted method) to make a preliminary estimate of the gas concentration distribution in the car. For example, assuming that the dimensions of the car are 10m long, 2.5m wide, and 2.5m high, and the concentration data of the four corners are known to be 10ppm, 15ppm, 20ppm, and 25ppm, respectively, a spatial model is established through the Kriging interpolation algorithm, and the concentration at the central point of the car (5m, 1.25m, 1.25m) is predicted to be 18ppm. The interpolation results are presented in the form of a three-dimensional concentration distribution map, providing initial conditions for flow field modeling.

[0083] Step S234: Obtain the air pressure sensor data and temperature sensor data in the vehicle compartment, and perform flow field modeling processing based on the air pressure sensor data and the temperature sensor data to obtain the air flow field data in the vehicle compartment.

[0084] The embodiment of the present invention collects real-time data through air pressure sensors and temperature sensors arranged in the car, for example, the air pressure is 101.3kPa and the temperature is 25°C, and preprocesses these data (such as filtering and calibration). The preprocessed air pressure data and temperature data are input into the flow field modeling software (such as ANSYS Fluent), and a CFD (computational fluid dynamics) model is established in combination with the internal structural geometry information of the car. Assuming that the ventilation rate in the car is 1m / s and the inlet velocity is 5m / s, the airflow field distribution data in the car is calculated through grid division, boundary condition setting and turbulence model selection (such as k-epsilon model), including velocity vector field, vortex area and pressure distribution. These data provide the necessary physical field support for further accurate prediction of the gas leakage location.

[0085] The present invention can accurately obtain the gas concentration data of each component by performing multi-component spectral analysis processing according to standardized spectral data. This process utilizes the high sensitivity and high selectivity of spectral analysis technology, can effectively distinguish and quantify the concentrations of multiple gases in the car, and provides basic data for subsequent analysis. Then, the time series comparison analysis processing is performed according to the gas concentration data of each component to obtain the concentration change trend data. This analysis helps to understand the change of gas concentration over time, so as to timely discover potential leakage or abnormal conditions and take measures in advance. Further, spatial interpolation processing is performed according to the concentration change trend data of the four corners of the car to obtain preliminary gas concentration distribution data. Spatial interpolation processing can expand discrete concentration data to the entire car space to form a continuous concentration distribution map, which provides a spatial reference for the precise positioning of the leakage source. Finally, the air pressure sensor data and temperature sensor data in the car are obtained, and the flow field modeling processing is performed according to these data to obtain the air flow field data in the car. The air flow field data can reflect the flow of gas in the car. Combined with the concentration distribution data, the diffusion path and speed of the gas can be more accurately predicted, thereby providing comprehensive data support for the three-dimensional positioning and risk assessment of the leakage source. In summary, these steps have achieved all-round monitoring and precise positioning of hazardous goods leakage through scientific calculation and analysis methods, which not only improves the safety and reliability of hazardous goods transportation, but also provides strong data support for emergency management and decision-making.

[0086] Preferably, step S24 comprises the following steps:

[0087] Step S241: performing theoretical diffusion rate calculation processing on the dangerous goods status characteristic data, thereby obtaining expected diffusion parameter data;

[0088] The embodiment of the present invention uses molecular diffusion theory to calculate the diffusion rate based on the characteristic data of the dangerous goods state (including pressure, temperature gradient and chemical activity). Taking ammonia as an example, when the temperature is 25°C and the pressure is 101.3 kPa, the diffusion rate is calculated using Fick's first law: The diffusion coefficient D can be calculated by the Chapman-Enskog formula and is approximately 2.24×10 -5 m 2 / s. Assuming the concentration gradient is 0.5ppm / m, then the theoretical diffusion rate is J = -2.24×10 -5 ×0.5=-1.12×10 -5 ppm·m / s. According to the temperature and pressure differences in different regions, the expected diffusion parameters at each location are calculated to form three-dimensional diffusion parameter data.

[0089] Step S242: performing computational fluid dynamics simulation processing on the preliminary gas concentration distribution data and the air flow field data in the vehicle compartment, so as to obtain actual diffusion field data;

[0090] The embodiment of the present invention uses preliminary gas concentration distribution data and airflow field data in the car to construct a gas diffusion model in computational fluid dynamics (CFD) simulation software (such as ANSYS Fluent). The concentration data is used as the initial condition, the airflow field data is used as the boundary condition, and the gas diffusion process is simulated in combination with a turbulence model (such as the k-epsilon model). Assuming that the ventilation speed in the car is 1m / s and the inlet concentration is 10ppm, the actual diffusion field data is calculated through grid division and time step setting, including transient concentration distribution and diffusion rate field. The simulation shows that the concentration in the center of the car is 8ppm and the concentration in the tail is 5ppm, forming a three-dimensional concentration change graph.

[0091] Step S243: performing gas reverse tracking calculation processing according to the expected diffusion parameter data and the actual diffusion field data, thereby obtaining initial leakage source position data;

[0092] The embodiment of the present invention compares the expected diffusion parameter data with the actual diffusion field data, and calculates the initial source position of the gas leakage through a reverse tracking algorithm (such as an inverse diffusion model based on the particle tracking method). Assuming that the actual concentration at a certain point is higher than the expected diffusion parameter, the time difference between the concentration change rate at this point and the surrounding area is analyzed to infer the diffusion center point. Through interpolation calculation, it is found that the initial leakage source position is the left front of the car, with coordinates of approximately (1.2m, 0.8m, 0.5m).

[0093] Step S244: performing Kalman filtering-based optimization processing on the initial leakage source position data, thereby obtaining accurate leakage source coordinate data;

[0094] The embodiment of the present invention inputs the initial leakage source position data into the Kalman filter model, and optimizes the initial position in combination with the real-time gas concentration change and the airflow field data. The filtering process includes two steps: state prediction and state update. With the predicted position error of 0.2m and the actual observation error of 0.1m, after 5 iterative calculations, the final optimization results in the precise leakage source coordinates of (1.15m, 0.75m, 0.55m). The optimization result has higher accuracy and can be used for subsequent risk assessment.

[0095] Step S245: performing diffusion dynamics analysis and processing according to the precise leakage source coordinate data and the actual diffusion field data, thereby obtaining diffusion rate data;

[0096] The embodiment of the present invention uses accurate leak source coordinate data and actual diffusion field data, combined with diffusion kinetic equations (such as the exponential relationship between diffusion rate and time), to calculate the gas diffusion rate. Assume that the diffusion process conforms to the first-order kinetic model R = k·e -t / τ , where k is the initial diffusion rate and τ is the diffusion time constant. Taking ammonia as an example, the initial diffusion rate k = 1.5ppm / s, the diffusion time constant τ = 2.5s, and the change trend of the diffusion rate at different time points is calculated. The results show that the rate drops from 1.5ppm / s to 0.6ppm / s within 5 seconds.

[0097] Step S246: performing gradient calculation processing on the diffusion rate data and the concentration data of each component gas to obtain concentration gradient data; performing risk assessment processing based on the leakage source coordinate data, diffusion rate data and concentration gradient data to obtain leakage risk distribution data.

[0098] The embodiment of the present invention performs gradient calculation on the diffusion rate data and the gas concentration data of each component, and uses the finite difference method to solve the concentration change gradient. For example, assuming that the concentration difference between the center and the edge of the compartment is 5ppm and the distance is 2m, the concentration gradient is Combining the leakage source coordinate data and diffusion rate data, the risk assessment model (such as the fault tree analysis model) is used to calculate the leakage risk distribution to obtain the leakage risk distribution data, including the specific values ​​and locations of high-risk areas (such as within 2m around the leakage source) and low-risk areas, to facilitate the subsequent adoption of targeted emergency measures.

[0099] The present invention obtains expected diffusion parameter data by performing theoretical diffusion rate calculation processing on the state characteristic data of dangerous goods. This process is based on the physical model and the known characteristics of dangerous goods, and can predict the diffusion behavior under ideal conditions, providing a theoretical basis for the subsequent comparison of actual diffusion field data. Then, the preliminary gas concentration distribution data and the airflow field data in the car are simulated by computational fluid dynamics (CFD) to obtain the actual diffusion field data. CFD simulation can accurately simulate the diffusion process of gas in the car, taking into account the influence of multiple factors such as airflow, temperature, pressure, etc., so that the diffusion field data is closer to the actual situation and the accuracy of leakage monitoring is improved. Further, the gas reverse tracking calculation processing is performed according to the expected diffusion parameter data and the actual diffusion field data to obtain the initial leakage source position data. The reverse tracking method uses the known diffusion data to reverse the position of the leakage source, providing preliminary results for accurately locating the leakage source. Then, the initial leakage source position data is optimized based on Kalman filtering to obtain accurate leakage source coordinate data. The Kalman filter can effectively handle measurement noise and system errors, optimize the estimation accuracy of the leakage source position, and make the positioning of the leakage source more accurate and reliable. In addition, diffusion kinetics analysis and processing are performed based on the precise leakage source coordinate data and the actual diffusion field data to obtain diffusion rate data. Diffusion kinetics analysis can describe the speed and mechanism of gas diffusion in detail, providing a scientific basis for assessing leakage risks and formulating emergency measures. Finally, gradient calculation and processing are performed on the diffusion rate data and the gas concentration data of each component to obtain concentration gradient data, and risk assessment processing is performed based on the leakage source coordinate data, diffusion rate data and concentration gradient data to obtain leakage risk distribution data. Concentration gradient data can reflect the rate of change of gas concentration in space. Combined with other data for risk assessment, the potential hazards of leakage accidents can be fully understood, providing decision support for emergency response and accident handling. In summary, these steps have achieved all-round monitoring and precise positioning of hazardous goods leakage through scientific calculation and analysis methods, which not only improves the safety and reliability of hazardous goods transportation, but also provides strong data support for emergency management and decision-making.

[0100] Preferably, step S3 comprises the following steps:

[0101] Step S31: extracting features and normalizing the leakage risk distribution data to obtain neural network input feature data;

[0102] The embodiment of the present invention extracts features from the leakage risk distribution data (including leakage source coordinates, diffusion rate, and concentration gradient, etc.), uses the principal component analysis (PCA) method to reduce the dimension, and extracts key features such as the distance from the leakage source to the surrounding high-risk area, the diffusion rate change rate, and the concentration gradient change trend as feature vectors. Subsequently, data normalization is performed, and the Min-Max normalization method is used to map the data to the [0, 1] range. The formula is: Where X is the value of the original data point, that is, the data value that needs to be standardized, X min is the minimum leakage rate measured in historical data, X max is the maximum leakage rate measured in the historical data, X' is the standardized value, and the result is mapped to the specified range; for example, the leakage source coordinates (x, y, z) are standardized to (0.65, 0.3, 0.45). After processing, the neural network input feature data obtained includes the standardized feature vector and time series data, which provide input for subsequent model prediction.

[0103] Step S32: obtaining a neural network model preset in the vehicle-mounted edge computing unit, wherein the parameters of the neural network model include network structure data, weight coefficient data, and bias parameter data;

[0104] The embodiment of the present invention loads a preset neural network model from the vehicle-mounted edge computing unit. The model adopts a multi-layer perceptron (MLP) structure, including an input layer, two hidden layers, and an output layer. The parameters include network structure data (the number of input layer nodes is 10, the number of hidden layer nodes is 64 and 32 respectively, the activation function is ReLU, and the number of output layer nodes is 1), weight coefficient data (initialized to uniform distribution, the range is [-0.05, 0.05]), and bias parameter data (the initial value is set to 0). When loading the model, first check the latest model version in the device storage, verify its integrity, and then load it into the memory for subsequent calibration.

[0105] Step S33: Acquire historical leakage accident case data; and perform classification matching processing on the historical leakage accident case data based on the type of dangerous goods, thereby obtaining leakage accident classification data;

[0106] The embodiment of the present invention obtains historical leakage accident data from the accident case database, including the time of the accident, the type of hazardous materials, the location of the leakage and the consequences, etc. When classifying the data, it is first divided into different categories according to the type of hazardous materials (such as flammable gas, toxic gas, asphyxiating gas, etc.). A classification algorithm based on text features is used to convert the case description information (such as "liquefied natural gas leakage, regional temperature rise, and rapid risk diffusion") into a feature vector to match the corresponding hazardous materials category. After classification, the leakage accident classification data is obtained. For example, if a specific case belongs to the "methane leakage" category, its feature data is used for model calibration.

[0107] Step S34: performing online calibration processing on the parameters in the neural network model according to the leakage accident classification data, thereby obtaining a leakage risk prediction model;

[0108] The embodiment of the present invention uses leakage accident classification data to calibrate the neural network model online, and uses the Adam optimization algorithm to adjust the model parameters. The calibration process includes inputting historical case feature data, comparing the model output with the actual accident risk results, calculating the prediction error and updating the weight and bias parameters. Assuming that the initial model predicts a risk value of 0.8, while the actual risk value is 0.6, the error is 0.2, and the weight is updated through back propagation to make the predicted value closer to the actual value. After the calibration is completed, the average prediction error of the model is reduced from the initial 10% to 3%, and a calibrated leakage risk prediction model is obtained.

[0109] Step S35: Use the leakage risk prediction model to predict the leakage risk of the neural network input feature data, so as to obtain situation evolution prediction data.

[0110] The embodiment of the present invention inputs the neural network input feature data into the calibrated leakage risk prediction model to perform leakage risk prediction. The prediction process calculates the situation evolution prediction results through forward propagation, including the expansion rate of high-risk areas, the changing trend of the leakage impact range, etc. For example, the model output shows that in the next 5 minutes, the high-risk area will expand from the leakage point to within 5m, the concentration gradient will increase to 2ppm / m, and the diffusion rate will slow down to 0.8ppm / s. Combined with the prediction data, a visual report is generated to provide a decision-making basis for the safety management of the compartment environment.

[0111] The present invention obtains neural network input feature data by extracting features and normalizing data on leakage risk distribution data. This process can extract the most representative and valuable information from complex data, and at the same time, through standardization, ensure the consistency and comparability of data, providing high-quality input data for the training of neural network models.

[0112] Next, the preset neural network model in the vehicle-mounted edge computing unit is obtained, wherein the parameters of the neural network model include network structure data, weight coefficient data, and bias parameter data. The preset neural network model provides an infrastructure for subsequent risk prediction. By adjusting these parameters, the performance of the model can be optimized to better adapt to specific application scenarios. Furthermore, historical leakage accident case data is obtained, and classification matching processing based on the type of hazardous materials is performed to obtain leakage accident classification data. This process uses the experience and knowledge in the historical data to provide rich background information for model training, which helps to improve the accuracy and reliability of the model. Then, the parameters in the neural network model are calibrated online according to the leakage accident classification data to obtain a leakage risk prediction model. Online calibration can dynamically adjust the model parameters according to the latest data and information to ensure that the model is always in the best state, thereby improving the accuracy and timeliness of the prediction. Finally, the leakage risk prediction model is used to predict the leakage risk of the neural network input feature data to obtain the situation evolution prediction data. This process can predict the development trend and possible impact range of leakage accidents in advance, provide a scientific basis for emergency management and decision-making, and help relevant personnel take effective measures in time to reduce accident risks and losses. In summary, these steps, through scientific calculation and analysis methods, have achieved accurate prediction of the risk of hazardous goods leakage and early warning of situation evolution, which not only improves the safety and reliability of hazardous goods transportation, but also provides strong data support for emergency management and decision-making.

[0113] Preferably, step S4 comprises the following steps:

[0114] Step S41: Acquire response characteristic data of the ventilation system, the refrigeration system and the neutralizer spraying system;

[0115] The embodiment of the present invention obtains the response characteristic data of the ventilation system, the refrigeration system and the neutralizer spray system through the system monitoring platform or the installed sensors. These data include the startup delay time, the maximum response time, the response amplitude and the nonlinear characteristics in the adjustment process of each system. For example, the ventilation system may take 5 seconds to reach the maximum wind speed, the refrigeration system may take 3 minutes to cool down to the set temperature after startup, and the response characteristics of the neutralizer spray system may include spraying time, flow rate and spraying range. By monitoring these response characteristics in real time, the accuracy of the formulation of subsequent control instructions is ensured.

[0116] Step S42: Perform time series response analysis based on the situation evolution prediction data and the response characteristic data of each execution system, and perform optimization processing based on response time, execution sequence and response intensity to obtain the optimal control time series data.

[0117] The embodiment of the present invention analyzes the timing response of the ventilation system, refrigeration system and neutralizer spray system based on the response characteristic data obtained in step S41 and the situation evolution prediction data (such as leakage source changes, gas concentration trends, etc.) obtained from step S35. The analysis content includes the response time of the system, the priority execution order of the system and the response intensity. By establishing an optimization model, combining the timeliness requirements and responsiveness of the control system, an optimization algorithm (such as particle swarm optimization, genetic algorithm or ant colony optimization) is used to determine the optimal control timing. For example, if the concentration of the chemical substance increases rapidly, the neutralizer spray system should be started first, followed by adjusting the ventilation system to accelerate the dilution of the gas. The optimal control timing data finally obtained is the best order and time node that each system should execute.

[0118] Step S43: performing control parameter calculation processing on the ventilation system, the refrigeration system and the neutralizer spraying system respectively according to the optimal control time series data and the situation evolution prediction data, so as to obtain the adjustment instruction data of each system;

[0119] The embodiment of the present invention calculates control parameters for each execution system based on the optimal control timing data obtained in step S42 and in combination with the situation evolution prediction data obtained in step S35. First, according to the situation evolution prediction data (such as the changing trend of gas concentration, temperature change, etc.), the corresponding adjustment parameters are calculated for each system. For example, when the leakage source is close, the wind speed of the ventilation system may need to be increased; and when the temperature is high, the refrigeration system may need to start quickly and increase the cooling intensity. Finally, the specific adjustment instruction data for each system is calculated through the adjustment parameter formula, including wind speed, temperature setting value, neutralizer spraying amount, etc.

[0120] Step S44: performing system interference analysis and coordination optimization processing on the adjustment instruction data of each system to obtain preliminary coordinated control instruction data;

[0121] In this step, the embodiment of the present invention analyzes the interference effects between different systems. For example, increasing the wind speed of the ventilation system may cause the energy efficiency of the refrigeration system to decrease, or the spray effect of the neutralizer spray system may be weakened by the airflow. Through mathematical modeling and system simulation, the interaction of each system is evaluated, and the adjustment instructions of each system are optimized through a coordinated optimization algorithm (such as multi-objective optimization, game theory or system coordinated control algorithm) to ensure coordinated work between the systems, eliminate or reduce interference effects, and thus obtain preliminary collaborative control instruction data.

[0122] Step S45: performing control quantity constraint check processing on the coordinated control instruction data, thereby obtaining control instruction data that satisfies the physical constraints of the actuator;

[0123] The embodiment of the present invention performs constraint checks on the physical limitations of each actuator (such as the maximum wind speed of the ventilation system, the maximum cooling capacity of the refrigeration system, the spraying amount of the neutralizer spraying system, etc.). According to the actual hardware capabilities and safety range of the system, the preliminary collaborative control instruction data is adjusted. For example, if the control instruction requires the ventilation system wind speed to exceed the maximum wind speed, it is adjusted to the maximum value; if the spraying amount exceeds the maximum spraying capacity of the neutralizer spraying system, it is limited to the maximum output value of the system. Through this constraint check, it is ensured that the generated control instruction data can be correctly executed by each actuator.

[0124] Step S46: performing control sequence generation processing according to the control instruction data, thereby obtaining control instruction sequence data for time-sharing execution;

[0125] The embodiment of the present invention generates a control sequence based on the adjustment instruction data of each system obtained in step S45, combined with the optimal control timing data. The generation of the control sequence takes into account the response timeliness, operation sequence and physical constraints of the system, and reasonably allocates and sorts the control instructions through a timing scheduling algorithm (such as linear programming, dynamic programming or mixed integer programming). For example, within a certain period of time, the control instructions of each system need to be adjusted in stages to avoid system overload or waste of resources. The control sequence data finally generated includes the control timing and execution order of each system.

[0126] Step S47: Reliability evaluation and priority sorting are performed on the control instruction sequence data executed in time-sharing manner, thereby obtaining collaborative control instruction data.

[0127] The embodiment of the present invention performs reliability assessment on the generated control instruction sequence to ensure that the execution of each stage has sufficient reliability. For example, through redundant design and fault tolerance mechanism, the stability and response time of each execution system under different operating conditions are evaluated. If some systems may not work properly under the current operating conditions, the execution order is readjusted or the backup system is replaced according to the priority. At the same time, based on factors such as the priority of the system and the severity of the leakage risk, the control instruction sequence is prioritized to ensure that the most critical system is executed first. Finally, complete collaborative control instruction data is generated to command the various systems to work together.

[0128] By acquiring the response characteristic data of the ventilation system, the refrigeration system and the neutralizer spray system, the present invention can understand the working performance and reaction speed of each system under different conditions in detail, and provide basic data for the subsequent control strategy formulation. Then, according to the situation evolution prediction data and the response characteristic data of each execution system, the time series response analysis is performed, and the optimization processing is performed to obtain the optimal control time series data. This process ensures that each system can respond in the most appropriate order and intensity at the most appropriate time, thereby improving the efficiency and effect of emergency treatment. Further, the control parameters of each system are calculated according to the optimal control time series data and the situation evolution prediction data to obtain the adjustment instruction data of each system. This step enables each system to be accurately adjusted according to the actual situation, ensuring that the impact of the leakage of dangerous goods is quickly and effectively controlled in an emergency. Then, the adjustment instruction data of each system is subjected to system interference analysis and coordinated optimization processing to obtain preliminary collaborative control instruction data. This process takes into account the mutual influence between the systems, avoids conflicts and interferences between the systems, and ensures that the systems can work together to form a joint force. Next, the collaborative control instruction data is subjected to control quantity constraint check processing to obtain control instruction data that meets the physical constraints of the actuator. This step ensures the feasibility and safety of the control instructions during actual execution, and avoids system failures caused by exceeding the equipment capacity. Further, the control sequence generation process is performed according to the control instruction data to obtain the control instruction sequence data for time-sharing execution. This step decomposes the complex control task into orderly steps, ensuring the orderliness and controllability of the control process. Finally, the reliability evaluation and priority sorting of the control instruction sequence data for time-sharing execution are performed to obtain the collaborative control instruction data. This process further optimizes the control strategy, ensures that the most critical tasks can be executed first in an emergency, and improves the overall reliability and effectiveness of the emergency response. In summary, these steps realize the refined management and collaborative control of the emergency response to hazardous materials leakage through scientific calculation and analysis methods, which not only improves the safety and reliability of hazardous materials transportation, but also provides strong technical support for emergency management and decision-making.

[0129] Preferably, step S5 comprises the following steps:

[0130] Step S51: Obtaining GPS positioning data of the vehicle's current location, electronic map road network data, and preset emergency shelter location data and safety channel data;

[0131] The embodiment of the present invention obtains the current location information, including latitude, longitude, altitude and other data, through the vehicle's GPS positioning system. At the same time, the road network data of the electronic map is obtained through the vehicle-mounted equipment in the vehicle or the external communication network. These data include the type of road, name, traffic restrictions, section length and connection relationship of the road network. The vehicle's control system will also pre-set the location data of multiple emergency shelters (such as shelters, medical stations, etc.) and safe channel data (such as dedicated emergency channels, refuge routes, etc.). These data are usually obtained through real-time synchronization with relevant management platforms or map services to ensure that the data used is the latest and can be dynamically updated according to the current location of the vehicle.

[0132] Step S52: Perform real-time road condition analysis and processing based on GPS positioning data and electronic map road network data, thereby obtaining current road network traffic status data;

[0133] The embodiment of the present invention uses the vehicle's GPS positioning data and electronic map road network data to analyze the current road traffic status in real time. First, based on the vehicle's current position and real-time traffic information of surrounding roads (such as traffic cameras, road sensors, feedback data from other vehicles, etc.), analyze the road traffic flow, whether there is congestion, whether there are traffic accidents on the road, etc. The road conditions are evaluated through traffic data streams (such as the status of traffic lights, traffic police commands, accident information, etc.), thereby obtaining the current road network traffic status data. At this time, the system can output information including whether the road is unobstructed, the degree of congestion, traffic accidents or construction areas, etc., for reference in subsequent decision-making steps.

[0134] Step S53: performing dangerous area division processing on the leakage risk distribution data and the situation evolution prediction data, and performing emergency channel screening processing according to the current road network traffic status data, emergency shelter location data and safety channel data, so as to obtain available emergency channel data;

[0135] The embodiment of the present invention divides the dangerous area according to the leakage risk distribution data obtained from step S41 and the situation evolution prediction data obtained from step S35. For example, the dangerous area is divided into different levels (such as severe danger area, mild danger area, low risk area, etc.) using the changing trend of gas concentration, diffusion rate data, etc. Then, combined with the traffic status data of the current road network and the location data of emergency shelters, safe passages that can effectively evacuate people or transport materials in dangerous situations are screened out. In this process, the system will exclude infeasible paths according to the current traffic conditions (such as road closures, traffic control, etc.), and finally obtain available emergency passage data, including selectable passage paths, shelters, traffic restrictions, etc.

[0136] Step S54: Perform safety assessment and handling time estimation processing according to the collaborative control instruction data and the available emergency channel data, thereby obtaining channel safety level data and expected handling time data;

[0137] The embodiment of the present invention performs a safety assessment on each emergency channel based on the collaborative control instruction data generated in step S43 and the available emergency channel data obtained in step S53. This assessment mainly depends on multiple factors, such as road safety, congestion, traffic control, etc. At the same time, combined with other risks that may be caused by leakage accidents, the system analyzes the traffic safety of each channel and derives the safety level data of each channel (such as low, medium, high, etc.). In addition, based on the current road capacity, traffic density and emergency response capabilities, the expected disposal time of each emergency channel (i.e., the time required from entering the channel to successfully evacuating or resolving the danger) is estimated. These data will provide a basis for subsequent path optimization and emergency disposal.

[0138] Step S55: performing multi-objective optimization and traffic simulation analysis on the channel safety level data and the expected handling time data, thereby obtaining multiple candidate emergency path data;

[0139] The embodiment of the present invention analyzes the channel safety level data and the expected handling time data through a multi-objective optimization algorithm, aiming to provide multiple alternative emergency paths for vehicles. During the optimization process, the system considers objectives including the shortest travel time, the lowest risk exposure, the best smoothness of travel, etc., and uses traffic simulation software (such as traffic flow models, simulation platforms, etc.) to simulate and test different channels. The simulation analysis will simulate the travel effects of multiple emergency paths based on factors such as traffic flow, road conditions, and risk levels, and provide specific data for each alternative emergency path (such as expected travel time, possible congestion, whether it will be blocked by traffic accidents or other events, etc.).

[0140] Step S56: Perform dynamic risk assessment on multiple alternative emergency path data according to the situation evolution prediction data to obtain the optimal emergency path data; generate navigation instructions for the optimal emergency path data and the collaborative control instruction data, and perform time matching processing to obtain the emergency response strategy data.

[0141] The embodiment of the present invention performs a dynamic risk assessment on multiple alternative emergency paths based on the situation evolution prediction data obtained in step S35. The dynamic risk assessment takes into account the situation that the leakage accident may change at any time, such as changes in wind speed and changes in the gas diffusion range. By real-time monitoring of situation changes, the system will update the alternative paths in real time and select the optimal emergency path with the lowest risk and the shortest evacuation time. Subsequently, the system generates navigation instructions for the vehicle based on the optimal emergency path data and the collaborative control instruction data obtained in step S43. These navigation instructions not only include path planning, but also adjust the navigation route in real time according to the adjustment data of each system to ensure that the vehicle can quickly and safely reach the target emergency shelter point. Finally, after time series matching processing, complete emergency disposal strategy data is generated, including path planning, emergency response instructions, risk avoidance measures and other information.

[0142] By acquiring GPS positioning data of the vehicle's current position, electronic map road network data, and preset emergency shelter location data and safety channel data, the present invention can fully grasp the traffic and emergency resource conditions around the vehicle, providing basic data for subsequent route planning.

[0143] Next, the real-time road condition analysis and processing are carried out according to the GPS positioning data and the electronic map road network data to obtain the current road network traffic status data. This process can understand the road congestion and traffic capacity in real time, and provide a basis for selecting the optimal emergency path. Further, the leakage risk distribution data and the situation evolution prediction data are processed for dangerous area division, and the emergency channel screening and processing are carried out according to the current road network traffic status data, the emergency shelter location data and the safe channel data to obtain the available emergency channel data. This step ensures that the safest and most feasible emergency channel is selected near the dangerous area, and improves the efficiency and safety of emergency response. Then, the safety assessment and disposal time estimation processing are carried out according to the collaborative control instruction data and the available emergency channel data to obtain the channel safety level data and the expected disposal time data. This process provides a scientific basis for the final path selection by quantitatively evaluating the safety and disposal efficiency of each emergency channel. Next, the channel safety level data and the expected disposal time data are processed by multi-objective optimization and traffic simulation analysis to obtain multiple alternative emergency path data. Multi-objective optimization and traffic simulation analysis can comprehensively consider safety and efficiency, generate multiple feasible emergency paths, and improve the flexibility and reliability of path selection. Finally, according to the situation evolution prediction data, a dynamic risk assessment is performed on multiple alternative emergency path data to obtain the optimal emergency path data; navigation instructions are generated for the optimal emergency path data and the collaborative control instruction data, and time-series matching processing is performed to obtain the emergency response strategy data. This step ensures that the selected emergency path has the lowest risk and the highest efficiency under the current circumstances through dynamic evaluation and optimization. At the same time, the generated navigation instructions can guide vehicles and relevant personnel to quickly and accurately perform emergency response measures. In summary, these steps realize the refined management and collaborative control of emergency response to hazardous materials leaks through scientific calculation and analysis methods, which not only improves the safety and reliability of hazardous materials transportation, but also provides strong technical support for emergency management and decision-making.

[0144] Preferably, the present invention further provides a cargo logistics monitoring and management system based on the Internet of Things, which is used to execute the above-mentioned cargo logistics monitoring and management method based on the Internet of Things, and the cargo logistics monitoring and management system based on the Internet of Things includes:

[0145] The dangerous goods state detection module is used to obtain the temperature distribution data and micro-deformation data on the surface of the dangerous goods container; based on the principle of thermodynamic equilibrium, the internal state of the container is non-contact inverted according to the temperature distribution data and micro-deformation data to obtain the dangerous goods state characteristic data, wherein the dangerous goods state characteristic data includes pressure trend data, temperature gradient data and chemical activity data;

[0146] The leakage risk positioning module is used to obtain gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; perform three-dimensional positioning processing on the leakage source position based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtain leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data;

[0147] The risk situation prediction module is used to use the neural network model preset in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data;

[0148] The collaborative control optimization module is used to optimize the distributed control strategy based on the situation evolution prediction data and the response characteristic data of the on-board actuators obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spray system;

[0149] The emergency response planning module is used to dynamically plan the emergency avoidance path according to the collaborative control command data and the preset navigation data to obtain the emergency response strategy data.

[0150] The dangerous goods status detection module in the present invention can accurately obtain the characteristic data of the dangerous goods status, including pressure trend, temperature gradient and chemical activity, by acquiring the temperature distribution data and micro-deformation data of the container surface and performing non-contact inversion processing based on the principle of thermodynamic equilibrium, which provides a solid data basis for subsequent risk assessment and disposal, effectively improves the safety and reliability of dangerous goods transportation, and does not need to directly contact dangerous goods, avoiding possible safety risks. The leakage risk positioning module uses the spectral gas sensors at the four corners of the carriage to obtain gas spectrum array data, and combines the dangerous goods status characteristic data for three-dimensional positioning processing, which can quickly and accurately determine the specific location of the leakage source and the leakage risk distribution, including key information such as leakage source coordinates, diffusion rate and concentration gradient, so as to buy precious time for timely and effective emergency measures, and help to minimize the harm that may be caused by leakage accidents. The risk situation prediction module uses the neural network model in the on-board edge computing unit to predict and analyze the leakage risk distribution data, and obtain the situation evolution prediction data, so that managers can predict the development direction and possible severity of the accident in advance, formulate corresponding response strategies in advance, enhance the active control ability of the dangerous goods logistics process, and improve the scientificity and effectiveness of emergency disposal. The collaborative control optimization module optimizes the distributed control strategy based on the situation evolution prediction data and the response characteristic data of the on-board actuators, obtains the collaborative control command data, and accurately regulates the ventilation, refrigeration and neutralizer spray systems, which can quickly and effectively reduce the concentration of dangerous gases in the car, improve the interior environment, create favorable conditions for the safe evacuation of personnel and emergency disposal of accidents, and also reduce the corrosion and damage of dangerous goods to vehicle equipment. The emergency response planning module dynamically plans the emergency avoidance path based on the collaborative control command data and preset navigation data, and obtains the emergency response strategy data to ensure that when a dangerous goods leakage accident occurs, the vehicle can quickly and safely leave the dangerous area to avoid collisions with other vehicles and personnel and other secondary accidents. At the same time, it is also convenient for rescue forces to arrive at the accident site in time to carry out rescue work, improve the overall efficiency and success rate of emergency disposal, and minimize accident losses.

[0151] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0152] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A cargo logistics monitoring and management method based on the Internet of Things, characterized in that: The following steps are involved: Step S1: Acquire temperature distribution data and micro-deformation data on the surface of the dangerous goods container; Based on the principle of thermodynamic equilibrium, the internal state of the container is processed contactlessly according to the temperature distribution data and micro-deformation data to obtain the state characteristic data of dangerous goods, wherein the state characteristic data of dangerous goods includes pressure trend data, temperature gradient data and chemical activity data; Step S2: acquiring gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; performing three-dimensional positioning processing on the leakage source position based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtaining leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data; Step S3: using a preset neural network model in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data; Step S4: performing distributed control strategy optimization processing according to the situation evolution prediction data and the response characteristic data of the vehicle-mounted actuator obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spraying system; Step S5: Dynamically plan the emergency avoidance path according to the collaborative control instruction data and the preset navigation data to obtain emergency response strategy data.

2. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: acquiring real-time temperature data collected by a distributed temperature sensor array on the surface of the dangerous goods container; Step S12: performing temperature field interpolation and reconstruction processing according to the real-time temperature acquisition data, thereby obtaining temperature distribution data on the container surface; Step S13: Acquire wavelength data collected by the fiber grating strain sensor on the surface of the dangerous goods container; Step S14: demodulating the wavelength acquisition data to obtain optical fiber strain data; Step S15: performing strain field reconstruction calculation processing according to the optical fiber strain data, thereby obtaining micro-deformation data of the container surface; Step S16: Based on the principle of thermodynamic equilibrium, non-contact inversion processing is performed on the internal state of the container according to the temperature distribution data and the micro-deformation data, so as to obtain the characteristic data of the state of dangerous goods, wherein the characteristic data of the state of dangerous goods includes pressure trend data, temperature gradient data and chemical activity data.

3. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 2 is characterized in that: Step S16 includes the following steps: Step S161: performing thermal stress calculation processing on the container wall according to the temperature distribution data, thereby obtaining thermal stress distribution data; Step S162: superimposing and analyzing the micro-deformation data and thermal stress distribution data of the container surface to obtain the comprehensive stress state data of the container wall; Step S163: performing elastic mechanics inversion calculation processing according to the comprehensive stress state data of the container wall, thereby obtaining the internal pressure data of the container; Step S164: performing thermodynamic state equation calculation processing according to the temperature distribution data and the internal pressure data of the container, thereby obtaining the temperature field data of the dangerous goods; Step S165: performing thermodynamic equilibrium analysis on the temperature field data of the dangerous goods, thereby obtaining chemical activity prediction data of the dangerous goods; Step S166: extracting features from the container internal pressure data and the hazardous material chemical activity prediction data, thereby obtaining hazardous material state feature data.

4. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 3 is characterized in that: Step S2 includes the following steps: Step S21: Acquire gas spectrum array data of spectrum gas sensors at four corners of the vehicle compartment; Step S22: performing baseline correction processing on the gas spectrum array data to obtain standardized spectrum data; Step S23: performing multi-component analysis through standardized spectral data and flow field modeling processing to obtain airflow field data in the vehicle compartment; Step S24: Perform three-dimensional positioning processing on the leakage source based on fluid mechanics and diffusion according to the dangerous goods status characteristic data and the airflow field data in the car to obtain leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data.

5. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 4 is characterized in that: Step S23 includes the following steps: Step S231: performing multi-component spectrum analysis processing according to the standardized spectrum data, thereby obtaining gas concentration data of each component; Step S232: performing time series comparison analysis based on the gas concentration data of each component, thereby obtaining concentration change trend data; Step S233: performing spatial interpolation processing according to the concentration change trend data of the four corners of the carriage, thereby obtaining preliminary gas concentration distribution data; Step S234: Obtain the air pressure sensor data and temperature sensor data in the vehicle compartment, and perform flow field modeling processing based on the air pressure sensor data and the temperature sensor data to obtain the air flow field data in the vehicle compartment.

6. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 5 is characterized in that: Step S24 includes the following steps: Step S241: performing theoretical diffusion rate calculation processing on the dangerous goods status characteristic data, thereby obtaining expected diffusion parameter data; Step S242: performing computational fluid dynamics simulation processing on the preliminary gas concentration distribution data and the air flow field data in the vehicle compartment, so as to obtain actual diffusion field data; Step S243: performing gas reverse tracking calculation processing according to the expected diffusion parameter data and the actual diffusion field data, thereby obtaining initial leakage source position data; Step S244: performing Kalman filtering-based optimization processing on the initial leakage source position data, thereby obtaining accurate leakage source coordinate data; Step S245: performing diffusion dynamics analysis and processing according to the precise leakage source coordinate data and the actual diffusion field data, thereby obtaining diffusion rate data; Step S246: performing gradient calculation processing on the diffusion rate data and the concentration data of each component gas to obtain concentration gradient data; performing risk assessment processing based on the leakage source coordinate data, diffusion rate data and concentration gradient data to obtain leakage risk distribution data.

7. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 6, characterized in that: Step S3 includes the following steps: Step S31: extracting features and normalizing the leakage risk distribution data to obtain neural network input feature data; Step S32: obtaining a neural network model preset in the vehicle-mounted edge computing unit, wherein the parameters of the neural network model include network structure data, weight coefficient data, and bias parameter data; Step S33: Acquire historical leakage accident case data; and perform classification matching processing on the historical leakage accident case data based on the type of dangerous goods, thereby obtaining leakage accident classification data; Step S34: performing online calibration processing on the parameters in the neural network model according to the leakage accident classification data, thereby obtaining a leakage risk prediction model; Step S35: Use the leakage risk prediction model to predict the leakage risk of the neural network input feature data, so as to obtain situation evolution prediction data.

8. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 7, characterized in that: Step S4 includes the following steps: Step S41: Acquire response characteristic data of the ventilation system, the refrigeration system and the neutralizer spraying system; Step S42: Perform time series response analysis based on the situation evolution prediction data and the response characteristic data of each execution system, and perform optimization processing based on response time, execution sequence and response intensity to obtain the optimal control time series data. Step S43: performing control parameter calculation processing on the ventilation system, the refrigeration system and the neutralizer spraying system respectively according to the optimal control time series data and the situation evolution prediction data, so as to obtain the adjustment instruction data of each system; Step S44: performing system interference analysis and coordination optimization processing on the adjustment instruction data of each system to obtain preliminary coordinated control instruction data; Step S45: performing control quantity constraint check processing on the collaborative control instruction data, thereby obtaining control instruction data that satisfies the physical constraints of the actuator; Step S46: performing control sequence generation processing according to the control instruction data, thereby obtaining control instruction sequence data for time-sharing execution; Step S47: Reliability evaluation and priority sorting are performed on the control instruction sequence data executed in time-sharing manner, thereby obtaining collaborative control instruction data.

9. The method for monitoring and managing cargo logistics based on the Internet of Things according to claim 8, characterized in that: Step S5 includes the following steps: Step S51: obtaining GPS positioning data of the vehicle's current location, electronic map road network data, and preset emergency shelter location data and safety channel data; Step S52: Perform real-time road condition analysis and processing based on GPS positioning data and electronic map road network data, thereby obtaining current road network traffic status data; Step S53: performing dangerous area division processing on the leakage risk distribution data and the situation evolution prediction data, and performing emergency channel screening processing according to the current road network traffic status data, emergency shelter location data and safety channel data, so as to obtain available emergency channel data; Step S54: Perform safety assessment and handling time estimation processing according to the collaborative control instruction data and the available emergency channel data, thereby obtaining channel safety level data and expected handling time data; Step S55: performing multi-objective optimization and traffic simulation analysis on the channel safety level data and the expected handling time data, thereby obtaining multiple candidate emergency path data; Step S56: Perform dynamic risk assessment on multiple alternative emergency path data according to the situation evolution prediction data to obtain the optimal emergency path data; generate navigation instructions for the optimal emergency path data and the collaborative control instruction data, and perform time matching processing to obtain the emergency response strategy data.

10. A cargo logistics monitoring and management system based on the Internet of Things, characterized in that: Used to execute the cargo logistics monitoring and management method based on the Internet of Things as claimed in claim 1, the cargo logistics monitoring and management system based on the Internet of Things includes: The dangerous goods state detection module is used to obtain the temperature distribution data and micro-deformation data on the surface of the dangerous goods container; based on the principle of thermodynamic equilibrium, the internal state of the container is non-contact inverted according to the temperature distribution data and micro-deformation data to obtain the dangerous goods state characteristic data, wherein the dangerous goods state characteristic data includes pressure trend data, temperature gradient data and chemical activity data; The leakage risk positioning module is used to obtain gas spectrum array data through the spectrum gas sensors at the four corners of the carriage; perform three-dimensional positioning processing on the leakage source based on fluid mechanics diffusion according to the dangerous goods state characteristic data and the gas spectrum array data, and obtain leakage risk distribution data, wherein the leakage risk distribution data includes leakage source coordinate data, diffusion rate data and concentration gradient data; The risk situation prediction module is used to use the neural network model preset in the vehicle-mounted edge computing unit to perform leakage risk prediction analysis on the leakage risk distribution data to obtain situation evolution prediction data; The collaborative control optimization module is used to optimize the distributed control strategy based on the situation evolution prediction data and the response characteristic data of the on-board actuators obtained in real time to obtain collaborative control instruction data, wherein the collaborative control instruction data specifically regulates the ventilation system, the refrigeration system and the neutralizer spray system; The emergency response planning module is used to dynamically plan the emergency avoidance path according to the collaborative control command data and the preset navigation data to obtain the emergency response strategy data.

Images