Method and system for identifying anomalies in logistics transport

Abstract

The application relates to the technical field of logistics monitoring, in particular to an abnormality identification method and system for logistics transportation, which collects transportation live data of a transportation carrier and combines the transportation live data into transportation monitoring records, samples and inspects the records at a transfer station to obtain sampling and inspection data, analyzes the overall quality by using the sampling data, obtains inferred quality information, analyzes the transportation monitoring records according to the inferred quality information, identifies transportation abnormalities, can comprehensively master the transportation conditions, makes the sampling and inspection more targeted, accurately evaluates the overall quality of goods, quickly and accurately identifies transportation abnormalities, discovers and handles problems in time, guarantees the quality of goods, reduces logistics cost, improves transportation efficiency and service level.

Description

Technical Field

[0001] This application relates to the technical field of logistics monitoring, and in particular to a method and system for identifying anomalies in logistics transportation. Background Technology

[0002] Modern logistics transportation encompasses various modes of transport, including road, rail, air, and water. These different modes have significantly different characteristics and environmental conditions. Goods undergo multiple transfer points during transit, increasing the risk of damage and the likelihood of unforeseen circumstances. Current technology typically only inspects goods at the beginning and end of transit, making it difficult to detect problems in time. Conversely, conducting thorough inspections during all transfers would be inefficient, wasting unnecessary time and effort. Summary of the Invention Therefore, it is necessary to provide a method and system for identifying anomalies in logistics transportation to address the aforementioned technical problems and to promptly detect abnormal phenomena in logistics transportation.

[0003] Firstly, this application provides a method for identifying anomalies in logistics transportation, the method comprising: Collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record; Based on the aforementioned transportation monitoring records, sampling and quality inspection of the target cargo batches are conducted at the transit station to obtain sampling and quality inspection data. Based on the sampling quality inspection data, the overall quality of the target goods batch is analyzed to obtain the inferred quality information of the target goods batch. Based on the inferred quality information, the abnormal data of the transportation monitoring records is analyzed to obtain the abnormal identification information of the transportation vehicles between various transfer stations.

[0004] Secondly, this application also provides an anomaly identification system for logistics transportation, used to implement the anomaly identification method for logistics transportation as described in any one of the first aspects, comprising: The data acquisition module is used to collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record. The sampling and quality inspection module is used to conduct sampling and quality inspection of the target goods batch at the transit station based on the transportation monitoring records, so as to obtain sampling and quality inspection data. The quality estimation module is used to analyze the overall quality of the target goods batch based on the sampling quality inspection data, and obtain the estimated quality information of the target goods batch. An anomaly identification module is used to parse the abnormal data of the transportation monitoring records based on the inferred quality information to obtain anomaly identification information of the transportation vehicle between various transfer stations.

[0005] The aforementioned method for identifying anomalies in logistics transportation collects real-time transportation data and accumulates it into monitoring records. This allows for a comprehensive understanding of the transportation situation. Sampling and quality inspection based on these records makes the inspection more targeted. Analyzing the overall quality based on the sampled data yields inferential information, providing a basis for anomaly judgment. Analyzing abnormal data accurately identifies anomalies between various transit stations. This helps to promptly detect transportation problems, ensure cargo quality, reduce corporate losses, improve logistics transportation efficiency and service quality, and enhance corporate competitiveness. Attached Figure Description

[0006] Figure 1 This is a schematic diagram illustrating the steps of an anomaly identification method in logistics transportation in one embodiment; Figure 2 This is a schematic diagram of the structure of an anomaly identification system for logistics transportation in one embodiment. Detailed Implementation

[0007] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0008] The logistics transportation anomaly identification method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown: S1: Collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record; S2: Based on the transportation monitoring records, sample and inspect the target cargo batch at the transit station to obtain sampling and inspection data; S3: Analyze the overall quality of the target goods batch based on the sampling quality inspection data to obtain the inferred quality information of the target goods batch; S4: Based on the inferred quality information, analyze the abnormal data of the transportation monitoring record to obtain the abnormal identification information of the transportation vehicle between each transfer station.

[0009] Specifically, in step S1 of the embodiments provided in this application, various sensors, including but not limited to temperature sensors, humidity sensors, vibration sensors, and acceleration sensors, are installed at key locations on the transport vehicle (such as trucks, ships, airplanes, etc.). These sensors are connected by wired or wireless means to form a sensor network. The sensors collect environmental data inside and outside the transport vehicle in real time, such as temperature, humidity, vibration frequency, and acceleration, and transmit this data to a data acquisition device. The data acquisition device can be an on-board computer or a dedicated data acquisition module, which performs preliminary processing and storage on the data collected by the sensors.

[0010] More specifically, various factors in the transportation environment (such as temperature, humidity, vibration, etc.) can affect the quality of goods. By monitoring the transportation environment in real time, we can keep track of environmental changes during transportation, providing a basis for subsequent quality analysis and anomaly identification. Sensor networks can collect data in real time, ensuring the accuracy and timeliness of the data. Compared with manual recording or periodic inspections, real-time monitoring can capture instantaneous changes during transportation and more comprehensively reflect the actual situation of the transportation environment.

[0011] More specifically, before the transport vehicle departs, ensure that the clock of the data acquisition device is synchronized with the standard time to guarantee the accuracy of the timestamps. When the transport vehicle leaves the transit station, the data acquisition device automatically adds a timestamp to the real-time transport data collected from that moment. The timestamp can be accurate to the second or even millisecond to accurately record the data collection time.

[0012] More specifically, timestamps provide time-dimensional information for real-time transportation data, enabling subsequent sorting and analysis of the data in chronological order. Through timestamps, it is possible to clearly understand the environmental changes at each point in time during transportation, which helps to identify the time points that lead to cargo quality issues. During logistics transportation, transport vehicles will pass through multiple transit stations, and timestamps can help distinguish transportation data between different transit stations, facilitating separate analysis of each transportation stage.

[0013] More specifically, the data acquisition equipment or back-end data analysis system sorts the real-time transportation data according to the timestamp, arranges the data in chronological order, and integrates the sorted data into a complete dataset to form a transportation monitoring record of the vehicle between two transfer stations. This record contains all the real-time transportation data during the transportation phase, such as the changes in temperature, humidity, vibration, etc. over time.

[0014] More specifically, by arranging data in chronological order, various data points during the transportation process can be connected sequentially to form a complete transportation monitoring record. This allows for a clear view of the changes in the transportation environment between two transit points, facilitating subsequent analysis and processing. Data arranged in chronological order is also easier to analyze and mine. For example, by analyzing the temperature change curve in the transportation monitoring record, it can be determined whether there are any abnormal temperatures during transportation; by analyzing vibration data, the impact of the degree of bumps during transportation on the goods can be assessed.

[0015] Specifically, in step S2 of the embodiment provided in this application, data related to the transportation environment, such as temperature, humidity, vibration frequency, and acceleration, are extracted from the transportation monitoring records. Statistical methods are used to analyze the extracted data, and statistical quantities such as the average, maximum, minimum, and standard deviation of each environmental indicator are calculated. For example, the average temperature, maximum temperature, and minimum temperature during transportation, as well as the temperature fluctuation range, are calculated. Based on the results of the statistical analysis, the transportation environment characteristics of the transport vehicle between two transfer stations are summarized, such as whether the temperature is stable and whether the vibration is severe. Different transportation environment factors will have different degrees of impact on the quality of goods. By analyzing the transportation environment characteristics, we can understand the environmental conditions faced by goods during transportation, providing a basis for subsequent prediction of changes in the quality of goods. Different transportation environment characteristics lead to different types of quality problems in goods. For example, high temperature environments can cause food to spoil, and severe vibration can damage fragile items. After understanding the transportation environment characteristics, we can determine the focus and methods of sampling quality inspection in a targeted manner.

[0016] More specifically, data on the type and arrangement of goods in the transport vehicle for the target goods batch are obtained, and a digital model of the target goods batch is generated based on this data. The digital model can simulate the state and interaction of goods during transportation. The transportation impact simulation is performed on the digital model according to the characteristics of the transportation environment. For example, the spoilage process of food under high temperature conditions or the damage of vibration to fragile items is simulated. Data statistics are performed on the quality loss prediction information of each item in the target goods batch to obtain an impact prediction list. This list records the quality problems of different goods in the current transportation environment and their corresponding probabilities.

[0017] More specifically, by predicting the quality impact of target cargo batches, we can understand in advance the quality problems that may occur during transportation, providing guidance for subsequent sampling and quality inspection. The impact prediction list can help determine the focus and scope of sampling, prioritizing the inspection of goods that may have quality problems, thereby improving the efficiency and accuracy of sampling and quality inspection.

[0018] More specifically, based on the impact prediction list, a representativeness analysis is performed on the sampling of each item within the target cargo batch to obtain the representative value of each item. Items with higher representative values ​​are better able to reflect the overall quality of the cargo batch. Based on the cargo type and arrangement data of the target cargo batch in the transport vehicle, the sampling difficulty of each item is analyzed to obtain the sampling difficulty level. For example, items located at the bottom of the cargo stack are more difficult to sample. Combining the representative value and sampling difficulty, sampling targets are selected for the target cargo batch to generate a sampling plan. The sampling plan should comprehensively consider the representativeness and feasibility of the sampling. Sampling quality inspection is carried out on the target cargo batch according to the sampling plan, and appropriate testing methods and equipment are used to conduct quality testing on the sampled goods to obtain sampling quality inspection data.

[0019] More specifically, by comprehensively considering the representativeness and ease of sampling, the generated sampling plan can ensure that the sampled goods are representative and operable, thereby improving the effectiveness of sampling quality inspection. Through sampling quality inspection, actual quality data of the target batch of goods can be obtained, providing a basis for subsequent analysis of the overall quality of the goods.

[0020] Specifically, in step S3 of the embodiment provided in this application, a digital model of the target goods batch is called from a pre-established database. The digital model contains detailed information such as the type, quantity, and placement of the goods. Data obtained from sampling inspection, such as the quality indicators of the goods (e.g., weight, size, appearance defects, etc.) and test results (qualified or unqualified), are marked on the corresponding goods in the digital model. The marking method can be to directly display specific values ​​or status indicators on the digital model.

[0021] More specifically, by marking sampling quality inspection information on the digital model, the position and distribution of the quality of the sampled goods in the entire batch of goods can be seen intuitively. This helps to quickly understand the relationship between the sampled goods and the whole batch of goods, providing an intuitive basis for subsequent overall quality prediction. Combining sampling quality inspection data with the digital model achieves effective data integration and correlation, which facilitates further analysis and processing of the data and improves the accuracy and efficiency of the analysis.

[0022] More specifically, statistical analysis is performed on the sampled quality inspection labeling information to identify the distribution patterns and characteristics of the sampled goods' quality. For example, this involves calculating the proportion of qualified products, the average value and standard deviation of quality indicators in the sampled goods, etc. Based on the quality distribution patterns and characteristics of the sampled goods, an appropriate information inference method is selected. Common methods include probabilistic statistical methods and machine learning algorithms. For instance, regression analysis can be used to predict the quality indicators of unsampled goods. Using the selected information inference method, quality information is inferred for unsampled goods in the digital model, and the inference results are labeled on the digital model to form the inferred quality information for the target batch of goods.

[0023] More specifically, since it is impossible to inspect all goods, by inferring information from unsampled goods, we can understand the quality of the entire batch to a certain extent, and achieve a comprehensive assessment of the overall quality of the goods. The inferred quality information can serve as an important basis for subsequent anomaly identification. By comparing the inferred quality information with the expected quality standards, we can promptly identify potential quality anomalies and take corresponding measures to deal with them.

[0024] Specifically, in step S4 of the embodiments provided in this application, specific indicators for measuring quality loss are determined based on the nature and characteristics of the goods, such as the quantity of damaged goods, weight reduction, degree of functional loss, etc. Data related to quality loss are extracted from the inferred quality information and calculated according to the predetermined measurement indicators. For example, if the quantity of damaged goods is used as the indicator, the quantity of goods marked as damaged in the inferred quality information is counted; if the weight reduction is used as the indicator, the difference between the actual weight of the goods and the standard weight is calculated. All quality loss data are summarized to obtain the quality loss data of the target goods batch, such as the total quantity of damaged goods, total weight loss, etc.

[0025] More specifically, by quantifying the quality loss of goods through data statistics, the quality loss situation can be presented in a specific data form, which is convenient for subsequent analysis and comparison. Quality loss data is the basis for further analysis of the cause of loss and identification of anomalies. Only by clarifying the specific situation of quality loss can subsequent steps be carried out in a targeted manner.

[0026] More specifically, quality loss data is presented in charts (such as bar charts, line charts, etc.) or other visualization methods to more intuitively observe the distribution and trend of the data. Data analysis methods are used to extract key features from the quality loss data, such as the central tendency (mean, median), dispersion (standard deviation, variance), and trend (increasing, decreasing, fluctuating). Based on the extracted features, the overall loss characteristics of the target batch of goods are summarized, such as whether the quality loss is concentrated in a certain period of time and whether the degree of loss shows a gradually increasing trend.

[0027] More specifically, by identifying data characteristics, we can better grasp the patterns and characteristics of cargo quality losses, providing clues for tracing the causes of losses. Overall loss characteristics can help determine whether quality losses are in line with normal conditions. If there are situations that do not conform to normal characteristics, then there is an anomaly.

[0028] More specifically, by combining professional knowledge in the field of logistics and transportation, experience from similar past cases, and factors such as the characteristics of the goods and the transportation environment, the overall loss characteristics are analyzed. Based on the analysis results, hypotheses about the causes of the loss of goods quality are proposed. For example, if the overall loss characteristics show that the loss of quality increases significantly during high-temperature periods, then the reason is that the high temperature has damaged the goods. If the loss is concentrated on the bumpy sections of the transportation process, it may be that the vibration has caused damage to the goods. The proposed hypotheses are screened and sorted to eliminate unreasonable or less likely causes, resulting in several expected causes of loss.

[0029] More specifically, by tracing the expected causes of loss through theory, the scope of investigation can be narrowed down to a few possible factors, improving the efficiency of subsequent analysis. The expected causes of loss clarify the direction of data to be searched in the transportation monitoring records, which helps to conduct more targeted data matching.

[0030] More specifically, for each anticipated cause of loss, relevant transportation monitoring record data items are identified. For example, if the anticipated cause of loss is high temperature, the relevant data items include temperature records during transportation; if the anticipated cause of loss is vibration, the relevant data items are the vibration frequency and amplitude recorded by vibration sensors. Data related to the anticipated cause of loss is filtered out from the transportation monitoring records and extracted. The extracted data is then organized and analyzed to generate verification data for each anticipated cause of loss, such as the duration of high-temperature periods and the number of times vibration exceeds a certain threshold.

[0031] More specifically, the confirmatory data generated through data matching can verify whether the expected cause of loss is valid. If the confirmatory data matches the expected cause of loss, it indicates that the cause is highly likely. The confirmatory data provides objective evidence for subsequent judgment of anomalies, making anomaly identification more scientific and accurate.

[0032] More specifically, the confirmatory data of each expected loss cause are compared to analyze their correlations and differences. For example, the overlap between high-temperature periods and quality loss periods, and the correlation between vibration intensity and the degree of cargo damage are compared. Appropriate methods are selected to assess the credibility of each expected loss cause, such as probability statistics methods or expert evaluation methods. Based on the evaluation method, the credibility parameters of each expected loss cause are calculated. The credibility parameters can be expressed as percentages or other quantitative indicators.

[0033] More specifically, by comparing and analyzing and generating credibility parameters, we can distinguish the likelihood of different expected loss causes and give priority to causes with higher credibility. Credibility parameters provide a basis for decision-making in the subsequent handling of expected loss causes and help to allocate resources reasonably for anomaly handling.

[0034] More specifically, expected loss causes with credibility parameters below a certain threshold are eliminated because these causes are less likely to lead to quality loss. For expected loss causes with higher credibility but inaccurate or incomplete descriptions, adjustments and improvements are made to make them more consistent with the actual situation. If multiple expected loss causes are essentially similar, they can be merged into a more concise cause description. The processed expected loss causes are summarized to form anomaly identification information for transport vehicles between various transit stations, clearly pointing out the abnormal factors that lead to cargo quality loss.

[0035] More specifically, by eliminating, adjusting, and integrating the expected causes of loss, interfering factors can be removed, making the anomaly identification information more accurate and clear. Clear anomaly identification information helps logistics companies take targeted measures to handle anomalies, such as adjusting transportation routes and improving packaging methods.

[0036] In one embodiment, the steps of collecting real-time transportation data of the transport vehicles for the target cargo batch and combining the real-time transportation data of the transport vehicles between various transit points along the logistics transportation route into a transportation monitoring record include: S11: By using a sensor network pre-installed on the transport vehicle, the transport environment of the transport vehicle is monitored in real time to obtain real-time transport data of the transport vehicle. S12: After the transport vehicle departs from the transfer station on the logistics transport route, a corresponding timestamp is generated for the real-time transport data starting at that moment; S13: Arrange the real-time transportation data at each moment according to the timestamp until the transport vehicle arrives at another transit station on the logistics transportation route, so as to obtain the transportation monitoring record of the transport vehicle between the two transit stations.

[0037] Specifically, various sensors are installed at different key locations on the transport vehicle. For example, temperature sensors can be installed in the cargo storage area to monitor the temperature around the cargo; humidity sensors are also placed near the cargo to ensure accurate humidity information; vibration sensors are installed on the vehicle chassis or cargo mounting frame to sense vibrations during transport; and acceleration sensors can be installed on the main body of the vehicle to monitor the vehicle's acceleration, deceleration, and other motion states. These sensors are connected to the data acquisition equipment via wired or wireless communication to form a complete sensor network.

[0038] More specifically, the sensor collects transportation environment data at a certain sampling frequency (e.g., once per second). The collected data is transmitted to the data acquisition device via a network. The data acquisition device performs preliminary processing on the sensor data, such as filtering and calibration, and then stores the processed data in a local storage device. At the same time, the data can be uploaded to a remote server via a wireless network for subsequent analysis and processing.

[0039] More specifically, factors such as temperature, humidity, and vibration in the transportation environment directly affect the quality of goods. For example, high temperature and humidity can cause food to spoil, while severe vibration can damage fragile goods. By monitoring the transportation environment in real time, we can promptly detect the potential impact of environmental changes on goods. Real-time transportation data forms the basis for subsequent sampling inspections, overall quality analysis, and anomaly identification. Accurate and comprehensive transportation environment data helps to more precisely assess the quality changes of goods during transportation.

[0040] More specifically, before the transport vehicle departs, ensure that the clock of the data acquisition device is synchronized with the standard time. Synchronization can be achieved with a time server on the Internet via the Network Time Protocol (NTP) to ensure time accuracy. When the transport vehicle leaves the transit point, the data acquisition device detects a departure signal (such as a vehicle start signal) and automatically adds a timestamp to the real-time transport data collected from that moment onwards. The timestamp can be accurate to the millisecond level to ensure the accurate chronological order of the data.

[0041] More specifically, timestamps provide a time dimension identifier for real-time transportation data, enabling accurate tracing of data collection time during subsequent analysis. For example, when a quality problem is discovered with goods, the timestamp can be used to pinpoint the specific transportation stage, thereby identifying potential influencing factors. During logistics transportation, vehicles pass through multiple transit points, and timestamps can clearly distinguish transportation data between different transit points, facilitating separate analysis and evaluation of each transportation stage.

[0042] More specifically, sorting algorithms (such as quicksort, mergesort, etc.) are used to sort the real-time transportation data stored in the data acquisition device or server according to the timestamp. The sorted data is arranged in chronological order to form an ordered dataset. The sorted data is then integrated into a complete transportation monitoring record and stored in a database or file system. The transportation monitoring record contains all transportation environment data of the transportation vehicle between two transit stations, as well as the corresponding timestamp information.

[0043] More specifically, by arranging data in a time sequence, various data points during the transportation process can be connected in chronological order to form a complete transportation monitoring record. This allows for a clear view of the changes in the transportation environment between two transit points, facilitating the analysis of trends in environmental factors during transportation. An orderly transportation monitoring record also makes data analysis and mining easier. For example, by analyzing time series data, abnormal fluctuations in environmental factors during transportation can be identified, thereby further investigating potential anomalies.

[0044] In one embodiment, the step of sampling and inspecting the target cargo batch at the transit station based on the transportation monitoring records to obtain sampling and inspection data includes: S21: Based on the transportation monitoring records, perform a characteristic analysis of the transportation environment of the transport vehicle to obtain the transportation environment characteristics of the transport vehicle between the two transfer stations; S22: Based on the transportation environment characteristics, predict the quality impact of the target cargo batch to obtain a list of predicted impacts of the target cargo batch on the transportation environment characteristics; S23: Based on the impact prediction list, generate a sampling plan for the target cargo batch, so as to conduct sampling quality inspection on the target cargo batch after it arrives at the transit station and obtain sampling quality inspection data.

[0045] Specifically, various data (such as temperature, humidity, vibration frequency, acceleration, etc.) in transportation monitoring records are cleaned to remove outliers and noise. For example, for temperature data that significantly exceeds the normal range, reasonable thresholds can be set for correction or removal. Statistical methods are used to calculate the statistical characteristics of various transportation environmental indicators, such as the average, maximum, minimum, and standard deviation of temperature, to understand the overall temperature level and fluctuation. For vibration frequency, its frequency distribution can be statistically analyzed, and key features can be extracted from the statistical results. For example, a large temperature standard deviation indicates severe temperature fluctuations during transportation; if the vibration frequency occurs frequently within a certain range, that range can be used as the characteristic range of vibration.

[0046] More specifically, the characteristics of the transportation environment directly affect the quality of goods. By analyzing the characteristics of the transportation environment, we can clearly understand the environmental conditions that goods experience during transportation, providing a basis for subsequent prediction of changes in goods quality. Accurate transportation environment characteristics are a key input for quality impact prediction. Only by understanding the specific circumstances of the transportation environment can we more accurately assess its impact on goods quality.

[0047] More specifically, data on the type of goods (such as food, electronic products, etc.) and the arrangement of goods (such as stacking method, fixing method, etc.) of the target goods batch are collected. A digital model of the target goods batch is constructed using computer modeling technology. This model can simulate the physical state and interaction of goods during transportation. The characteristics of the transportation environment are input into the digital model, and the quality changes of goods during transportation are simulated through simulation software. For example, the spoilage process of food under high temperature environment or the damage of vibration to fragile electronic products is simulated. The simulation results are analyzed and organized to obtain a list of predicted impacts of the corresponding transportation environment characteristics of the target goods batch. The list should include possible quality problems (such as food spoilage, electronic product damage, etc.) and their corresponding probabilities.

[0048] More specifically, by predicting the impact of quality issues, a preliminary assessment of potential quality problems in goods can be made before sampling inspection. This helps to determine the focus and scope of sampling. Based on the impact prediction list, sampling objects can be selected in a targeted manner, avoiding blind sampling and thus improving the efficiency and accuracy of sampling inspection.

[0049] More specifically, based on the impact prediction list, a representativeness analysis of the sampling is conducted on each item within the target cargo batch. For example, goods significantly affected by the transportation environment have a higher representative value, while goods less affected have a relatively lower representative value. Considering the cargo type and arrangement data within the transport vehicle, the sampling difficulty of each item is assessed. For instance, goods at the bottom of the cargo stack are more difficult to sample, while those at the top or edge are relatively easier. Combining the representative value and sampling difficulty, the sampling target and quantity are determined, and specific sampling methods, such as random sampling and stratified sampling, are developed to generate a complete sampling plan. The target cargo batch is then sampled according to the sampling plan, and the sampled goods are subjected to quality testing using appropriate testing equipment and methods. The test results are recorded to obtain sampling quality inspection data.

[0050] More specifically, by comprehensively considering the representativeness and ease of sampling, the generated sampling plan can ensure that the sampled goods are representative and operable, thereby improving the scientific nature of sampling quality inspection. Sampling quality inspection data is an important basis for evaluating the overall quality of the target goods batch. Through scientific sampling plans and accurate testing methods, reliable quality information can be obtained, providing support for subsequent overall quality analysis and anomaly identification.

[0051] In one embodiment, the step of predicting the quality impact of a target cargo batch based on the transportation environment characteristics to obtain a list of predicted impacts of the target cargo batch on the transportation environment characteristics includes: S221: Obtain cargo type data and placement data of the target cargo batch in the transport vehicle, and generate a digital model of the target cargo batch based on the cargo type data and placement data; S222: Simulate the transportation impact on the digital model based on the transportation environment characteristics to obtain the quality loss prediction information for each piece of cargo; S223: Statistically analyze the quality loss prediction information of each item in the target batch of goods to obtain an impact prediction list.

[0052] Specifically, the specific type of each item in the target batch of goods is obtained through order information and cargo lists from the logistics system, such as food, electronic products, and fragile items. At the same time, detailed information such as the specifications and materials of the goods is recorded. The placement, stacking method, and fixing status of the goods in the transport vehicle are determined by means of on-site inspection, image recognition, or design drawings of the transport vehicle. For example, a high-definition camera is used to photograph the goods in the transport vehicle, and then the placement of the goods is identified by image analysis software.

[0053] More specifically, using professional modeling software (such as 3D modeling software), a three-dimensional digital model of the target batch of goods is constructed based on the collected data on the type and arrangement of goods. In the model, the shape, size, position of each item of goods and their relative relationships are accurately presented. The goods in the model are given corresponding physical properties, such as mass, hardness, and elasticity, so as to facilitate subsequent simulation of the impact of transportation.

[0054] More specifically, different types of goods have different sensitivities to the transportation environment, and the way they are placed will also affect the forces and interactions of the goods during transportation. By constructing a digital model, these factors can be accurately incorporated into the simulation process, more realistically reflecting the actual state of the goods during transportation. The digital model is the foundation for simulating the impact of transportation, providing specific objects and parameters for subsequent simulation analysis, making the simulation results more accurate and reliable.

[0055] More specifically, environmental characteristics of the transportation environment (such as temperature, humidity, vibration frequency, acceleration, etc.) are used as input parameters and loaded into the simulation environment of the digital model. These parameters can be obtained from transportation monitoring records and set according to time series to simulate the dynamic changes of the environment during transportation. Physical simulation algorithms (such as finite element analysis, multibody dynamics simulation, etc.) are used to simulate the impact of transportation on the digital model. During the simulation, factors such as the forces, heat transfer, and chemical reactions of the goods in the transportation environment are considered to calculate the quality loss of each item of goods at different stages of transportation. After the simulation, the quality loss prediction information of each item of goods is output, including the degree of quality loss, possible quality problems (such as damage, deterioration, etc.) and the probability of occurrence.

[0056] More specifically, by simulating the transportation process, it is possible to predict the potential quality loss of goods under different transportation environments in advance, providing a basis for subsequent sampling inspection and anomaly handling. Understanding the quality loss prediction information of each piece of goods helps to assess the risks during transportation and take timely measures to reduce losses, such as adjusting transportation routes and improving packaging.

[0057] More specifically, the quality loss prediction information for each item of goods is compiled, including the numerical value of quality loss, the type of quality problem, and the probability of occurrence, and stored in a database or table. The compiled data is then statistically analyzed to calculate statistical indicators such as the frequency of occurrence of different quality problems and the average degree of quality loss. For example, the proportion of damaged goods to the total number of goods is calculated, as well as the average quality loss rate of each type of goods. Based on the statistical analysis results, an impact prediction list is generated. The list should include information such as the type of goods, the possible quality problems, the probability of occurrence, and the degree of quality loss, and be sorted according to the probability of occurrence or the degree of quality loss, so as to intuitively understand the impact on the target batch of goods.

[0058] More specifically, through data statistics, the quality loss prediction information of each item can be comprehensively analyzed to fully assess the impact on the target batch of goods during transportation. The impact prediction list provides an important reference for sampling quality inspection, helps to determine the focus and scope of sampling, and improves the efficiency and accuracy of sampling quality inspection.

[0059] In one embodiment, the step of generating a sampling plan for the target cargo batch based on the impact prediction list includes: S231: Based on the aforementioned impact prediction list, perform a sampling representativeness analysis on each item in the target goods batch to obtain the sampling representative value of each item. S232: Based on the data on the type and arrangement of goods in the transport vehicle for the target batch of goods, analyze the sampling difficulty of each item of goods to obtain the sampling difficulty of each item of goods. S233: Combining the sampling representativeness and sampling difficulty, select sampling targets for the target batch of goods to generate a sampling plan.

[0060] Specifically, based on the information in the impact prediction list, indicators related to sampling representativeness are determined, such as the degree to which goods are affected by the transportation environment, the probability of quality problems, and the proportion of different types of goods in the batch. A corresponding weight is assigned to each indicator, and the sampling representativeness of each item is quantified through weighted calculation. For example, goods that are greatly affected by the transportation environment and have a high probability of quality problems are given higher weights; goods that account for a large proportion in the batch are also given appropriate weights. Based on the results of the quantitative analysis, the sampling representative value of each item is calculated. The higher the sampling representative value, the more representative the quality of the entire batch of goods is.

[0061] More specifically, the purpose of sampling is to infer the quality status of the entire batch of goods by inspecting a portion of the goods. By analyzing the representative value of the sampling, the goods that best reflect the quality characteristics of the batch can be selected for sampling, thereby improving the effectiveness and accuracy of sampling. Selecting goods with high representative value for quality inspection can make the test results more representative of the actual quality of the entire batch of goods, thereby improving the reliability of the overall quality assessment of the batch of goods.

[0062] More specifically, different types of goods have varying degrees of sampling difficulty. For example, liquid goods require special sampling tools and methods, making sampling relatively difficult; while sampling solid goods is relatively easy. The arrangement of goods in the transport vehicle affects the convenience of sampling. If goods are stacked tightly, located deep within the transport vehicle, or obstructed by other goods, the sampling difficulty increases; conversely, if goods are neatly arranged and easily accessible, the sampling difficulty is lower. Based on these factors, the sampling difficulty of each type of goods can be quantitatively assessed using a grading method, such as dividing the sampling difficulty into three levels: easy, medium, and difficult, or using specific numerical values.

[0063] More specifically, when designing a sampling plan, the actual operational difficulty of sampling needs to be considered. If the sampling difficulty is too high, it will lead to excessive time and resources being consumed in the sampling process, or even make it impossible to complete the sampling task. Therefore, analyzing the sampling difficulty can ensure that the sampling plan is practically feasible. After understanding the sampling difficulty of each item, sampling resources can be allocated reasonably, and priority can be given to sampling items with lower sampling difficulty and certain representativeness, thereby improving sampling efficiency.

[0064] More specifically, by comprehensively considering the representative value and sampling difficulty of each item, a weighted summation method can be used. Different weights are assigned to the representative value and sampling difficulty to calculate the comprehensive score of each item. The items are then ranked according to their comprehensive scores, with priority given to items with higher comprehensive scores as sampling targets. At the same time, the rationality of the sampling quantity must be considered to ensure that the sampling quantity meets the requirements of quality assessment without excessively increasing sampling costs. Based on the selected sampling targets, the specific sampling methods (such as random sampling, stratified sampling, etc.), sampling time and location, etc., are determined to form a complete sampling plan.

[0065] More specifically, a sampling plan needs to ensure the representativeness of the samples while considering their practical feasibility. By combining the representativeness of the samples with the ease of sampling when selecting the sampling targets, a balance can be found between the two. This ensures that the sampling plan can accurately reflect the quality of the batch of goods and can be smoothly implemented in actual operation. A reasonable sampling plan can improve sampling efficiency, reduce unnecessary sampling costs, and at the same time ensure the accuracy and reliability of the sampling results, providing strong support for subsequent goods quality assessment and anomaly identification.

[0066] In one embodiment, the step of analyzing the overall quality of the target goods batch based on the sampling quality inspection data to obtain the inferred quality information of the target goods batch includes: S31: Call the digital model of the target goods batch, and mark the goods corresponding to the sampling inspection on the digital model according to the sampling inspection data to obtain the sampling inspection marking information of the digital model. S32: Based on the sampling quality inspection labeling information, perform information inference processing on the corresponding unsampled goods in the digital model, and label the results of the information inference on the digital model to obtain the inferred quality information of the target goods batch.

[0067] Specifically, the digital model of the target batch of goods is accurately retrieved from a pre-built and stored database. This digital model should contain detailed information about the goods, such as their location, type, and specifications. The sampling inspection data is matched one-to-one with the goods in the digital model. For example, if goods numbered 001-010 are sampled and inspected, the corresponding 10 goods are found in the digital model. The specific data obtained from the sampling inspection, such as the quality parameters (weight, dimensional deviation, etc.) and quality status (qualified, unqualified, minor defects, etc.) of the goods, are marked on the digital model in a visual way. Different colors, symbols, or text descriptions can be used to represent different quality statuses.

[0068] More specifically, by marking sampling quality inspection information on the digital model, the specific situation of the sampled goods in the entire batch of goods can be seen intuitively, which facilitates a quick understanding of the distribution and characteristics of the goods quality. The marked digital model provides an accurate reference for inferring information about unsampled goods, making the inference process more scientific and reasonable.

[0069] More specifically, statistical analysis is performed on the sampling quality inspection labeling information to identify the patterns and characteristics of the sampled goods' quality. For example, the proportion of qualified products in the sampled goods and the frequency of different quality problems can be calculated. Based on the quality distribution patterns and characteristics of the sampled goods, appropriate information inference methods are selected. Common methods include probability and statistics-based methods and machine learning algorithms (such as regression analysis and decision trees). For example, if there is a certain correlation between the quality of the sampled goods and their location, regression analysis can be used to infer the quality of the unsampled goods. Using the selected inference method, quality information is inferred for the unsampled goods in the digital model. For example, based on the quality of goods in a certain area of ​​the sampled goods, the quality of the unsampled goods in the same area can be inferred. The inferred information is then labeled on the digital model, such as labeling the possible quality status of the unsampled goods and the predicted values ​​of quality parameters, thereby obtaining the inferred quality information of the target batch of goods.

[0070] More specifically, since it is impossible to inspect all goods, by inferring information from unsampled goods, we can understand the quality of the entire batch to a certain extent, and achieve a comprehensive assessment of the overall quality of the goods. The inferred quality information can serve as an important basis for subsequent anomaly identification. By comparing the inferred quality information with the expected quality standards, we can promptly identify potential quality anomalies and take corresponding measures to deal with them.

[0071] In one embodiment, the step of parsing abnormal data from the transportation monitoring records based on the inferred quality information to obtain anomaly identification information of the transport vehicle between various transfer stations includes: S41: Based on the inferred quality information, perform data statistics on the quality loss of the target goods batch to obtain the quality loss data of the target goods batch; S42: Identify the data features of the quality loss data to obtain the overall quality loss characteristics of the target batch of goods; S43: Based on the overall loss characteristics, theoretically trace the causes of loss to obtain several expected causes of loss; S44: Based on each of the expected causes of loss, perform data matching on the transportation monitoring records to generate verification data for each of the expected causes of loss; S45: Compare and analyze the verification data of each of the expected causes of loss to generate a confidence parameter for each of the expected causes of loss; S46: Based on the credibility parameters, eliminate, adjust and merge the expected causes of loss to obtain anomaly identification information of the transport vehicle between each transfer station.

[0072] Specifically, based on the nature and purpose of the goods, specific indicators for measuring quality loss are determined. For example, for food, the quantity of spoilage and the amount of weight reduction can be used; for electronic products, the functional failure rate can be used. Data related to quality loss are selected from the inferred quality information and calculated according to established standards. For example, the quantity of goods that are inferred to be damaged or spoiled is counted, and the total value loss of these goods is calculated. The results of each calculation are compiled and summarized to form the quality loss data of the target batch of goods, which can usually be recorded in the form of tables or documents.

[0073] More specifically, presenting quality loss in concrete data form allows for a more intuitive and accurate understanding of the extent of damage to goods during transportation, providing clear quantitative basis for subsequent analysis. Unified quality loss data facilitates comparison with historical transportation data, industry standards, etc., and helps to identify whether there are any abnormally severe losses during this transportation.

[0074] More specifically, charting tools (such as bar charts, line charts, and pie charts) are used to visually display the quality loss data, allowing for a clearer observation of the data's distribution and trends. Statistical methods, such as calculating the mean, median, and standard deviation, are employed to analyze the central tendency and dispersion of the data. Cluster analysis is used to identify similar groups within the data. The time-series characteristics of the data are observed to determine whether the losses exhibit specific patterns with transportation time. Based on the analysis results, the overall characteristics of the quality loss of the target cargo batch are summarized, such as whether the losses are concentrated in a certain type of cargo or whether the losses intensify during a specific time period.

[0075] More specifically, by identifying data characteristics, we can gain a deeper understanding of the distribution patterns of quality losses across dimensions such as time and cargo type, providing clues for tracing the causes of losses. Overall loss characteristics can reflect differences from normal conditions, helping to quickly identify possible abnormal transportation situations, such as a sudden peak in losses indicating a problem in a certain stage of transportation.

[0076] More specifically, by combining professional knowledge in the field of logistics and transportation, cargo characteristics, and experience from similar past transportation cases, the overall loss characteristics are analyzed. Based on the analysis results, multiple hypotheses are proposed for the causes of cargo quality loss. For example, if cargo is found to be severely damaged during high-temperature periods, it is hypothesized that high temperature is the cause of the quality problem; if a certain type of fragile item suffers significant losses, it is speculated that vibration or collision during transportation is one of the causes. The proposed hypotheses are initially screened to eliminate obviously unreasonable hypotheses, and similar causes are merged and organized to obtain several expected causes of loss.

[0077] More specifically, identifying the expected causes of loss from numerous possible factors greatly narrows the scope of subsequent investigation and analysis, improving work efficiency. Clearly defined expected causes of loss help to more effectively search for relevant data in transportation monitoring records for further verification.

[0078] More specifically, for each anticipated cause of loss, relevant transportation monitoring record data items are identified. For example, if the anticipated cause is high temperature, the relevant data item is the temperature record during transportation; if the anticipated cause is vibration, the relevant data item is the vibration frequency and amplitude recorded by vibration sensors. Corresponding data is filtered out from the transportation monitoring records and extracted. This can be done by writing data query scripts or using the filtering function of data analysis software. The extracted data is then organized and preliminarily analyzed to generate verification data for each anticipated cause of loss. For example, the duration of high-temperature periods and the number of times vibration exceeded the threshold are statistically analyzed.

[0079] More specifically, by matching the expected causes of loss with transportation monitoring records, it is possible to verify whether these causal assumptions are supported by actual data, providing a basis for the credibility of subsequent judgments on the causes of anomalies. Verifying data is objective evidence for judging the causes of loss, ensuring that subsequent decisions are not based on subjective guesswork, but on reliable data information.

[0080] More specifically, the confirmatory data of different expected loss causes are compared to analyze their correlation with the overall loss characteristics. For example, the overlap between high-temperature periods and peak periods of quality loss, and the correlation between vibration intensity and the number of fragile items lost are compared. Based on the data characteristics and the nature of the problem, appropriate methods are selected to assess credibility, such as probability and statistics methods and expert evaluation methods. If the data is abundant and has statistical regularity, probability and statistics methods can be used to calculate the probability of each cause leading to loss. If complex professional judgment is involved, domain experts can be invited to score and evaluate the data. Using the selected evaluation method, the credibility parameters of each expected loss cause are calculated, usually expressed as a percentage or numerical value.

[0081] More specifically, the credibility parameter can intuitively reflect the probability of each expected cause of loss, helping decision-makers distinguish which causes are more likely to actually lead to quality loss. In subsequent handling of abnormal situations, the credibility parameter provides a quantitative reference for decisions such as determining the causes to be prioritized and allocating resources reasonably.

[0082] More specifically, a credibility threshold is set, and expected loss causes with credibility below the threshold are eliminated, as these causes are unlikely to be the main factors leading to quality loss. For expected loss causes with higher credibility but inaccurate or incomplete descriptions, adjustments and improvements are made, such as refining the specific description of the cause and supplementing relevant conditions or scope. If multiple expected loss causes are essentially similar or related, they can be merged to form a more concise and comprehensive cause description. The processed expected loss causes are then organized and summarized, and combined with relevant data from transportation monitoring records, such as the time and location of the anomaly, to generate anomaly identification information for the transport vehicle between various transfer stations, presented in a clear and concise report format.

[0083] More specifically, eliminating invalid information, adjusting inaccurate descriptions, and integrating similar causes can make anomaly identification information more refined and accurate, highlighting key anomaly factors. Accurate and clear anomaly identification information helps logistics companies quickly develop targeted solutions, take effective measures to handle transportation anomalies, reduce losses, and prevent similar problems from recurring.

[0084] In one embodiment, such as Figure 2 As shown, an anomaly identification system for logistics transportation is provided, used to implement the anomaly identification method for logistics transportation as described in any one of the first aspects, including: The data acquisition module is used to collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record. The sampling and quality inspection module is used to conduct sampling and quality inspection of the target goods batch at the transit station based on the transportation monitoring records, so as to obtain sampling and quality inspection data. The quality estimation module is used to analyze the overall quality of the target goods batch based on the sampling quality inspection data, and obtain the estimated quality information of the target goods batch. An anomaly identification module is used to parse the abnormal data of the transportation monitoring records based on the inferred quality information to obtain anomaly identification information of the transportation vehicle between various transfer stations.

[0085] In this embodiment, the specific implementation of each module in the above system embodiment is described in the above method embodiment, and will not be repeated here.

[0086] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0087] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0088] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0089] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identifying anomalies in logistics transportation, characterized in that, include: Collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record; Based on the aforementioned transportation monitoring records, sampling and quality inspection of the target cargo batches are conducted at the transit station to obtain sampling and quality inspection data. Based on the sampling quality inspection data, the overall quality of the target goods batch is analyzed to obtain the inferred quality information of the target goods batch. Based on the inferred quality information, the abnormal data of the transportation monitoring records is analyzed to obtain the abnormal identification information of the transportation vehicles between various transfer stations.

2. The anomaly identification method for logistics transportation as described in claim 1, characterized in that, The steps of collecting real-time transportation data of the transport vehicles for the target cargo batch and combining the real-time transportation data of the transport vehicles between various transfer stations along the logistics transportation route into a transportation monitoring record include: By using a sensor network pre-installed on the transport vehicle, the transport environment of the transport vehicle can be monitored in real time, and the real-time transport data of the transport vehicle can be obtained. Once the transport vehicle departs from a transit point on the logistics transport route, a corresponding timestamp is generated for the real-time transport data starting at that moment. The transportation data at each time point is arranged in chronological order based on the timestamp until the transport vehicle arrives at another transit station on the logistics route, so as to obtain the transportation monitoring record of the transport vehicle between the two transit stations.

3. The anomaly identification method for logistics transportation as described in claim 1, characterized in that, The steps for obtaining sampling inspection data by sampling and inspecting the target cargo batch at the transit station based on the transportation monitoring records include: Based on the transportation monitoring records, the characteristics of the transportation environment of the transport vehicle are analyzed to obtain the transportation environment characteristics of the transport vehicle between the two transfer stations; Based on the aforementioned transportation environment characteristics, a quality impact prediction for the target cargo batch is performed, resulting in a list of predicted impacts of the target cargo batch on the aforementioned transportation environment characteristics. Based on the aforementioned impact prediction list, a sampling plan is generated for the target cargo batch to conduct sampling quality inspection on the target cargo batch after it arrives at the transit station, thereby obtaining sampling quality inspection data.

4. The anomaly identification method for logistics transportation as described in claim 3, characterized in that, The steps of predicting the quality impact of the target cargo batch based on the transportation environment characteristics and obtaining a list of predicted impacts of the target cargo batch on the transportation environment characteristics include: Acquire cargo type data and placement data of the target cargo batch in the transport vehicle, and generate a digital model of the target cargo batch based on the cargo type data and placement data; Based on the characteristics of the transportation environment, the digital model is used to simulate the transportation impact in order to obtain the quality loss prediction information for each piece of cargo. Data statistics were collected on the quality loss prediction information of each item in the target batch of goods to obtain an impact prediction list.

5. The anomaly identification method for logistics transportation as described in claim 3, characterized in that, The steps for generating a sampling plan for the target cargo batch based on the aforementioned impact prediction list include: Based on the aforementioned impact prediction list, a sampling representativeness analysis is performed on each item within the target goods batch to obtain the sampling representative value of each item. Based on the data on the type and arrangement of the target goods in the transport vehicle, the sampling difficulty of each item is analyzed to obtain the sampling difficulty of each item. By combining the representative value of the sampling with the ease of sampling, the sampling target is selected for the target batch of goods in order to generate a sampling plan.

6. The method for identifying anomalies in logistics transportation as described in claim 1, characterized in that, The steps for analyzing the overall quality of the target goods batch based on the sampling quality inspection data to obtain the inferred quality information of the target goods batch include: The digital model of the target goods batch is invoked, and the information of the goods corresponding to the sampling inspection is marked on the digital model according to the sampling inspection data, so as to obtain the sampling inspection marking information of the digital model. Based on the sampling quality inspection labeling information, information inference processing is performed on the corresponding unsampled goods in the digital model, and the results of the information inference are labeled on the digital model to obtain the inferred quality information of the target goods batch.

7. The anomaly identification method for logistics transportation as described in claim 6, characterized in that, The steps of parsing abnormal data from the transportation monitoring records based on the inferred quality information to obtain anomaly identification information of the transportation vehicle between various transfer stations include: Based on the inferred quality information, statistical analysis of the quality loss data of the target goods batch is performed to obtain the quality loss data of the target goods batch. The data loss data is analyzed to identify data features, thereby obtaining the overall loss characteristics of the target batch of goods quality. Based on the overall loss characteristics, the causes of the loss are theoretically traced to obtain several expected causes of the loss. Based on each of the expected causes of loss, the transportation monitoring records are matched to generate verification data for each of the expected causes of loss. Comparative analysis is performed on the confirmatory data of each of the expected causes of loss to generate a confidence parameter for each of the expected causes of loss. Based on the credibility parameters, the expected causes of loss are eliminated, adjusted, and merged to obtain anomaly identification information of the transport vehicle between various transfer stations.

8. An anomaly identification system for logistics transportation, characterized in that, A method for identifying anomalies in logistics transportation as described in any one of claims 1-7 includes: The data acquisition module is used to collect real-time transportation data of the transport vehicles for the target batch of goods, and combine the real-time transportation data of the transport vehicles between various transfer stations on the logistics transportation route into a transportation monitoring record. The sampling and quality inspection module is used to conduct sampling and quality inspection of the target goods batch at the transit station based on the transportation monitoring records, so as to obtain sampling and quality inspection data. The quality estimation module is used to analyze the overall quality of the target goods batch based on the sampling quality inspection data, and obtain the estimated quality information of the target goods batch. An anomaly identification module is used to parse the abnormal data of the transportation monitoring records based on the inferred quality information to obtain anomaly identification information of the transportation vehicle between various transfer stations.

Images