Supply chain management system based on Internet of Things technology

Abstract

The invention provides a supply chain management system based on the Internet of Things technology, and relates to the technical field of supply chain management, and the system comprises the steps: predicting the temperature and humidity conditions in the future according to the temperature and humidity change trends of cargoes in the transportation process; according to the future temperature and humidity conditions and the cargo position data, abnormal points are identified through preset dynamic factors; according to the abnormal point, the real-time road condition, the weather information and the traffic jam condition, a particle optimization algorithm is utilized to obtain a corresponding transportation path; according to the transportation path and the current road condition and weather condition, evaluating the potential risk of each path to obtain a risk evaluation result; and determining a corresponding risk coping strategy according to the risk assessment result. According to the invention, comprehensive monitoring, accurate prediction and intelligent decision making of the cargo transportation process are realized, so that the efficiency and safety of supply chain transportation management are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain management, and particularly to a supply chain management system based on Internet of Things technology. Background Art

[0002] Intelligent supply chain transportation management systems play an increasingly important role in the modern logistics industry. Some traditional supply chain transportation management methods rely on manual operations and simple informatization means, and there are many deficiencies in aspects such as cargo transportation tracking, monitoring, and risk management.

[0003] First of all, some traditional solutions lack an effective data analysis mechanism and are unable to deeply mine and analyze the collected data to reveal key information such as the temperature and humidity change trends of goods during transportation. The lack of such information makes the control of the cargo transportation environment less precise and difficult to ensure the quality and safety of goods.

[0004] Furthermore, some traditional solutions have weak capabilities in predicting and identifying potential risks. Due to the lack of comprehensive analysis of historical data and real-time data, some traditional solutions are difficult to accurately predict future temperature and humidity conditions and may have problems in failing to identify emerging abnormal points. Therefore, this increases the uncertainty and risks during the cargo transportation process. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a supply chain management system based on Internet of Things technology, which realizes comprehensive monitoring, accurate prediction, and intelligent decision-making of the cargo transportation process, thereby effectively improving the efficiency and safety of supply chain transportation management.

[0006] To solve the above technical problem, the technical solution of the present invention is as follows:

[0007] In a first aspect, a supply chain management system based on Internet of Things technology includes:

[0008] An acquisition module, configured to collect multi-dimensional data during the cargo transportation process in real time, where the multi-dimensional data includes cargo location, temperature, and humidity;

[0009] An analysis module, configured to analyze the multi-dimensional data, and obtain the temperature and humidity change trends of the goods during transportation by comparing the temperature and humidity data at different time points;

[0010] A prediction module, configured to predict future temperature and humidity conditions based on the temperature and humidity change trends of the goods during transportation;

[0011] An identification module, configured to identify abnormal points according to the future temperature and humidity conditions and based on the cargo location data through a preset dynamic factor;

[0012] An optimization module, which is used to obtain corresponding transportation routes by using a particle optimization algorithm according to abnormal points, real-time road conditions, weather information, and traffic congestion situations;

[0013] An evaluation module, which is used to evaluate the potential risks of each route according to the transportation route and the current road conditions and weather conditions to obtain a risk assessment result; and determine corresponding risk response strategies according to the risk assessment result.

[0014] Furthermore, analyze multi-dimensional data, and obtain the temperature and humidity change trends of goods during transportation by comparing the temperature and humidity data at different time points, including:

[0015] Collect multi-dimensional data in real time during the transportation of goods through temperature and humidity sensors, including timestamps, temperature, and humidity values; perform standardization processing on the multi-dimensional data to obtain standardized multi-dimensional data;

[0016] Sort the standardized multi-dimensional data according to timestamps to construct time series of temperature and humidity;

[0017] Identify the initial change trends of temperature and humidity according to the time series of temperature and humidity;

[0018] Generate an initial temperature and humidity change trend model based on the initial change trends;

[0019] Generate neighborhood solutions of the current solution through a random perturbation strategy; define an objective function for evaluating the fitting effect; determine whether to accept the neighborhood solution as the new current solution according to the Metropolis criterion of the simulated annealing algorithm, and update the temperature parameter in the simulated annealing algorithm;

[0020] Repeat the steps of neighborhood solution generation, objective function evaluation, and acceptance criterion judgment until the termination condition is met to obtain the final temperature and humidity change trends.

[0021] Furthermore, predict future temperature and humidity conditions according to the temperature and humidity change trends of goods during transportation, including:

[0022] Construct an autoregressive integrated moving average model according to the temperature and humidity change trends of goods during transportation;

[0023] Use the collected temperature and humidity data to train the autoregressive integrated moving average model to obtain a trained autoregressive integrated moving average model;

[0024] Predict future temperature and humidity conditions according to the trained autoregressive integrated moving average model.

[0025] Further, use the collected temperature and humidity data to train an autoregressive integrated moving average (ARIMA) model to obtain a trained ARIMA model, including:

[0026] Encode the three parameters p, d, and q of the ARIMA model as individuals in a genetic algorithm, where p represents the order of the autoregressive term, d represents the number of differencing times, and q represents the order of the moving average term;

[0027] Generate an initial population that contains multiple individuals, and each individual represents a parameter combination of an ARIMA model;

[0028] Define a fitness function for evaluating the quality of each individual; select corresponding individuals for reproduction according to the fitness function, and through crossover operations, exchange the parameters of two individuals to generate new offspring individuals; randomly vary the individuals, and continuously adjust the three parameters p, d, and q of the ARIMA model through the genetic algorithm until a preset number of iterations is reached to obtain optimized parameters p, d, and q;

[0029] Calculate the error of each ARIMA model through the training set data and the test set data;

[0030] According to the error of the ARIMA model, select the final parameters p, d, and q based on the error, and train the ARIMA model with the final parameters p, d, and q to obtain a trained ARIMA model.

[0031] Further, according to the future temperature and humidity conditions and the cargo location data, identify abnormal points through a preset dynamic factor, including:

[0032] Collect the cargo location data in real time, and calculate the dynamic factor based on the cargo location data, transportation route, historical data, and weather forecast information;

[0033] Associate the real-time temperature and humidity data with the cargo location data to form a comprehensive data set that includes timestamps, temperature and humidity values, and location information;

[0034] Divide the comprehensive data set into different segments according to the cargo location data, and each segment represents a time period during the cargo transportation process;

[0035] For the data of each time period, calculate the absolute difference between the actual temperature and humidity value and the future temperature and humidity;

[0036] Multiply the absolute difference by the dynamic factor to obtain a weighted difference value;

[0037] Set a threshold value. If the weighted difference value exceeds the threshold, mark the corresponding point as an abnormal point.

[0038] Furthermore, the calculation formula for the dynamic factor is as follows:

[0039]

[0040] where α, β, γ, and δ represent weight coefficients; D m represents the total distance, which is the maximum distance from the starting point to the ending point; D c represents the current distance, which is the distance from the current position to the starting point; C v represents the climate variation index; ΔA represents the altitude change, which is the altitude change amount on the route; A m represents the maximum altitude, which is the altitude of the highest point on the route; T c represents the terrain complexity; H a represents the historical anomaly index; T f and H f represent the future temperature and humidity values respectively; T c and H c represent the current actual temperature and humidity values respectively; T r and H r represent the normal change ranges of the regional temperature and humidity respectively;

[0041] where where n is the number of climate data points along the route; x i is the temperature or humidity value of the i-th data point; μ is the average value of the temperature or humidity along the route; σ is the standard deviation of the temperature or humidity along the route; i is the index value of the data point;

[0042] where where L is the total length of the route; h(x) is the altitude at the route position x; is the slope of the altitude at the position x;

[0043] where where m is the number of historical data points under similar conditions; y j is the temperature and humidity value at the j-th historical data point; is the average value of the historical temperature and humidity data under similar conditions, and j is the index value of the historical data point.

[0044] Furthermore, according to the anomaly points, real-time road conditions, weather information, and traffic congestion situations, use the particle optimization algorithm to obtain the corresponding transportation routes, including:

[0045] Obtain real-time road conditions, including road surface conditions and visibility; obtain real-time weather data, including temperature, humidity, wind speed, rainfall, and snow conditions; obtain real-time traffic congestion information;

[0046] Preprocess real-time road conditions, real-time weather data, and traffic congestion information to obtain preprocessed real-time road conditions, real-time weather data, and traffic congestion information;

[0047] Represent the transportation route as a series of nodes, where each node contains location information, as well as preprocessed real-time road conditions, real-time weather data, and traffic congestion information;

[0048] Based on each node, generate a set of initial paths as particles, where each particle represents a solution;

[0049] Define an objective function for evaluating the quality of a path according to the real-time road conditions, weather data, and traffic congestion information in the nodes;

[0050] Use the objective function to evaluate each particle and calculate its fitness value; according to the fitness value of the particle and a preset update strategy, combined with the changes in real-time data, dynamically adjust the state of the particle to search for the corresponding path, and repeat the particle evaluation and state update process until a preset number of iterations is reached, so as to select the corresponding path from the final particle swarm as the transportation route.

[0051] Furthermore, according to the transportation route and the current road conditions and weather conditions, evaluate the potential risks of each path to obtain a risk assessment result, including:

[0052] Obtain the basic information of each transportation route, including the path length, the areas passed through, and the key nodes; combined with GIS data, analyze the terrain and landforms, road grades, and historical accident data of the path;

[0053] Obtain real-time road condition data, including the degree of road surface slipperiness, visibility, and traffic flow; obtain real-time weather forecast and actual data, including key meteorological information for the area where the transportation route is located, and the key meteorological information includes temperature, humidity, wind speed, and precipitation probability;

[0054] Determine the risk assessment factors affecting transportation safety, including road surface quality, road width, bad weather index, visibility, congestion degree, and accident frequency;

[0055] Set corresponding weights for each risk assessment factor to establish a quantitative risk calculation model, and the risk calculation model outputs a specific risk value according to each risk assessment factor;

[0056] Input the real-time collected road and weather data into the risk calculation model to perform real-time risk assessment on each transportation route to obtain a specific risk value.

[0057] In a second aspect, a supply chain management method based on Internet of Things technology includes:

[0058] Real-time collect multi-dimensional data during the goods transportation process, where the multi-dimensional data includes the location, temperature, and humidity of the goods;

[0059] Analyze the multi-dimensional data, and by comparing the temperature and humidity data at different time points, obtain the temperature and humidity change trends of the goods during transportation;

[0060] Predict the future temperature and humidity conditions based on the temperature and humidity change trends of the goods during transportation;

[0061] Based on the future temperature and humidity conditions and the goods location data, identify abnormal points through a preset dynamic factor;

[0062] Based on the abnormal points, real-time road conditions, weather information, and traffic congestion situations, use the particle optimization algorithm to obtain the corresponding transportation route;

[0063] Evaluate the potential risks of each route according to the transportation route and the current road conditions and weather conditions to obtain the risk assessment result;

[0064] Determine the corresponding risk response strategy according to the risk assessment result.

[0065] In a third aspect, a computing device includes:

[0066] One or more processors;

[0067] A storage device for storing one or more programs, which when executed by the one or more processors, cause the one or more processors to implement the method.

[0068] In a fourth aspect, a computer-readable storage medium stores a program that, when executed by a processor, implements the method.

[0069] The above solution of the present invention has at least the following beneficial effects.

[0070] By real-time collecting multi-dimensional data (such as the location, temperature, and humidity of the goods) during the goods transportation process, the transportation status of the goods becomes transparent and traceable, providing a comprehensive view of the goods status for supply chain managers, which helps to timely discover problems and make responses.

[0071] By analyzing and predicting the temperature and humidity change trends of goods during transportation, the transportation environment can be adjusted in a timely manner to ensure that the goods are transported under optimal conditions and reduce the risk of goods damage caused by temperature and humidity changes. By identifying abnormal points through preset dynamic factors, the system can promptly detect abnormal situations during transportation, such as sudden temperature and humidity changes, transportation delays, etc., and thus take corresponding countermeasures to ensure the safety and timeliness of goods transportation. Using the particle optimization algorithm combined with real-time road conditions, weather information, and traffic congestion conditions to provide optimal route planning for goods transportation helps reduce transportation time and costs and improve transportation efficiency.

[0072] Risk assessment and response strategies: By evaluating the potential risks of each path and formulating corresponding risk response strategies, it is possible to respond promptly in case of emergencies, reduce potential losses, and improve the resilience and reliability of the supply chain. This method provides rich data and insights for supply chain managers, helping them make more informed decisions, thereby enhancing the efficiency and effectiveness of the entire supply chain management. Description of the Drawings

[0073] Figure 1 is a flowchart showing a supply chain management method based on Internet of Things technology provided by an embodiment of the present invention.

[0074] Figure 2 is a schematic diagram of a supply chain management system based on Internet of Things technology provided by an embodiment of the present invention. Detailed Embodiments

[0075] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0076] As Figure 1 shown, an embodiment of the present invention proposes a supply chain management method based on Internet of Things technology, and the method includes the following steps:

[0077] Step 1, collect multi-dimensional data during the transportation of goods in real time, and the multi-dimensional data includes the location, temperature, and humidity of the goods;

[0078] Step 2, analyze the multi-dimensional data, and obtain the temperature and humidity change trends of the goods during transportation by comparing the temperature and humidity data at different time points;

[0079] Step 3, predict the future temperature and humidity conditions according to the temperature and humidity change trends of the goods during transportation;

[0080] Step 4: Based on the future temperature and humidity conditions and the cargo location data, identify abnormal points through a preset dynamic factor.

[0081] Step 5: Based on the abnormal points, real-time road conditions, weather information, and traffic congestion, use the particle swarm optimization algorithm to obtain the corresponding transportation route.

[0082] Step 6: Based on the transportation route and the current road conditions and weather conditions, evaluate the potential risks of each route to obtain a risk assessment result.

[0083] Step 7: Based on the risk assessment result, determine the corresponding risk response strategy.

[0084] In the embodiment of the present invention, multi-dimensional data (such as cargo location, temperature, humidity) during the cargo transportation process is collected in real time through the Internet of Things technology, making the transportation status of the cargo transparent and traceable, providing a comprehensive view of the cargo status for supply chain managers, and helping to detect problems in a timely manner and make responses. By analyzing and predicting the temperature and humidity change trends of the cargo during transportation, the transportation environment can be adjusted in a timely manner to ensure that the cargo is transported under optimal conditions and reduce the risk of cargo damage caused by temperature and humidity changes. By identifying abnormal points through a preset dynamic factor, the system can timely detect abnormal situations during transportation, such as sudden changes in temperature and humidity, transportation delays, etc., and thus take corresponding countermeasures to ensure the safety and timeliness of cargo transportation. Using the particle swarm optimization algorithm combined with real-time road conditions, weather information, and traffic congestion to provide an optimal route planning for cargo transportation helps to reduce transportation time and costs and improve transportation efficiency. Risk assessment and response strategy: By evaluating the potential risks of each route and formulating corresponding risk response strategies, sudden situations can be quickly responded to, potential losses can be reduced, and the resilience and reliability of the supply chain can be improved. This method provides rich data and insights for supply chain managers, helping them make more informed decisions, thereby enhancing the efficiency and effectiveness of the entire supply chain management.

[0085] In another preferred embodiment of the present invention, in the above Step 1, collecting multi-dimensional data in real time during the cargo transportation process, where the multi-dimensional data includes cargo location, temperature, humidity, may include:

[0086] Select data collection devices for the goods transportation environment, such as GPS locators, temperature and humidity sensors, etc., to ensure that these devices have real-time data transmission functions, so as to send data to the central management system in a timely manner through Internet of Things technology. Install the selected data collection devices on the goods or transportation vehicles (such as trucks, containers, etc.); configure the device parameters to ensure that it can accurately measure and record multi-dimensional data such as the location, temperature, and humidity of the goods. Build a stable and reliable data transmission network to support real-time data exchange between the device and the system. Wireless communication technologies (such as 4G / 5G, LoRa, Wi-Fi, etc.) can be used to ensure that data can be transmitted continuously during transportation; start the data collection device to start real-time collection of data such as the location, temperature, and humidity of the goods, and the device will send the collected data to the central management system or cloud platform through the established data transmission network. The central management system or cloud platform receives the real-time data from the data collection device, cleans and organizes the received data, and stores it according to identifiers such as timestamps. Regularly verify and calibrate the collected data to ensure the accuracy and reliability of the data. Data quality assessment can be carried out by comparing with other data sources, device self-checks, or regular manual inspections. During the data collection process, if abnormal data (such as values exceeding the preset range) is found, it should be immediately recorded and the alarm mechanism should be triggered to further analyze the abnormal data to determine whether adjustments need to be made to the transportation environment or equipment.

[0087] In a preferred embodiment of the present invention, in step 2 above, analyzing the multi-dimensional data by comparing the temperature and humidity data at different time points to obtain the temperature and humidity change trends of the goods during transportation may include:

[0088] Step 21, collect multi-dimensional data in real time during the goods transportation process through temperature and humidity sensors, including timestamps, temperature, and humidity values; perform standardization processing on the multi-dimensional data to obtain standardized multi-dimensional data, specifically including: during the goods transportation process, collect data in real time through temperature and humidity sensors; ensure that the sensors record timestamps, temperature values, and humidity values at a certain frequency (such as per second or per minute); transmit the collected data to the central data processing system wirelessly or wiredly; clean the collected raw data to remove outliers, noise, or duplicate data; perform standardization processing on the temperature and humidity data, such as using z-score standardization to convert the data into a distribution with a mean of 0 and a standard deviation of 1.

[0089] Step 22, sort the standardized multi-dimensional data according to the timestamps to construct time series of temperature and humidity, specifically including: sort the standardized multi-dimensional data in the order of timestamps, with time as the horizontal axis and temperature and humidity values as the vertical axis, to construct two time series: one for temperature and one for humidity.

[0090] Step 23: Identify the initial change trends of temperature and humidity based on their time series, specifically including: making a preliminary judgment on the constructed temperature time series and humidity time series to determine whether the temperature and humidity show an upward trend, a downward trend, or a fluctuating stability; using the time series data to calculate the slope between adjacent data points to quantify the increase and decrease speeds of temperature and humidity, where a positive slope indicates an increasing trend, a negative slope indicates a decreasing trend, and a value close to zero indicates stability; identifying the peaks (highest points) and valleys (lowest points) in the time series, marking their timestamps and values, and finding the turning points, i.e., the points where the change trend changes significantly.

[0091] Step 24: Generate an initial temperature and humidity change trend model based on the initial change trends, specifically including: selecting a suitable time series analysis model, such as the autoregressive integrated moving average model (ARIMA), based on the initially identified change trends, such as stationary, seasonal fluctuations, etc.; using the existing time series data to train the selected model and adjust the model parameters, such as the order, seasonal period, etc., to fit the data; calculating the goodness of fit of the model, such as the coefficient of determination (R 2 value), to evaluate the degree of interpretation of the model for the data, where an R 2 value close to 1 indicates a good model fitting effect, and a value close to 0 indicates a poor fitting effect.

[0092] Step 25: Generate a neighborhood solution of the current solution through a random perturbation strategy; define an objective function for evaluating the fitting effect; and determine whether to accept the neighborhood solution as the new current solution according to the Metropolis criterion of the simulated annealing algorithm and update the temperature parameter in the simulated annealing algorithm, specifically including:

[0093] Applying small random perturbations to the parameters of the current model to generate a new parameter combination as the neighborhood solution; defining the mean squared error (MSE) as the objective function for evaluating the model fitting effect, where MSE calculates the average squared difference between the predicted value and the actual observed value; calculating the MSE difference between the neighborhood solution and the current solution, if the difference is negative (i.e., the neighborhood solution is better), then unconditionally accept the neighborhood solution, if the difference is positive, then determine whether to accept the neighborhood solution with a certain acceptance probability (based on the temperature parameter and the size of the difference); gradually reducing the value of the temperature parameter according to the rules of the simulated annealing algorithm, and the reduction of the temperature parameter will reduce the probability of accepting a worse solution and promote the convergence of the algorithm.

[0094] Step 26: Repeat the steps of generating neighborhood solutions, evaluating the objective function, and judging the acceptance criterion until the termination condition is met to obtain the final temperature and humidity change trends, specifically including:

[0095] Repeat step 25, continuously generate neighborhood solutions, evaluate the objective function and apply the acceptance criterion, set the termination conditions, such as reaching the preset maximum number of iterations, the MSE value shows no significant improvement after consecutive iterations (i.e., converges below a certain threshold), or the temperature parameter drops to a very low value, and check whether the termination conditions are met after each iteration; when the termination conditions are met, stop the iterative process and use the current solution (i.e., the solution accepted in the last round of iteration) as the final temperature and humidity change trend model.

[0096] In the embodiment of the present invention, in step 21, the multi-dimensional data is standardized, which can eliminate the dimensional differences between the data, making the data of different dimensions comparable in subsequent analysis. This helps to accurately capture the true change trends of temperature and humidity and improve the accuracy of the analysis. In step 22, by sorting according to the time stamp to construct the time series of temperature and humidity, the continuous changes of temperature and humidity during the transportation of goods can be clearly shown. This helps to discover potential periodic patterns or abnormal fluctuations and provides strong support for subsequent prediction and decision-making. In step 23, identifying the initial change trends of temperature and humidity can provide a basis for subsequent model construction. By initially understanding the change characteristics of the data, it is possible to better select the appropriate model type and parameter settings, thereby improving the fitting effect and prediction accuracy of the model. From step 24 to step 26, by generating an initial model, using a random perturbation strategy to generate neighborhood solutions, evaluating the fitting effect, and updating the solution according to the simulated annealing algorithm, the change trend model of temperature and humidity can be continuously optimized. This iterative optimization process not only improves the fitting accuracy of the model but also enhances the robustness of the model, enabling it to better adapt to the complex and changeable environment during the actual transportation process. Accurately grasping the change trends of temperature and humidity during the transportation of goods is crucial for formulating effective risk response strategies and optimizing transportation conditions. The analysis results obtained through the above steps can provide more accurate and comprehensive decision-making support for supply chain managers, thereby ensuring the safety and quality of goods transportation.

[0097] In a preferred embodiment of the present invention, step 3 above, predicting the future temperature and humidity conditions according to the change trends of temperature and humidity during the transportation of goods, includes:

[0098] Construct an autoregressive integrated moving average model according to the temperature and humidity change trends of goods during transportation, specifically including: collecting historical temperature and humidity data of goods during transportation to ensure the quality and integrity of the data; cleaning the data to remove outliers, missing values or duplicate values, and performing necessary stationarity tests. If the data is non-stationary, differencing or other transformations are required to make it stationary; select an appropriate ARIMA model form according to the characteristics of the data (such as seasonality, periodicity, etc.). The ARIMA model is a combination of an autoregressive model (AR), a moving average model (MA), and differencing (I); set initial ranges or values for the three key parameters p (order of autoregressive terms), d (number of differencing), and q (order of moving average terms) of the ARIMA model.

[0099] Use the collected temperature and humidity data to train the autoregressive integrated moving average model to obtain the trained autoregressive integrated moving average model, including: encoding the three parameters p, d, and q of the autoregressive integrated moving average model as individuals in the genetic algorithm, where p represents the order of autoregressive terms; d represents the number of differencing; q represents the order of moving average terms; generate an initial population, which contains multiple individuals, and each individual represents a parameter combination of an autoregressive integrated moving average model; define a fitness function for evaluating the quality of each individual; select corresponding individuals for reproduction according to the fitness function, and through the crossover operation, exchange the parameters of two individuals to generate new offspring individuals; randomly vary the individuals, and continuously adjust the three parameters p, d, and q of the autoregressive integrated moving average model through the genetic algorithm until the preset number of iterations is reached to obtain optimized parameters p, d, and q; calculate the error of each autoregressive integrated moving average model through the training set data and the test set data; select the final parameters p, d, and q according to the error of the autoregressive integrated moving average model, and train the ARIMA model with the final parameters p, d, and q to obtain the trained autoregressive integrated moving average model;

[0100] Among them, the specific implementation process of optimizing the ARIMA model parameters using the genetic algorithm is as follows:

[0101] Encode the parameters p, d, and q of the ARIMA model as individuals in a genetic algorithm, where each individual represents a possible combination of parameters; randomly generate an initial population consisting of multiple individuals, with each individual being a possible combination of p, d, and q; define a fitness function to evaluate the quality of each individual, and this function can be the mean squared error (MSE). The smaller the error, the higher the fitness. Based on the fitness function, select excellent individuals from the current population for reproduction. This usually involves methods such as roulette wheel selection and tournament selection. Through the crossover operation, partially exchange the parameters of two selected individuals to generate new offspring individuals, which helps to explore new possible solutions in the search space. Randomly mutate the offspring individuals, that is, randomly change the values of some of their parameters. The mutation operation helps to increase the diversity of the population and prevent premature convergence. Repeat the processes of selection, crossover, and mutation until the preset number of iterations is reached. Through the iterative process of the genetic algorithm, continuously optimize the parameters p, d, and q of the ARIMA model, and finally obtain a set of optimized parameter combinations;

[0102] Use the optimized parameters p, d, and q, along with historical temperature and humidity data, to train the ARIMA model, and estimate the coefficients of the model through the maximum likelihood estimation method; divide the historical data into a training set and a test set. The training set is used for model training, and the test set is used to evaluate the prediction performance of the model; use the test set data to calculate the prediction error of the ARIMA model, and the error metrics include the mean squared error (MSE); evaluate the prediction performance of the ARIMA model based on the calculated error metrics.

[0103] According to the trained autoregressive integrated moving average model, predict the future temperature and humidity conditions, specifically including: determining the future time point or time period to be predicted; using the trained ARIMA model and the optimized parameters p, d, and q to predict the temperature and humidity conditions at the future time point or time period, and output the prediction results, including the predicted temperature and humidity values at each future time point; analyze the prediction results and take corresponding measures according to the actual situation, such as adjusting transportation conditions, issuing alarms, etc.

[0104] In an embodiment of the present invention, the parameters (p, d, q) of the ARIMA model are optimized through a genetic algorithm, which can more accurately fit historical data and predict future temperature and humidity conditions accordingly. This optimization method can reduce prediction errors and make the prediction results more accurate and reliable. The genetic algorithm has global search capabilities and can find the optimal solution within a large parameter space, thus adapting to different temperature and humidity change trends. This means that regardless of how the temperature and humidity of the goods change, this method can find the most suitable ARIMA model parameters for prediction. The selection of parameters for the traditional ARIMA model may require a large amount of manual trial and error and adjustment, while the automated optimization process combined with the genetic algorithm can significantly reduce such manual intervention and improve work efficiency. During the transportation of goods, the temperature and humidity conditions may change at any time, and this method can adjust the parameters of the ARIMA model in real time to adapt to these changes, ensuring the real-time and accuracy of the prediction.

[0105] In a preferred embodiment of the present invention, step 4, according to the future temperature and humidity conditions and the goods location data, identifying outliers through a preset dynamic factor may include:

[0106] Step 41, collect the goods location data in real time, and calculate the dynamic factor according to the goods location data, transportation route, historical data, and weather forecast information;

[0107] Step 42, associate the real-time temperature and humidity data with the goods location data to form a comprehensive data set including a timestamp, temperature and humidity values, and location information, specifically including: determining that the real-time collected temperature and humidity data and location data are synchronized in time; associating the temperature and humidity data and location data at the same time point to form a new comprehensive data set, and each record includes a timestamp, real-time temperature and humidity readings, and the specific location of the goods.

[0108] Step 43, divide the comprehensive data set into different segments according to the goods location data, and each segment represents a time period during the goods transportation process, specifically including: dividing the entire transportation process into several time periods according to the transportation route of the goods and the estimated transportation time; dividing the comprehensive data set into corresponding parts according to these time periods, and each part represents a specific transportation time period.

[0109] Step 44, for the data of each time period, calculate the absolute difference between the actual temperature and humidity value and the future temperature and humidity, specifically including: using the previously trained ARIMA model to predict the temperature and humidity value at the end of each time period; for the data points within each time period, calculate the absolute difference between the actual temperature and humidity readings and the predicted temperature and humidity values.

[0110] Step 45: Multiply the absolute difference by the dynamic factor to obtain a weighted difference value, specifically including: multiplying the absolute difference calculated in Step 44 by the dynamic factor obtained in Step 41 to obtain the weighted difference value, which takes into account the specific impacts of the current environmental conditions and the transportation route.

[0111] Step 46: Set a threshold. If the weighted difference value exceeds the threshold, mark the corresponding point as an abnormal point, specifically including: setting a threshold based on historical data and business requirements to determine whether the weighted difference value is abnormal; if the weighted difference value of a certain data point exceeds this threshold, then this point is regarded as an abnormal point.

[0112] In the embodiment of the present invention, by combining real-time temperature and humidity data and cargo location data, as well as the dynamic factor calculated considering the transportation route, historical data, and weather forecast information, abnormal points can be identified more accurately. This method not only considers the absolute values of temperature and humidity but also the dynamic changes of the cargo during transportation, thereby improving the accuracy of anomaly detection. This method can monitor the temperature and humidity conditions of the cargo in real time and issue an early warning in a timely manner when an anomaly is detected. This is particularly important for goods that require strict control of transportation environmental conditions, as potential problems can be discovered and addressed in a timely manner to ensure the safety and quality of the goods. The introduction of the dynamic factor enables this method to flexibly adapt to different transportation scenarios and conditions. Whether it is changes in the route, the impact of weather, or interference from other unknown factors, the dynamic factor can be adjusted according to the actual situation, thereby maintaining the sensitivity and accuracy of anomaly detection. By forming a comprehensive dataset containing timestamps, temperature and humidity values, and location information, this method provides strong support for data-driven decision-making. Enterprises can analyze and optimize based on these data, improve transportation strategies, and enhance logistics efficiency and service quality.

[0113] In a preferred embodiment of the present invention, the calculation formula of the dynamic factor is:

[0114]

[0115] where α, β, γ, and δ represent weight coefficients; D m represents the total distance, which is the maximum distance from the starting point to the ending point; D c represents the current distance, which is the distance from the current location to the starting point; C v represents the climate variation index; ΔA represents the altitude change, which is the altitude change amount on the route; A m represents the maximum altitude, which is the highest altitude on the route; T c represents the terrain complexity; H a represents the historical anomaly index; T f and H frespectively represent future temperature and humidity values; T c and H c respectively represent the current actual temperature and humidity values; T r and H r respectively represent the normal change ranges of regional temperature and humidity;

[0116] Among them, where n is the number of climate data points along the way; x i is the temperature or humidity value of the i-th data point; μ is the average value of temperature or humidity along the way; σ is the standard deviation of temperature or humidity along the way; i is the index value of the data point;

[0117] Among them, where L is the total length of the route; h(x) is the altitude at the route position x; is the slope of the altitude at position x;

[0118] Among them, where m is the number of historical data points under similar conditions; y j is the temperature and humidity value at the j-th historical data point; is the average value of historical temperature and humidity data under similar conditions, and j is the index value of the historical data point.

[0119] In the embodiment of the present invention, this formula comprehensively considers multiple factors such as distance, climate, terrain, altitude, historical anomalies, and future and current temperature and humidity, making the calculation of the dynamic factor more comprehensive and accurate. This consideration of multiple factors helps to more accurately reflect the actual changes in temperature and humidity during the cargo transportation process. By adjusting the weight coefficients, this formula can flexibly adapt to different transportation scenarios and requirements. For example, in areas with variable climate or complex terrain, the relevant weights can be adjusted accordingly to improve the sensitivity and accuracy of anomaly detection. The future temperature and humidity prediction values are incorporated into the formula and compared with the current actual temperature and humidity values, achieving an effective combination of prediction and real-time monitoring. This combination helps to timely detect potential temperature and humidity anomalies. By considering the normal change ranges of regional temperature and humidity, this formula can reduce false alarms caused by normal fluctuations to a certain extent. This helps to improve the reliability of anomaly detection and reduce unnecessary interference and costs. The historical anomaly index in the formula reflects the frequency or degree of anomalies that occurred in the past under similar conditions, which helps to combine historical experience with real-time data and improve the accuracy of anomaly detection.

[0120] In a preferred embodiment of the present invention, step 5 above, according to the anomaly points, real-time road conditions, weather information, and traffic congestion conditions, using the particle optimization algorithm to obtain the corresponding transportation route, may include:

[0121] Step 51, obtain real-time road conditions, including road surface conditions and visibility; obtain real-time weather data, including temperature, humidity, wind speed, rainfall, and snow conditions; obtain real-time traffic congestion information, specifically including: collecting real-time road surface conditions (such as dry, slippery, waterlogged, etc.) and visibility information through a traffic management system, road surface sensors, or in-vehicle devices; obtaining real-time weather data, including temperature, humidity, wind speed, rainfall, and snow conditions, from a meteorological service station or an online weather API; using traffic monitoring cameras, GPS data, or a third-party traffic information service platform to obtain real-time traffic congestion conditions, including road congestion index, average vehicle speed, etc.

[0122] Step 52, preprocess the real-time road conditions, real-time weather data, and traffic congestion information to obtain preprocessed real-time road conditions, real-time weather data, and traffic congestion information, specifically including: removing duplicate, incorrect, or incomplete data records; unifying data from different sources and formats into a standard format; extracting key features that affect path planning from the original data, such as road surface slipperiness level, visibility grade, rainfall amount, etc.; fusing data from different sources to form a comprehensive data set containing multi-dimensional information.

[0123] Step 53, represent the transportation path as a series of nodes, each node containing location information, as well as preprocessed real-time road conditions, real-time weather data, and traffic congestion information, specifically including: dividing the transportation path into a series of nodes, each node representing a specific geographical location or intersection; associating preprocessed real-time road conditions, weather data, and traffic congestion information with each node; forming a complete representation of the transportation path from the starting point to the ending point by arranging these nodes in an orderly manner.

[0124] Step 54, generate a set of initial paths as particles based on each node, each particle representing a solution, specifically including: randomly generating a set of initial paths as a particle swarm based on the defined nodes, each particle representing a possible transportation path solution; ensuring that the generated particles have sufficient diversity in path selection to cover different possible paths.

[0125] Step 55, define an objective function for evaluating the quality of a path based on the real-time road conditions, weather data, and traffic congestion information in the node, where the calculation formula of the objective function is:

[0126]

[0127] where, A a is the number of accidents in the history of this path; A max is the maximum value of the number of accidents among all paths; C t is the traffic congestion index; C maxis the congestion index with the worst traffic among all paths; V s is the average vehicle speed on the current path; V max is the maximum designed vehicle speed of this path; H h is the real-time humidity of the area where the path is located; H max is the maximum humidity of this path; T t is the real-time temperature on the current path; T i is the ideal temperature range (e.g., 20°C - 25°C); T max and T min are the maximum and minimum temperature values of this path respectively; D d is the current road damage degree of the path; D t is the threshold of the damage degree; α is a parameter controlling the steepness of the curve; V c is the current visibility on the path; V max is the ideal maximum visibility on this path; w1, w2, w3 and w4 represent weight coefficients; α1, β, γ, ∈ and ζ represent weight coefficients. Among them:

[0128]

[0129] Among them, D cr represents the total crack length (unit: meter); D p represents the total number of potholes; D pot represents the average depth of potholes (unit: centimeter); D de represents the pavement deformation area (unit: square meter); D wear represents the wear depth (unit: centimeter); L total is the total road length; N max is the maximum allowable number of potholes; is the maximum pothole depth; A total is the total road area; is the maximum wear depth; w6, w7, w8, w9 and w 10 represent weight coefficients.

[0130] Step 56: Evaluate each particle using the objective function to calculate its fitness value; according to the fitness value of the particle and the preset update strategy, combined with the changes in real-time data, dynamically adjust the state of the particle to search for the corresponding path, and repeat the particle evaluation and state update process until the preset number of iterations is reached, so as to select the corresponding path from the final particle swarm as the transportation path. Specifically, it includes: evaluating each particle (i.e., each path) using the defined objective function to calculate its fitness value, which reflects the quality of the path in meeting the requirements of the objective function; according to the fitness value of the particle and the preset update strategy (such as the velocity update formula, position update formula, etc.), combined with the changes in real-time data, dynamically adjust the state of the particle (i.e., path selection); repeat the particle evaluation and state update process until the preset number of iterations is reached; select the path with the optimal fitness value from the final particle swarm as the recommended transportation path. Among them, the velocity update formula determines the moving direction and speed of the particle in the search space. In PSO, each particle adjusts its velocity according to its own historical best position and the global best position.

[0131] v i (t + 1) = w·v i (t) + c1·r1·(pbest i -x i (t)) + c2·r2·(gbest - x i (t));

[0132] Among them, v i (t) is the velocity of particle i at time t; w is the inertia weight, used to balance the global and local search capabilities; c1 and c2 are learning factors, usually set as positive numbers; r1 and r2 are random numbers within the range of [0, 1]; pbest i is the historical best position of particle i; gbest is the global best position; x i (t) is the position of particle i at time t.

[0133] Position update formula: The position update formula adjusts the position of the particle according to the updated velocity.

[0134] x i (t + 1) = x i (t) + v i (t + 1);

[0135] Among them, x i (t + 1) is the position of particle i at time t + 1.

[0136] In the embodiments of the present invention, by obtaining road conditions, weather data, and traffic congestion information in real time and incorporating them into route planning considerations, the system can respond immediately to changes in the external environment. This real-time nature ensures flexible adjustment of transportation routes, effectively avoiding adverse conditions and improving transportation efficiency. Based on the preprocessed real-time data, the system can anticipate potential risk points during the route planning stage, such as the impact of bad weather, traffic congestion, or poor road conditions. This helps to avoid these risks in advance and reduce delays and losses during transportation. The particle optimization algorithm can search for the optimal solution globally while considering multiple influencing factors. This ensures that the selected route is not only optimal in a single dimension (such as distance or time), but also globally optimal after comprehensively considering multiple dimensions (such as safety, cost, timeliness, etc.). By continuously evaluating the fitness values of particles and dynamically adjusting the state of particles according to changes in real-time data, the system can continuously optimize the transportation route. This adaptive performance enables the system to remain efficient and stable in a changing environment. By selecting the optimal route, the system can reduce unnecessary detours and waiting times, thereby improving the overall transportation efficiency. At the same time, avoiding potential risk points also enhances the reliability of transportation and reduces the risk of cargo damage or delayed delivery.

[0137] In a preferred embodiment of the present invention, step 6, according to the transportation route and the current road conditions and weather conditions, evaluate the potential risks of each route to obtain a risk assessment result, may include:

[0138] Step 61, obtain the basic information of each transportation route, including the route length, passing areas, and key nodes; combine GIS data to analyze the terrain, road grade, and historical accident data of the route, specifically including: through a map service or a GIS (Geographic Information System) platform, obtain the basic data of each transportation route, including the total length of the route, the main passing areas (such as cities, villages, mountains, etc.), and key nodes (such as transportation hubs, bridges, tunnels, etc.); use GIS data to conduct a detailed analysis of the terrain of the route, such as whether it includes steep slopes, sharp turns, river crossings, etc., which may all affect transportation safety; combine the data of the road management department to analyze the road grade of the route (such as expressways, national highways, provincial highways, etc.) and the historical accident data on this route to understand the frequency and type of accidents.

[0139] Step 62: Obtain real-time road condition data, including road surface slipperiness, visibility, and traffic flow; obtain real-time weather forecast and actual data, including key meteorological information for the area where the transportation route is located. The key meteorological information includes temperature, humidity, wind speed, and precipitation probability. Specifically, collect data such as road surface slipperiness, visibility (affected by fog, haze, etc.), and traffic flow through real-time monitoring devices such as road surface sensors and traffic cameras. These data can be obtained through the traffic management center or relevant service platforms; cooperate with meteorological service agencies to obtain real-time weather forecast and actual data for the area where the transportation route is located, including key meteorological information such as temperature, humidity, wind speed, and precipitation probability.

[0140] Step 63: Determine risk assessment factors affecting transportation safety, including road surface quality, road width, severe weather index, visibility, congestion level, and accident frequency. Specifically, based on historical data and expert experience, screen out risk assessment factors that have a significant impact on transportation safety. These factors include road surface quality (such as potholes, cracks, etc.), road width, severe weather index (comprehensively considering meteorological factors such as temperature, humidity, wind speed, etc.), visibility, congestion level, and accident frequency, etc.; define each risk assessment factor and determine its quantification method. For example, road surface quality can be represented by regular road surface inspection scores, and congestion level can be calculated through traffic flow and average vehicle speed, etc.

[0141] Step 64: Set corresponding weights for each risk assessment factor and establish a quantitative risk calculation model. The risk calculation model outputs a specific risk value based on each risk assessment factor. Specifically, use the AHP method to conduct pairwise comparisons of each risk assessment factor through a questionnaire survey to determine their relative importance; based on the results of the questionnaire survey, construct a judgment matrix, where each element represents the relative importance ratio between two factors; by solving the maximum eigenvalue and the corresponding eigenvector of the judgment matrix, the weights of each risk assessment factor can be obtained, and these weights reflect the relative magnitudes of the impacts of each factor on transportation safety; to ensure the consistency of the judgment matrix, conduct a consistency test. If the consistency ratio (CR) is less than or equal to a predetermined threshold (usually 0.1), it is considered that the judgment matrix has satisfactory consistency, otherwise, the judgment matrix needs to be readjusted. For each risk assessment factor, its quantification method needs to be determined.

[0142] For example, the road surface quality can be quantified by the road surface detection score, and the visibility can be quantified by the measured value of the visibility meter, etc. Based on the determined weights and the quantified values of each risk assessment factor, a comprehensive risk calculation formula is established; historical data is used to verify the risk calculation formula to ensure that the risk calculation formula can accurately reflect the actual situation. If deviations or deficiencies are found in the risk calculation formula, the risk calculation formula is optimized and adjusted to improve its accuracy and reliability; data of each risk assessment factor is collected in real time and preprocessed (such as data cleaning, standardization, etc.) to ensure the accuracy and consistency of the input data; the preprocessed data is input into the established risk calculation formula, and a specific risk value is obtained according to the result output by the risk calculation formula. The calculated risk value is interpreted and applied. For example, the risk value can be compared with a preset threshold to determine whether the transportation risk level of the current path is acceptable. If the risk value exceeds the threshold, corresponding safety measures need to be taken to reduce the risk.

[0143] Among them, the risk calculation formula can specifically be:

[0144]

[0145] Among them, R is the finally calculated risk value; w i is the weight corresponding to the i-th risk assessment factor; f i is the quantified value of the i-th risk assessment factor, and this value can be obtained through actual measurement or can be converted through a certain scoring mechanism; n is the total number of risk assessment factors.

[0146] Step 65, input the road and weather data collected in real time into the risk calculation model to conduct real-time risk assessment on each transportation path to obtain a specific risk value, which specifically includes: input the road and weather data collected in real time into the established risk calculation model, and this data includes the road condition monitoring data obtained in Step 62 and the weather forecast and actual situation data; through model calculation, conduct real-time risk assessment on each transportation path to obtain a specific risk value, and this risk value can help decision-makers judge the safety of the current path and make reasonable transportation arrangements and adjustments accordingly; output the risk assessment result in the form of an alarm so that relevant personnel can timely understand and respond to potential risks; according to the risk assessment result, response strategies such as changing the transportation route, adjusting the transportation time, and increasing safety measures can be taken to reduce the risk.

[0147] In the embodiments of the present invention, by conducting a detailed risk assessment on each transportation route, potential high-risk areas can be discovered and avoided in a timely manner, thereby significantly reducing the probability of accidents during transportation and ensuring the safety of personnel and goods. Based on real-time road and weather data, this process can dynamically evaluate the risk levels of various routes, helping decision-makers select the safest and most efficient transportation routes under current conditions and improving logistics efficiency. Through a quantitative risk calculation model, multiple complex risk factors are comprehensively considered and output in the form of specific risk values, making the decision-making process more objective and reducing the uncertainty brought by subjective judgment. When encountering sudden weather changes or deteriorating road conditions, this process can quickly re-evaluate the route risks, providing strong support for timely adjustment of transportation plans and enhancing the flexibility and timeliness in dealing with emergencies. Continuously collecting and analyzing road, weather, and accident data can provide valuable operational insights for transportation enterprises, helping them continuously optimize transportation strategies and improve overall management levels.

[0148] In another preferred embodiment of the present invention, step 7 above, determining corresponding risk response strategies according to the risk assessment results, may include:

[0149] Based on the finally calculated risk values, to understand the specific risk levels of each transportation route; compare the risk values of different routes, identify high-risk and low-risk routes, and analyze the main factors leading to high risks, such as severe weather in specific areas, poor road conditions, or frequent historical accidents, etc. Develop risk response strategies:

[0150] For high-risk routes:

[0151] Consider avoiding or replacing this route and select an alternative route with lower risk; if it is necessary to use a high-risk route, safety measures should be increased, such as strengthening vehicle maintenance, equipping emergency rescue equipment, arranging experienced drivers, etc. For medium-risk routes:

[0152] Monitor and respond in a timely manner to any adverse changes on the route, such as sudden weather events or road repairs. Regularly inspect and maintain the vehicles to ensure they are suitable for driving on such risky routes. Provide driver training and guidance to enhance their ability to identify and respond to potential risks.

[0153] For low-risk routes:

[0154] Continue to maintain monitoring and regularly evaluate whether the risk levels of the routes have changed. Encourage the use of these routes for transportation to optimize resource utilization and reduce overall risks.

[0155] Establish an emergency response plan:

[0156] Develop a detailed emergency response plan for possible emergencies (such as traffic accidents, path closures due to bad weather, etc.). Ensure that all relevant personnel are aware of the emergency plan and conduct regular training and drills. Establish a 24-hour emergency contact channel to enable rapid coordination of resources and response when needed.

[0157] Monitor the risk status of each path in real time and adjust the risk response strategy according to the actual situation; regularly review and evaluate the effectiveness of the risk response strategy and update and optimize it as needed; communicate with relevant stakeholders (such as customers, suppliers, etc.) to jointly address the risk challenges during transportation.

[0158] As Figure 2 shown, an embodiment of the present invention also provides a supply chain management system based on Internet of Things technology, including:

[0159] An acquisition module for real-time collection of multi-dimensional data during the transportation of goods, where the multi-dimensional data includes the location, temperature, and humidity of the goods;

[0160] An analysis module for analyzing the multi-dimensional data, and obtaining the temperature and humidity change trends of the goods during transportation by comparing the temperature and humidity data at different time points;

[0161] A prediction module for predicting future temperature and humidity conditions based on the temperature and humidity change trends of the goods during transportation;

[0162] An identification module for identifying abnormal points according to the future temperature and humidity conditions and the goods location data through a preset dynamic factor;

[0163] An optimization module for obtaining the corresponding transportation path by using the particle swarm optimization algorithm according to the abnormal points, real-time road conditions, weather information, and traffic congestion;

[0164] An evaluation module for evaluating the potential risks of each path according to the transportation path and the current road conditions and weather conditions to obtain a risk assessment result; and determining the corresponding risk response strategy according to the risk assessment result.

[0165] It should be noted that this system corresponds to the above method, and all implementation manners in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0166] An embodiment of the present invention also provides a computing device, including: a processor and a memory storing a computer program. When the computer program is run by the processor, it executes the method as described above. All implementation manners in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0167] An embodiment of the present invention also provides a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to execute the method described above. All implementation manners in the above method embodiments are applicable to this embodiment and can also achieve the same technical effects.

[0168] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A supply chain management system based on Internet of Things technology, characterized in that: include: The acquisition module is used to collect multi-dimensional data in the process of cargo transportation in real time, including cargo location, temperature, and humidity; The analysis module is used to analyze multi-dimensional data and compare the temperature and humidity data at different time points to obtain the temperature and humidity change trends of the goods during transportation; The prediction module is used to predict the future temperature and humidity conditions based on the temperature and humidity change trends of the goods during transportation; Identification module, used to identify abnormal points through preset dynamic factors according to future temperature and humidity conditions and cargo location data; The optimization module is used to use the particle optimization algorithm to obtain the corresponding transportation route according to abnormal points, real-time road conditions, weather information and traffic congestion; The assessment module is used to assess the potential risks of each route based on the transportation route and current road conditions and weather conditions to obtain risk assessment results; based on the risk assessment results, determine the corresponding risk response strategy.

2. The supply chain management system based on Internet of Things technology according to claim 1 is characterized in that: Analyze multi-dimensional data and compare temperature and humidity data at different time points to obtain the temperature and humidity change trends of the goods during transportation, including: The temperature and humidity sensors are used to collect multi-dimensional data in real time during the transportation of goods, including timestamps, temperature and humidity values; the multi-dimensional data is standardized to obtain standardized multi-dimensional data; Sort the standardized multi-dimensional data by timestamp to construct the time series of temperature and humidity; Based on the time series of temperature and humidity, identify the initial trend of temperature and humidity changes; Based on the initial change trend, an initial temperature and humidity change trend model is generated; Generate a neighborhood solution of the current solution through a random perturbation strategy; define an objective function for evaluating the fitting effect; decide whether to accept the neighborhood solution as the new current solution according to the Metropolis criterion of the simulated annealing algorithm, and update the temperature parameter in the simulated annealing algorithm; The steps of neighborhood solution generation, objective function evaluation, and acceptance criterion judgment are repeated until the termination condition is met to obtain the final temperature and humidity change trends.

3. The supply chain management system based on Internet of Things technology according to claim 2 is characterized in that: According to the temperature and humidity change trends of the goods during transportation, the future temperature and humidity conditions are predicted, including: According to the temperature and humidity change trends of the goods during transportation, an autoregressive integrated moving average model is constructed; The collected temperature and humidity data are used to train the autoregressive integrated moving average model to obtain a trained autoregressive integrated moving average model; Based on the trained autoregressive integrated moving average model, future temperature and humidity conditions are predicted.

4. The supply chain management system based on Internet of Things technology according to claim 3 is characterized in that: The collected temperature and humidity data are used to train the autoregressive integrated moving average model to obtain the trained autoregressive integrated moving average model, including: The three parameters p, d and q of the autoregressive integrated moving average model are encoded as individuals in the genetic algorithm, where p represents the order of the autoregressive term; d represents the number of differences; and q represents the order of the moving average term. Generate an initial population, which contains multiple individuals, each of which represents a parameter combination of an autoregressive integrated moving average model; Define a fitness function for evaluating the quality of each individual; select the corresponding individual for reproduction according to the fitness function, and exchange the parameters of two individuals through crossover operation to generate new offspring individuals; randomly change the individuals, and continuously adjust the three parameters p, d and q of the autoregressive integrated moving average model through genetic algorithm until the preset number of iterations is reached to obtain the optimized parameters p, d and q; Calculate the error of each autoregressive integrated moving average model using the training set data and the test set data; According to the error of the autoregressive integrated moving average model, the final parameters p, d and q are selected according to the error, and the ARIMA model is trained with the final parameters p, d and q to obtain the trained autoregressive integrated moving average model.

5. The supply chain management system based on Internet of Things technology according to claim 4 is characterized in that: Based on future temperature and humidity conditions and cargo location data, abnormal points are identified through preset dynamic factors, including: Collect cargo location data in real time and calculate dynamic factors based on cargo location data, transportation routes, historical data and weather forecast information; Correlate the real-time temperature and humidity data with the location data of the goods to form a comprehensive data set containing timestamps, temperature and humidity values, and location information; Based on the cargo location data, the comprehensive dataset is divided into different segments, each segment representing a time period during the cargo transportation process; For the data of each time period, calculate the absolute difference between the actual temperature and humidity values ​​and the future temperature and humidity; Multiply the absolute difference by the dynamic factor to get a weighted difference value; Set a threshold. If the weighted difference value exceeds the threshold, the corresponding point will be marked as an outlier.

6. The supply chain management system based on Internet of Things technology according to claim 5 is characterized in that: According to abnormal points, real-time road conditions, weather information and traffic congestion, particle optimization algorithm is used to obtain the corresponding transportation path, including: Get real-time road conditions, including road conditions and visibility; get real-time weather data, including temperature, humidity, wind speed, rainfall and snow conditions; get real-time traffic congestion information; Preprocessing the real-time road condition, the real-time weather data and the traffic congestion information to obtain preprocessed real-time road condition, the real-time weather data and the traffic congestion information; The transportation route is represented as a series of nodes, each of which contains location information, as well as pre-processed real-time road conditions, real-time weather data, and traffic congestion information; Based on each node, a set of initial paths are generated as particles, each particle represents a solution; Define the objective function for evaluating the quality of the path based on the real-time road conditions, weather data, and traffic congestion information in the node; Each particle is evaluated using the objective function to calculate its fitness value. Based on the particle's fitness value and the preset update strategy, combined with changes in real-time data, the particle's state is dynamically adjusted to search for the corresponding path. The particle evaluation and state update process is repeated until the preset number of iterations is reached to select the corresponding path from the final particle swarm as the transportation path.

7. The supply chain management system based on Internet of Things technology according to claim 6 is characterized in that: Based on the transportation route and current road and weather conditions, the potential risks of each route are assessed to obtain risk assessment results, including: Obtain basic information on each transport route, including route length, transit areas and key nodes; combine GIS data to analyze the route's topography, road grade and historical accident data; Obtain real-time road condition data, including road slipperiness, visibility, and traffic flow; obtain real-time weather forecasts and live data, including key meteorological information for the transportation route area, including temperature, humidity, wind speed, and precipitation probability; Identify risk assessment factors that affect transportation safety, including road quality, road width, severe weather index, visibility, congestion level and accident frequency; Set corresponding weights for each risk assessment factor and establish a quantitative risk calculation model. The risk calculation model outputs a specific risk value based on each risk assessment factor. The road and weather data collected in real time are input into the risk calculation model to conduct real-time risk assessment on each transportation route to obtain a specific risk value.

8. A supply chain management method based on Internet of Things technology, characterized in that: A system for executing any one of claims 1 to 7, comprising: Real-time collection of multi-dimensional data during cargo transportation, including cargo location, temperature, and humidity; Analyze multi-dimensional data and compare temperature and humidity data at different time points to obtain the temperature and humidity change trends of the goods during transportation; Predict future temperature and humidity conditions based on the temperature and humidity change trends of the goods during transportation; Based on future temperature and humidity conditions and cargo location data, abnormal points are identified through preset dynamic factors; According to abnormal points, real-time road conditions, weather information and traffic congestion, particle optimization algorithm is used to obtain the corresponding transportation path; Based on the transportation route and current road conditions and weather conditions, the potential risks of each route are assessed to obtain risk assessment results; based on the risk assessment results, the corresponding risk response strategies are determined.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the system according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, and when the program is executed by a processor, the system according to any one of claims 1 to 7 is implemented.

Images