Logistics data management method and system based on data analysis

Abstract

The application discloses a logistics data management method and system based on data analysis, and relates to the technical field of data management.The method comprises the following steps: collecting logistics whole-process data in real time, and performing space-time alignment processing to obtain a space-time synchronous data stream; constructing a multi-dimensional logistics feature matrix according to the space-time synchronous data stream, calculating a dynamic correlation coefficient, and assigning a real-time weight to an environment-related feature; weighting and fusing the multi-dimensional logistics feature matrix according to the real-time weight, inputting the weighted multi-dimensional logistics feature matrix obtained into a logistics analysis model, and outputting a logistics state prediction result; collecting actual state data in a logistics execution process, and calculating a deviation rate; and storing the deviation rate, the space-time synchronous data stream, the actual state data and the logistics state prediction result as logistics data to the cloud.The application solves the problems of the prior art, such as difficulty in fusing multi-source heterogeneous logistics data, static weight of environment features and lack of real-time deviation monitoring.

Description

Technical Field

[0001] This invention relates to the field of data management technology, and in particular to a logistics data management method and system based on data analysis. Background Technology

[0002] With the rapid development of the logistics industry, the scale of logistics data is growing explosively. Modern logistics management relies on the real-time collection, analysis, and prediction of massive amounts of data to improve transportation efficiency, reduce costs, and enhance customer satisfaction.

[0003] However, existing logistics data management technologies still have the following problems: 1) Data heterogeneity and spatiotemporal asynchrony: The logistics process involves diverse data sources, including business order systems (discrete event data), vehicle terminals (high-frequency trajectory data), and third-party weather map services (low-frequency or regional data). These data vary greatly in timestamp frequency and lack a unified spatiotemporal benchmark, making data fusion difficult and hindering the formation of high-quality input data for analytical models.

[0004] 2) Staticization of Environmental Factor Impact Assessment: In logistics timeliness forecasting, the external environment (such as weather and road conditions) is a significant influencing factor. Existing technologies typically use fixed weights or simple linear models to handle environmental characteristics. However, the impact of environmental factors on transportation efficiency is not static. For example, light rain has a small impact on highways but a significant impact on congested urban roads; fixed weights cannot capture this dynamic, context-dependent impact mechanism, resulting in poor robustness of the forecasting model in the face of sudden environmental changes.

[0005] 3) Lack of real-time deviation monitoring and feedback: Traditional logistics forecasting typically involves one-time route planning or time-delivery estimation before departure. However, during transportation execution, the actual situation often deviates from expectations. Existing technologies lack effective means to calculate the deviation rate between the "predicted state" and the "actual objective state" in real time during execution, making it impossible to promptly detect potential delay risks or effectively use deviation data for subsequent model correction or decision support.

[0006] Therefore, there is an urgent need for a logistics data management method that can achieve spatiotemporal alignment of multi-source data, dynamically assess environmental impact, and monitor prediction deviations in real time. Summary of the Invention

[0007] This invention provides a logistics data management method and system based on data analysis, which solves the problems of existing technologies, such as the difficulty in integrating multi-source heterogeneous logistics data, static environmental feature weights, and lack of real-time deviation monitoring.

[0008] In a first aspect, embodiments of the present invention provide a logistics data management method based on data analysis, the method comprising: Real-time collection of logistics data throughout the entire process, and spatiotemporal alignment processing of the logistics data to obtain a spatiotemporally synchronized data stream; Based on the spatiotemporally synchronized data stream, a multidimensional logistics feature matrix is ​​constructed, the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix is ​​calculated, and real-time weights are assigned to environmental correlation features. Based on real-time weights, the multidimensional logistics feature matrix is ​​weighted and fused, and the resulting weighted multidimensional logistics feature matrix is ​​input into the preset logistics analysis model to output the logistics status prediction results. Collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results; Deviation rate, spatiotemporally synchronized data stream, actual status data, and logistics status prediction results are stored as logistics data in the cloud.

[0009] The technical solution provided in this application has at least the following beneficial effects: By collecting real-time logistics data including business orders, vehicle trajectories, sensor data, and external environmental data, and performing spatiotemporal alignment processing based on the timeline of high-frequency vehicle data, the problem of time asynchrony and frequency inconsistency of multi-source heterogeneous data is effectively solved, constructing a high-dimensional spatiotemporally synchronized data stream, providing a high-quality data foundation for subsequent analysis. After constructing a multi-dimensional logistics feature matrix, the dynamic correlation coefficient between environmental features and historical transportation efficiency is creatively calculated, and real-time weights are assigned according to the correlation coefficient and the current value of the features. This allows for dynamic perception of the actual impact of different environmental factors (such as rainfall and congestion) on transportation efficiency, avoiding prediction bias caused by fixed weights in traditional methods, and significantly improving the adaptability and accuracy of the logistics status prediction model. A deviation rate calculation mechanism is introduced, which verifies the prediction results in real time using actual status data, and stores the deviation rate along with the original data and prediction results in the cloud, realizing closed-loop management of logistics data and providing a quantitative basis for subsequent model optimization and transportation strategy adjustment.

[0010] In one optional implementation, real-time data collection of the entire logistics process is performed, and the data is then spatiotemporally aligned to obtain a spatiotemporally synchronized data stream, including: Real-time collection of logistics process data, including business order data with discrete timestamps, vehicle trajectory and sensor data with high-frequency timestamps, and external environment data; Using the time axis of vehicle trajectory and sensor data as a reference, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method, thus constructing a spatiotemporally synchronized data stream.

[0011] In one optional implementation, real-time data collection is performed across the entire logistics process, including: By listening to the database change log of the order management system through message queue middleware, if the order status in the database change log changes, the corresponding key fields are extracted to obtain business order data with discrete timestamps. By receiving high-frequency data streams from vehicles through in-vehicle terminals and IoT gateways, vehicle trajectory and sensor data with high-frequency timestamps can be obtained. The system calls third-party map service APIs and weather service APIs to obtain external environmental data for the area corresponding to the vehicle's current coordinates. The external environmental data includes real-time traffic congestion index and weather data.

[0012] In one alternative implementation, using the time axis of vehicle trajectory and sensor data as a reference, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method, constructing a spatiotemporally synchronized data stream, including: The average sampling frequency of vehicle trajectory and sensor data is calculated, and combined with the set grid interval, the continuous time axis is discretized into a time series of several consecutive grid time points. The data points of each high-frequency timestamp in the vehicle trajectory and sensor data are mapped to the corresponding grid time points in the time series. If there are data points where the high-frequency timestamp and the grid time point do not match, the physical features of the grid time point are obtained by linear interpolation based on the adjacent data points of the two high-frequency timestamps before and after the grid time point that failed to match, and a gridded physical feature vector sequence is obtained. By using the nearest neighbor method, discrete events in business order data are mapped to the nearest grid time point, and a state preservation strategy is adopted to obtain a gridded business state vector sequence. Based on the gridded physical feature vector sequence, calculate its spatial location index in the environmental map grid, and based on the spatial location index, map the external environmental data to the corresponding grid time point in the time series to obtain the gridded environmental feature vector; The gridded physical feature vector sequence, the gridded business state vector sequence, and the gridded environmental feature vector are spliced ​​together according to the grid time points of the time series to construct a spatiotemporally synchronized data stream.

[0013] In one optional implementation, a multidimensional logistics feature matrix is ​​constructed based on the spatiotemporally synchronized data stream. The dynamic correlation coefficient between environmental features and historical transportation efficiency in the multidimensional logistics feature matrix is ​​calculated, and real-time weights are assigned to the environmental features, including: Based on the spatiotemporally synchronized data stream, a multidimensional logistics feature matrix is ​​constructed, which includes basic attribute features, real-time trajectory features, and environmental correlation features. Using basic attribute features as query keys, historical samples of similar transportation tasks are retrieved from the historical database to obtain a historical sample set. The historical samples include historical environmental features and corresponding historical transportation efficiency. For each dimension of the environmental correlation features, calculate its linear correlation with historical transportation efficiency in the historical sample set, and use it as the dynamic correlation coefficient; Based on the dynamic correlation coefficient, real-time weights are assigned to each dimension of the environmental correlation features.

[0014] In one optional implementation, the basic attribute features include current coordinates, destination coordinates, cargo weight, cargo volume, and promised delivery time level; The real-time trajectory features include instantaneous acceleration and velocity variance; The environmental characteristics include rainfall, traffic congestion index, and visibility.

[0015] In one optional implementation, the multidimensional logistics feature matrix is ​​weighted and fused according to real-time weights, and the resulting weighted multidimensional logistics feature matrix is ​​input into a preset logistics analysis model to output logistics status prediction results, including: While keeping the basic attribute features and real-time trajectory features unchanged, the environmental correlation features in the multidimensional logistics feature matrix are weighted and fused according to the real-time weight of each dimension to obtain the weighted multidimensional logistics feature matrix. Set the time window length, convert the weighted multidimensional logistics feature matrix into a time series input tensor, and input the time series input tensor into the preset logistics analysis model to obtain the remaining transportation time; Add the remaining transit time to the current time to get the estimated arrival timestamp, and compare the estimated arrival timestamp with the order's promised time to generate a status label; The remaining transit time, estimated arrival timestamp, and status label are combined to output the logistics status prediction result.

[0016] In one alternative implementation, the logistics analysis model is constructed based on the LSTM algorithm.

[0017] In one optional implementation, actual state data during the logistics execution process is collected, and the deviation rate between the actual state data and the predicted logistics state is calculated, including: The actual status data at the current moment during the logistics execution process is obtained through the vehicle terminal and IoT gateway. The actual status data includes the current real-time coordinates, the current instantaneous speed, and the current system time. Calculate the Euclidean distance based on the current real-time coordinates and the order destination coordinates, and use it as the actual remaining distance. Take the average of the current instantaneous velocity at several moments as the current effective average velocity; Calculate the virtual actual remaining transportation time based on the actual remaining distance and the current effective average speed; Calculate the virtual actual arrival time based on the remaining transportation time and current system time in the logistics status prediction results; The difference between the estimated arrival timestamp and the virtual actual arrival time is taken as the absolute time deviation; Divide the absolute time deviation by the virtual actual remaining transportation time to obtain the deviation rate between the actual status data and the logistics status prediction results.

[0018] Secondly, embodiments of the present invention provide a logistics data management system based on data analysis, used to implement a logistics data management method. The system includes: The multi-source data acquisition unit is used to collect data from the entire logistics process in real time and perform spatiotemporal alignment processing on the data to obtain a spatiotemporally synchronized data stream. The spatiotemporal alignment processing unit is used to construct a multidimensional logistics feature matrix based on the spatiotemporally synchronized data stream, calculate the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix, and assign real-time weights to the environmental correlation features. The feature weight calculation unit is used to perform weighted fusion of the multidimensional logistics feature matrix according to the real-time weight, and input the obtained weighted multidimensional logistics feature matrix into the preset logistics analysis model to output the logistics status prediction result. The prediction and deviation analysis unit is used to collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results. The cloud storage unit is used to store deviation rate, spatiotemporally synchronized data stream, actual status data, and logistics status prediction results as logistics data in the cloud.

[0019] A third aspect of this invention provides an electronic device, which includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by at least one processor, such that the at least one processor can perform the method proposed in the first aspect of the present invention.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in the first aspect of the present invention. Attached Figure Description

[0021] Figure 1This is a schematic diagram of the electronic device structure of the hardware operating environment involved in the embodiments of the present invention; Figure 2 This is a flowchart illustrating the steps of a logistics data management method based on data analysis provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the functional units of a logistics data management system based on data analysis provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0023] The present invention will be further described below with reference to the accompanying drawings.

[0024] Reference Figure 1 , Figure 1 This is a schematic diagram of the electronic device structure of the hardware operating environment involved in the embodiments of the present invention.

[0025] like Figure 1 As shown, the electronic device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0026] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0027] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and an electronic program for a logistics data management system based on data analysis.

[0028] exist Figure 1 In the illustrated electronic device, the network interface 1004 is mainly used for data analysis-based logistics data management with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the electronic device of the present invention can be installed in the electronic device, and the electronic device calls the electronic program of the data analysis-based logistics data management system stored in the memory 1005 through the processor 1001, and executes the data analysis-based logistics data management method provided in the embodiment of the present invention.

[0029] Reference Figure 2 The present invention provides a logistics data management method based on data analysis, the method comprising: S201: Real-time collection of logistics process data, and spatiotemporal alignment processing of the logistics process data to obtain a spatiotemporally synchronized data stream; S202: Based on the spatiotemporally synchronized data stream, construct a multidimensional logistics feature matrix, calculate the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix, and assign real-time weights to environmental correlation features; S203: Based on real-time weights, perform weighted fusion of the multidimensional logistics feature matrix, and input the resulting weighted multidimensional logistics feature matrix into the preset logistics analysis model to output the logistics status prediction result; S204: Collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results; S205: Store deviation rate, spatiotemporally synchronized data stream, actual status data, and logistics status prediction results as logistics data in the cloud.

[0030] The technical solution provided in this application has at least the following beneficial effects: By collecting real-time logistics data including business orders, vehicle trajectories, sensor data, and external environmental data, and performing spatiotemporal alignment processing based on the timeline of high-frequency vehicle data, the problem of time asynchrony and frequency inconsistency of multi-source heterogeneous data is effectively solved, constructing a high-dimensional spatiotemporally synchronized data stream, providing a high-quality data foundation for subsequent analysis. After constructing a multi-dimensional logistics feature matrix, the dynamic correlation coefficient between environmental features and historical transportation efficiency is creatively calculated, and real-time weights are assigned according to the correlation coefficient and the current value of the features. This allows for dynamic perception of the actual impact of different environmental factors (such as rainfall and congestion) on transportation efficiency, avoiding prediction bias caused by fixed weights in traditional methods, and significantly improving the adaptability and accuracy of the logistics status prediction model. A deviation rate calculation mechanism is introduced, which verifies the prediction results in real time using actual status data, and stores the deviation rate along with the original data and prediction results in the cloud, realizing closed-loop management of logistics data and providing a quantitative basis for subsequent model optimization and transportation strategy adjustment.

[0031] In one optional implementation, real-time data collection of the entire logistics process is performed, and the data is then spatiotemporally aligned to obtain a spatiotemporally synchronized data stream, including: S2011: Real-time collection of logistics process data, including business order data with discrete timestamps, vehicle trajectory and sensor data with high-frequency timestamps, and external environment data; S2012: Based on the time axis of vehicle trajectory and sensor data, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method to build a spatiotemporally synchronized data stream.

[0032] In one optional implementation, real-time data collection is performed across the entire logistics process, including: S20111: Listen to the database change log of the order management system through message queue middleware. If the order status in the database change log changes, extract the corresponding key fields to obtain business order data with discrete timestamps. In this embodiment, Change Data Capture (CDC) technology is employed. A message queue middleware (such as Kafka or RocketMQ) monitors the Binlog logs of the order management system's database (such as MySQL or Oracle) in real time. When an order status changes (e.g., from "accepted" to "in transit"), the system automatically captures the change event, extracts key fields (order ID, cargo weight and volume, origin and destination coordinates, promised delivery time, etc.), and adds discrete timestamps. This approach avoids putting pressure on the production system due to high-frequency queries to the business database. S20112: Receive high-frequency data streams from vehicles via in-vehicle terminals and IoT gateways to obtain vehicle trajectory and sensor data with high-frequency timestamps; In this embodiment, the vehicle terminal (T-Box) and vehicle sensors collect data at high frequencies (e.g., 1Hz-10Hz) and report the data via MQTT or TCP long connections through the IoT gateway. Specifically, this includes latitude and longitude coordinates, instantaneous speed, engine speed, fuel consumption, and cabin temperature. These data have time-series characteristics and serve as the basis for constructing the spatiotemporal grid. S20113: Call the third-party map service application programming interface (API) and meteorological service API to obtain the external environment data of the area corresponding to the vehicle's current coordinates. The external environment data includes real-time traffic congestion index and weather data. In this embodiment, the system does not directly store the full map data of the third party, but calls it on demand. Specifically, with the vehicle's current Global Positioning System (GPS) coordinates as the center point, it calls the map service API (such as Gaode or Baidu Maps) to obtain the traffic congestion index (smooth / slow / congested level) within a radius of 500 meters; and calls the meteorological API to obtain the real-time weather code (such as light rain, heavy rain), visibility value, etc. for the current location.

[0033] In one alternative implementation, using the time axis of vehicle trajectory and sensor data as a reference, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method, constructing a spatiotemporally synchronized data stream, including: S20121: Calculate the average sampling frequency of vehicle trajectory and sensor data, and in combination with the set grid interval, discretize the continuous time axis into a time series of several consecutive grid time points. In this embodiment, it is assumed that the average sampling frequency of vehicle trajectory and sensor data is 1Hz, and the grid interval is set to 1 second. The system constructs a unified time axis in seconds; S20122: Map the data points of each high-frequency timestamp in the vehicle trajectory and sensor data to the corresponding grid time points in the time series. If there are data points where the high-frequency timestamp and the grid time point do not match, then obtain the physical characteristics of the grid time point by linear interpolation based on the adjacent data points of the two high-frequency timestamps before and after the grid time point that failed to match, and obtain the gridded physical feature vector sequence. In this embodiment, for the vehicle's own high-frequency data, if there is packet loss or timestamp offset (e.g., no data in one second, but two data entries in the next second), a linear interpolation method is used, for example... t Velocity data is missing at time 1, butt The speed at time 0 is 60 km / h. t At time 2, the speed is 62 km / h, then t The interpolation speed at time 1 is 61 km / h to ensure that the physical feature vector sequence is continuous and smooth on the time axis; S20123: By using the nearest neighbor method, discrete events in business order data are mapped to the nearest grid time point, and a state preservation strategy is adopted to obtain a gridded business state vector sequence; In this embodiment, the business order status is a discrete event. For example, the order status changes to "loading completed" at 10:00:00 and then to "received" at 12:00:00. At all grid time points between 10:00:00 and 12:00:00, a state preservation strategy is adopted to fill in the "in transit" status and related order attributes to construct a complete business state vector sequence. S20124: Based on the gridded physical feature vector sequence, calculate its spatial location index in the environmental map grid, and based on the spatial location index, map the external environmental data to the corresponding grid time point in the time series to obtain the gridded environmental feature vector; In this embodiment, environmental data is usually based on a geographic grid (such as a 1km×1km grid). The system calculates the geographic grid index where the vehicle trajectory point is located, queries the environmental data corresponding to the grid, and assigns it to the corresponding time grid point. If the environmental data is updated slowly (for example, the weather is updated every 10 minutes), the previous environmental feature value is used within the update cycle to obtain a gridded environmental feature vector. S20125: Construct a spatiotemporally synchronized data stream by splicing together the gridded physical feature vector sequence, the gridded business state vector sequence, and the gridded environmental feature vector according to the grid time points of the time series.

[0034] In one optional implementation, a multi-dimensional logistics feature matrix is ​​constructed based on the spatiotemporally synchronized data stream. The dynamic correlation coefficient between environmental features and historical transportation efficiency in the multi-dimensional logistics feature matrix is ​​calculated, and real-time weights are assigned to the environmental features, including: S2021: Based on the spatiotemporally synchronized data stream, construct a multi-dimensional logistics feature matrix, which includes basic attribute features, real-time trajectory features, and environmental correlation features. S022: Using basic attribute features as query keys, retrieve historical samples of similar transportation tasks from the historical database to obtain a set of historical samples. The historical samples include historical environmental features and corresponding historical transportation efficiencies. S2023: For each dimension of the environmental correlation features, calculate its linear correlation with historical transportation efficiency in the historical sample set, as the dynamic correlation coefficient, using the following formula: In the formula, The first of the environmental association features j The dynamic correlation coefficient of the dimensions; j For dimension indexing; The total number of historical samples; k For historical sample indexing; For the first k The first historical sample j eigenvalues ​​of dimension; The first of the environmental association features j Historical mean of the dimension; For the first k Historical transportation efficiency of a historical sample; This represents the historical average of transportation efficiency. like ≈1: This indicates that the larger the value of the environmental factor, the longer the transportation time (positive correlation). For example, the "congestion index" usually fits this situation. like ≈-1: This indicates that the larger the value of the environmental factor, the shorter the transportation time (negative correlation). For example, "tailwind" or "good highway conditions at night" may fit this situation. like ≈0: This indicates that the environmental factor has no significant linear relationship with transportation efficiency and can be considered as noise in the prediction. S2024: Based on the dynamic correlation coefficient, assign real-time weights to each dimension of the environmental correlation features, using the following formula: In the formula, The first of the environmental association features j Real-time weights of dimensions; The first of the environmental association features j eigenvalues ​​of dimension; This is the sensitivity adjustment coefficient (which can be set to 1.5). The first of the environmental association features j The preset warning threshold for the dimension; Normal range ( At this point, the environment is in a normal state, and its influence follows historical patterns, so the correlation coefficient can be used directly. As a weight, this reflects the model's respect for historical experience.

[0035] Exceeding limits ( ): At this time, the environment is in an extreme state (such as torrential rain or severe traffic congestion). There may be few such samples in historical data, so we directly use... Its impact may have been underestimated, therefore an amplification factor is introduced. and the proportion exceeding the limit; Function: Sensitivity adjustment coefficient The larger the size, the more sensitive the system is to extreme environments, for example... =1.5 means that when the rainfall exceeds the threshold by 1 time, its weight will become 1 + 1.5 × 1 = 2.5 times that under normal circumstances.

[0036] In one optional implementation, the basic attribute features include current coordinates, destination coordinates, cargo weight, cargo volume, and promised delivery time level; The real-time trajectory features include instantaneous acceleration and velocity variance; The environmental characteristics include rainfall, traffic congestion index, and visibility.

[0037] In one optional implementation, the multidimensional logistics feature matrix is ​​weighted and fused according to real-time weights, and the resulting weighted multidimensional logistics feature matrix is ​​input into a preset logistics analysis model to output logistics status prediction results, including: S2031: Keeping the basic attribute features and real-time trajectory features unchanged, the environmental correlation features in the multidimensional logistics feature matrix are weighted and fused according to the real-time weight of each dimension to obtain the weighted multidimensional logistics feature matrix. S2032: Set the time window length, convert the weighted multidimensional logistics feature matrix into a time series input tensor, and input the time series input tensor into the preset logistics analysis model to obtain the remaining transportation time; S2033: Add the remaining transit time to the current time to obtain the estimated arrival timestamp, and compare the estimated arrival timestamp with the order commitment time to generate a status label; S2034: Combine the remaining transit time, estimated arrival timestamp, and status label to output the logistics status prediction result; In this embodiment, if the expected arrival timestamp is less than the promised time, a status label "normal" is generated; If the promised time is less than the estimated arrival timestamp, which is less than the promised time plus the buffer period, a status label "Delay Risk" will be generated. If the expected arrival timestamp is greater than the promised time plus the buffer period, generate the status label "Delayed". The final output is a structured prediction result, which is then used by the scheduling center for decision-making.

[0038] In one alternative implementation, the logistics analysis model is constructed based on the Long Short-Term Memory (LSTM) algorithm.

[0039] In one optional implementation, actual state data during the logistics execution process is collected, and the deviation rate between the actual state data and the predicted logistics state is calculated, including: S2041: Obtain the actual status data at the current moment during the logistics execution process through the vehicle terminal and IoT gateway. The actual status data includes the current real-time coordinates, the current instantaneous speed, and the current system time. S2042: Calculate the Euclidean distance based on the current real-time coordinates and the order destination coordinates, and use it as the actual remaining distance; S2043: Take the average of the current instantaneous velocity at several moments as the current effective average velocity; In this embodiment, noise reduction: directly using instantaneous speed (such as 0 when stopping at a red light) would result in an infinitely large remaining time. Therefore, the average speed over several past moments (such as the past 5 grid points) is taken. Effectiveness: If the average speed is too low (below the minimum threshold), the vehicle's historical average cruising speed is used instead to deal with scenarios where the vehicle is stationary for a long time (such as resting or loading / unloading) but the transportation has not ended. S2044: Calculate the virtual actual remaining transportation time based on the actual remaining distance and the current effective average speed; S2045: Calculate the virtual actual arrival time based on the remaining transportation time and current system time in the logistics status prediction results; S2046: The difference between the estimated arrival timestamp and the virtual actual arrival time is taken as the absolute time deviation; S2047: Divide the absolute time deviation by the virtual actual remaining transportation time to obtain the deviation rate between the actual status data and the logistics status prediction results; For example, if the remaining journey takes 1000 minutes and the deviation is 10 minutes, the deviation rate is only 1% (negligible); if the remaining journey takes 10 minutes and the deviation is 10 minutes, the deviation rate is 100% (severe). This indicator can intuitively reflect the relative accuracy of the prediction and is a key indicator for triggering early warnings and model corrections.

[0040] This invention also provides a logistics data management system based on data analysis, referring to... Figure 3 The device may include the following units: The multi-source data acquisition unit 301 is used to acquire real-time data of the entire logistics process and perform spatiotemporal alignment processing on the data of the entire logistics process to obtain a spatiotemporally synchronized data stream. The spatiotemporal alignment processing unit 302 is used to construct a multidimensional logistics feature matrix based on the spatiotemporally synchronized data stream, calculate the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix, and assign real-time weights to the environmental correlation features. The feature weight calculation unit 303 is used to perform weighted fusion of the multidimensional logistics feature matrix according to the real-time weight, and input the obtained weighted multidimensional logistics feature matrix into the preset logistics analysis model to output the logistics status prediction result. The prediction and deviation analysis unit 304 is used to collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results. The cloud storage unit 305 is used to store deviation rate, spatiotemporally synchronized data stream, actual status data and logistics status prediction results as logistics data in the cloud.

[0041] Based on the same inventive concept, another embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. Memory, used to store computer programs; The processor, when executing a program stored in memory, implements the data analysis-based logistics data management method of the present invention.

[0042] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of representation, only one thick line is used in the diagram, but this does not indicate that there is only one bus or one type of bus. The communication interface is used for communication between the aforementioned terminal and other devices. The memory can include Random Access Memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory can also be at least one storage device located remotely from the aforementioned processor.

[0043] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0044] Furthermore, to achieve the above objectives, embodiments of the present invention also propose a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the data analysis-based logistics data management method of the present invention.

[0045] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable hardware devices (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0046] The embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (apparatus), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0047] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0048] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0049] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. "And / or" indicates that either one or both can be chosen. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes the element.

[0050] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A logistics data management method based on data analysis, characterized in that, The method includes: Real-time collection of logistics data throughout the entire process, and spatiotemporal alignment processing of the logistics data to obtain a spatiotemporally synchronized data stream; Based on the spatiotemporally synchronized data stream, a multidimensional logistics feature matrix is ​​constructed, the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix is ​​calculated, and real-time weights are assigned to environmental correlation features. Based on real-time weights, the multidimensional logistics feature matrix is ​​weighted and fused, and the resulting weighted multidimensional logistics feature matrix is ​​input into the preset logistics analysis model to output the logistics status prediction results. Collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results; Deviation rate, spatiotemporally synchronized data stream, actual status data, and logistics status prediction results are stored as logistics data in the cloud.

2. The logistics data management method based on data analysis according to claim 1, characterized in that, Real-time acquisition of logistics data across the entire process, followed by spatiotemporal alignment processing to obtain a spatiotemporally synchronized data stream, including: Real-time collection of logistics process data, including business order data with discrete timestamps, vehicle trajectory and sensor data with high-frequency timestamps, and external environment data; Using the time axis of vehicle trajectory and sensor data as a reference, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method, thus constructing a spatiotemporally synchronized data stream.

3. The logistics data management method based on data analysis according to claim 2, characterized in that, Real-time collection of data throughout the entire logistics process, including: By listening to the database change log of the order management system through message queue middleware, if the order status in the database change log changes, the corresponding key fields are extracted to obtain business order data with discrete timestamps. By receiving high-frequency data streams from vehicles through in-vehicle terminals and IoT gateways, vehicle trajectory and sensor data with high-frequency timestamps can be obtained. The system calls third-party map service APIs and weather service APIs to obtain external environmental data for the area corresponding to the vehicle's current coordinates. The external environmental data includes real-time traffic congestion index and weather data.

4. The logistics data management method based on data analysis according to claim 3, characterized in that, Using the timeline of vehicle trajectory and sensor data as a reference, business order data and external environment data are mapped onto a unified time grid through linear interpolation or the nearest neighbor method, constructing a spatiotemporally synchronized data stream, including: The average sampling frequency of vehicle trajectory and sensor data is calculated, and combined with the set grid interval, the continuous time axis is discretized into a time series of several consecutive grid time points. The data points of each high-frequency timestamp in the vehicle trajectory and sensor data are mapped to the corresponding grid time points in the time series. If there are data points where the high-frequency timestamp and the grid time point do not match, the physical features of the grid time point are obtained by linear interpolation based on the adjacent data points of the two high-frequency timestamps before and after the grid time point that failed to match, and a gridded physical feature vector sequence is obtained. By using the nearest neighbor method, discrete events in business order data are mapped to the nearest grid time point, and a state preservation strategy is adopted to obtain a gridded business state vector sequence. Based on the gridded physical feature vector sequence, calculate its spatial location index in the environmental map grid, and based on the spatial location index, map the external environmental data to the corresponding grid time point in the time series to obtain the gridded environmental feature vector; The gridded physical feature vector sequence, the gridded business state vector sequence, and the gridded environmental feature vector are spliced ​​together according to the grid time points of the time series to construct a spatiotemporally synchronized data stream.

5. The logistics data management method based on data analysis according to claim 4, characterized in that, Based on the spatiotemporally synchronized data stream, a multidimensional logistics feature matrix is ​​constructed. The dynamic correlation coefficients between environmental features and historical transportation efficiency in the multidimensional logistics feature matrix are calculated, and real-time weights are assigned to the environmental features, including: Based on the spatiotemporally synchronized data stream, a multidimensional logistics feature matrix is ​​constructed, which includes basic attribute features, real-time trajectory features, and environmental correlation features. Using basic attribute features as query keys, historical samples of similar transportation tasks are retrieved from the historical database to obtain a historical sample set. The historical samples include historical environmental features and corresponding historical transportation efficiency. For each dimension of the environmental correlation features, calculate its linear correlation with historical transportation efficiency in the historical sample set, and use it as the dynamic correlation coefficient; Based on the dynamic correlation coefficient, real-time weights are assigned to each dimension of the environmental correlation features.

6. The logistics data management method based on data analysis according to claim 5, characterized in that, The basic attribute features include current coordinates, destination coordinates, cargo weight, cargo volume, and promised delivery time level; The real-time trajectory features include instantaneous acceleration and velocity variance; The environmental characteristics include rainfall, traffic congestion index, and visibility.

7. The logistics data management method based on data analysis according to claim 6, characterized in that, Based on real-time weights, the multidimensional logistics feature matrix is ​​weighted and fused, and the resulting weighted multidimensional logistics feature matrix is ​​input into a preset logistics analysis model to output logistics status prediction results, including: While keeping the basic attribute features and real-time trajectory features unchanged, the environmental correlation features in the multidimensional logistics feature matrix are weighted and fused according to the real-time weight of each dimension to obtain the weighted multidimensional logistics feature matrix. Set the time window length, convert the weighted multidimensional logistics feature matrix into a time series input tensor, and input the time series input tensor into the preset logistics analysis model to obtain the remaining transportation time; Add the remaining transit time to the current time to get the estimated arrival timestamp, and compare the estimated arrival timestamp with the order's promised time to generate a status label; The remaining transit time, estimated arrival timestamp, and status label are combined to output the logistics status prediction result.

8. The logistics data management method based on data analysis according to claim 7, characterized in that, The logistics analysis model is built based on the LSTM algorithm.

9. The logistics data management method based on data analysis according to claim 8, characterized in that, Collect actual status data during the logistics execution process, and calculate the deviation rate between the actual status data and the predicted logistics status, including: The actual status data at the current moment during the logistics execution process is obtained through the vehicle terminal and IoT gateway. The actual status data includes the current real-time coordinates, the current instantaneous speed, and the current system time. Calculate the Euclidean distance based on the current real-time coordinates and the order destination coordinates, and use it as the actual remaining distance. Take the average of the current instantaneous velocity at several moments as the current effective average velocity; Calculate the virtual actual remaining transportation time based on the actual remaining distance and the current effective average speed; Calculate the virtual actual arrival time based on the remaining transportation time and current system time in the logistics status prediction results; The difference between the estimated arrival timestamp and the virtual actual arrival time is taken as the absolute time deviation; Divide the absolute time deviation by the virtual actual remaining transportation time to obtain the deviation rate between the actual status data and the logistics status prediction results.

10. A logistics data management system based on data analysis, used to implement the logistics data management method as described in any one of claims 1-9, characterized in that, The system includes: The multi-source data acquisition unit is used to collect data from the entire logistics process in real time and perform spatiotemporal alignment processing on the data to obtain a spatiotemporally synchronized data stream. The spatiotemporal alignment processing unit is used to construct a multidimensional logistics feature matrix based on the spatiotemporally synchronized data stream, calculate the dynamic correlation coefficient between environmental correlation features and historical transportation efficiency in the multidimensional logistics feature matrix, and assign real-time weights to the environmental correlation features. The feature weight calculation unit is used to perform weighted fusion of the multidimensional logistics feature matrix according to the real-time weight, and input the obtained weighted multidimensional logistics feature matrix into the preset logistics analysis model to output the logistics status prediction result. The prediction and deviation analysis unit is used to collect actual status data during the logistics execution process and calculate the deviation rate between the actual status data and the logistics status prediction results. The cloud storage unit is used to store deviation rate, spatiotemporally synchronized data stream, actual status data, and logistics status prediction results as logistics data in the cloud.

Images