Quality intelligent control methods, systems, and equipment based on big data analysis of logistics networks

Abstract

This invention discloses a quality intelligent control method, system, and equipment based on big data analysis of logistics networks. The method includes: aggregating and integrating initial data collected by data acquisition equipment to obtain integrated data and preprocessing it; performing correlation analysis on the preprocessed data in conjunction with transportation orders to obtain correlation feature analysis information and configuring a corresponding correlation analysis model; performing trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data to obtain correlation feature trend analysis information and performing quality trend analysis to obtain quality early warning analysis information; and optimizing the correlation feature analysis information according to quality optimization strategies to obtain quality optimization suggestions. The above-mentioned quality intelligent control method can integrate and aggregate initial data from multiple environments and perform correlation analysis and quality trend analysis in conjunction with transportation orders to obtain quality early warning analysis information and quality optimization suggestions, significantly improving the efficiency of quality control in the logistics transportation process.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, and in particular to a quality intelligent control method, system, and equipment based on big data analytics of logistics networks. Background Technology

[0002] With the rapid development of China's logistics and transportation industry and the continuous expansion of the logistics market, the industry faces the practical need for both speed and quality assurance. To ensure the quality of goods during transportation, quality control analysis is necessary, along with corresponding optimization of transportation strategies. However, existing logistics companies typically rely on manual inspection and analysis for quality control, which is not only time-consuming and labor-intensive but also susceptible to human factors, making it difficult to guarantee the accuracy of quality inspection and optimization. Furthermore, logistics data often involves multiple stages, including transportation, warehousing, loading and unloading, and distribution. Data from different stages is stored separately, creating "data silos." Due to the lack of effective data integration and aggregation mechanisms, logistics companies struggle to fully grasp the quality status of goods during transportation and flow, thus failing to promptly identify and resolve quality issues. Therefore, existing methods for logistics quality control suffer from low efficiency. Summary of the Invention

[0003] This invention provides a quality intelligent control method, system, and equipment based on big data analysis of logistics networks, aiming to solve the problem of low quality control efficiency in existing methods for logistics quality control.

[0004] In a first aspect, embodiments of the present invention provide a quality intelligent control method based on big data analysis of a logistics network, wherein the method is applied to a quality intelligent control platform, the quality intelligent control platform being communicatively connected with data acquisition devices set up in the logistics network to realize data information transmission, and the method includes:

[0005] The initial data collected by the data acquisition device is obtained, and the initial data is aggregated and integrated according to the pre-stored transportation orders to obtain the corresponding integrated data;

[0006] The integrated data is preprocessed according to preset preprocessing rules to obtain corresponding preprocessed data;

[0007] Based on the preset feature extraction model and the transportation order, the key feature information is correlated and analyzed to obtain the corresponding correlation feature analysis information;

[0008] Based on the association feature analysis information, an association analysis model corresponding to each transportation type is configured;

[0009] Trend analysis is performed on the correlation feature values ​​corresponding to each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information.

[0010] Based on the correlation analysis model, quality trend analysis is performed on the correlation feature trend analysis information to obtain quality early warning analysis information;

[0011] The associated feature analysis information is optimized and analyzed according to the preset quality optimization strategy to obtain quality optimization suggestions for each transportation type.

[0012] Secondly, embodiments of the present invention also provide a quality intelligent control system based on big data analysis of logistics networks. The system is configured on a quality intelligent control platform, which is communicatively connected to data acquisition devices installed in the logistics network to transmit data information. The system is used to execute the quality intelligent control method based on big data analysis of logistics networks as described in the first aspect above. The system includes the following units:

[0013] An integrated data acquisition unit is used to acquire the initial data collected by the data acquisition device, and to collect and integrate the initial data according to the pre-stored transportation orders to obtain the corresponding integrated data.

[0014] A preprocessing data acquisition unit is used to preprocess the integrated data according to preset preprocessing rules to obtain corresponding preprocessed data;

[0015] The correlation analysis unit is used to perform correlation analysis on key feature information based on the preset feature extraction model and the transportation order to obtain the corresponding correlation feature analysis information;

[0016] The model configuration unit is used to configure the correlation analysis model corresponding to each transportation type based on the correlation feature analysis information.

[0017] The correlation feature trend analysis unit is used to perform trend analysis on the correlation feature values ​​of each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information.

[0018] The quality trend analysis information acquisition unit is used to perform quality trend analysis on the correlation feature trend analysis information according to the correlation analysis model to obtain quality early warning analysis information.

[0019] The quality optimization suggestion acquisition unit is used to perform optimization analysis on the associated feature analysis information according to the preset quality optimization strategy in order to obtain quality optimization suggestions corresponding to each transportation type.

[0020] Thirdly, embodiments of the present invention also provide a computer device, wherein the device includes 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;

[0021] Memory, used to store computer programs;

[0022] When the processor executes the program stored in the memory, it implements the steps of the quality intelligent control method based on big data analysis of logistics networks described in the first aspect above.

[0023] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the quality intelligent control method based on big data analysis of logistics networks as described in the first aspect above.

[0024] This invention provides a quality intelligent control method, system, and device based on big data analysis of logistics networks. The method includes: aggregating and integrating initial data collected by data acquisition equipment to obtain integrated data and preprocessing it; performing correlation analysis on the preprocessed data in conjunction with transportation orders to obtain correlation feature analysis information and configuring a corresponding correlation analysis model; performing trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data to obtain correlation feature trend analysis information and performing quality trend analysis to obtain quality early warning analysis information; and performing optimization analysis on the correlation feature analysis information according to quality optimization strategies to obtain quality optimization suggestions. This quality intelligent control method can integrate and collect initial data from multiple stages of cargo transportation and perform correlation analysis and quality trend analysis in conjunction with transportation orders, thereby obtaining quality early warning analysis information reflecting transportation quality trends and quality optimization suggestions for optimizing the transportation process for different transportation types, significantly improving the efficiency of quality control in the logistics transportation process. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating the quality intelligent control method based on big data analysis of logistics networks provided in this embodiment of the invention;

[0027] Figure 2 This is a schematic diagram illustrating an application scenario of the quality intelligent control method based on big data analysis of logistics networks provided in this embodiment of the invention.

[0028] Figure 3 A schematic block diagram of a quality intelligent control system based on big data analysis of logistics networks provided in an embodiment of the present invention;

[0029] Figure 4 This is a schematic block diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0032] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0033] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0034] This invention application provides a quality intelligent control method based on big data analysis of logistics networks. This method is applied to a quality intelligent control platform, which executes stored software programs to implement the aforementioned quality intelligent control method based on big data analysis of logistics networks. The quality intelligent control platform 10 is an intelligent management and control platform configured within an enterprise for data analysis and logistics quality control, such as a management server or cluster server. Figure 2As shown, the quality intelligent control platform 10 communicates with data acquisition devices set up in the logistics network to realize the transmission of data information. The data acquisition devices can be vehicle-mounted data acquisition devices 20, logistics site data acquisition devices 21, warehouse data acquisition devices 22, and cargo data acquisition devices 23. They collect data on the cargo itself, as well as during its transportation in the vehicle, during unloading / loading in the logistics site, and during storage in the logistics site / warehouse, and realize quality control. The data acquisition devices can be temperature and humidity sensors, air pressure sensors, light sensors, vibration sensors, positioning sensors, RFID tags, cameras, etc. For example, temperature and humidity sensors can collect temperature and humidity data, air pressure sensors can collect air pressure data of the environment where the cargo is located, light sensors are used to collect light intensity data, vibration sensors can collect vibration data, positioning sensors can collect satellite positioning data (such as Beidou satellite positioning data), RFID tags (active RFID tags) can be scanned to obtain the tag data of the cargo, and cameras can collect images / videos of the appearance of the cargo and unloading / loading operations. Among them, RFID tags are fixedly installed on the goods; vibration sensors, positioning sensors and barometric pressure sensors are all fixedly installed on the vehicles; temperature and humidity sensors are fixedly installed on the vehicles and inside the warehouse; and cameras are fixedly installed in the vehicles, warehouses and logistics sites.

[0035] like Figure 1 As shown, the method includes steps S110 to S170.

[0036] S110. Obtain the initial data collected by the data acquisition device, and collect and integrate the initial data according to the pre-stored transportation orders to obtain the corresponding integrated data.

[0037] Initial data can be collected through data acquisition devices. This initial data consists of data collected periodically by the devices. For example, if the collection period is set to 5 minutes, data will be collected every 5 minutes, yielding a set of data. Before transportation, a transportation order is generated and stored. Each transported item corresponds to a transportation order. Through these orders, the data collected from logistics nodes can be aggregated and integrated. The logistics network includes numerous logistics nodes (such as transport vehicles). Data collected from different logistics nodes for the same transported item occurs in stages. Therefore, real-time data collection is necessary. Data for the same transported item will not exist simultaneously during transportation, loading / unloading, storage, and delivery stages, but it can be aggregated and integrated by acquiring data collected from the same item at different time periods. Transportation companies receive a huge number of transportation orders daily, resulting in a massive amount of transported item data. This enables intelligent management and control of logistics quality based on big data analysis. If a transportation order includes one or more transported items, and each item uniquely corresponds to a tag, initial data matching the tag information of the transported item in the order can be obtained, thus achieving data aggregation and integration. By matching the transported goods with the corresponding transport orders using the tag information, the data of the transport orders and transported goods can be integrated to obtain integrated data. The data collected by temperature and humidity sensors, light sensors, barometric pressure sensors, vibration sensors, positioning sensors, and RFID tags are all numerical data; the transport order records the type of goods, quantity of goods, packaging type, shelf life, etc., the vibration sensor can collect vibration acceleration, vibration frequency, tilt angle, speed, etc., and the positioning sensor can obtain real-time location, transport trajectory, dwell time, etc. The appearance images / videos captured by the camera can be used to extract numerical information, such as the outer contour coverage area of ​​the goods packaging and the RGB color values ​​of the goods packaging, as data information captured by the camera. By calculating whether the outer contour coverage area of ​​the goods packaging changes, it can be determined whether the goods packaging is damaged or deformed (if the change rate of the outer contour coverage area during transportation is less than 2%, it is determined that the goods packaging is not damaged or deformed). By performing uniformity analysis on the RGB color values ​​of the goods packaging, it can be determined whether the goods packaging is stained or intact (if the color uniformity is higher than 97%, it is determined that the goods packaging is not stained and the packaging is intact).

[0038] The rate of change of the area covered by the outer contour can be calculated using the following formula:

[0039] (1);

[0040] In the above formula, μ is the rate of change, and Smax S represents the maximum area covered by the outer contour. min S represents the minimum area covered by the outer contour. a This represents the average area covered by the outer contour.

[0041] Color uniformity can be calculated using the following formula:

[0042] (2);

[0043] In the above formula, ε represents the calculated color uniformity, N represents the number of pixels in the outline image corresponding to the outer contour of the goods packaging, and P... i Let P be the pixel value of the i-th pixel in the contour image. a This is the pixel average of all pixels in the contour image.

[0044] S120. The integrated data is preprocessed according to the preset preprocessing rules to obtain the corresponding preprocessed data.

[0045] The integrated data can be preprocessed using preprocessing rules to standardize the data values ​​and obtain the corresponding preprocessed data.

[0046] In a specific embodiment, step S120 includes the following sub-steps: supplementing and updating the missing values ​​in the initial data according to the missing value supplementation function configured in the preprocessing rules to obtain updated data corresponding to the initial data; performing outlier detection on the updated data according to the outlier range configured in the preprocessing rules to exclude outlier data in the updated data and obtain corresponding valid data; and performing format conversion on the valid data according to the standard format configured in the preprocessing rules to obtain corresponding preprocessed data.

[0047] Specifically, the preprocessing rules include a missing value imputation function, which can be a function constructed based on the Lagrange Mean Value Theorem. This function can be used to fill in and update missing values ​​in the initial data. If a value at a certain acquisition time point in the initial data is empty or zero, then that value is determined to be missing, and its location is the missing position. The values ​​at the previous and next acquisition time points of the missing position can be obtained, and the corresponding imputation values ​​are calculated based on the missing value imputation function. These imputation values ​​are then replaced at the missing position, thus filling in and updating the missing values ​​in the initial data and obtaining updated data corresponding to the initial data.

[0048] Anomaly detection is performed on the updated data according to the outlier range configured in the preprocessing rules to exclude abnormal data and obtain valid data corresponding to the updated data. Specifically, the collection interval time (the difference between the collection time of the previous data set and the collection time of the current data set) of two adjacent sets of data for each transported goods in the updated data can be obtained, and it can be determined whether the collection interval time is greater than the interval time threshold configured in the preprocessing rules (e.g., the interval time threshold can be set to 4 minutes). If the collection interval time is greater than the interval time threshold, it indicates that there are no anomalies in the collection process of the two sets of data and they can be retained. If the collection interval time is not greater than the interval time threshold, it indicates that there are anomalies in the collection process of the two sets of data, such as network communication failure causing data collection delay or sensor failure causing short-term continuous data collection. The latest set of data is retained, and the other set of data is excluded.

[0049] The preprocessing rules also include outlier ranges corresponding to each value. This allows the system to determine whether a value in the updated data falls within its corresponding outlier range. If a value is within this range, the data value of the transported goods corresponding to that value at the same collection time point is excluded as outlier data. Furthermore, to improve data collection accuracy, after excluding outlier data for the transported goods corresponding to a value within an outlier range at the same collection time point, the system further checks whether the preceding or following collection time point contains supplementary values. If so, the data value of those supplementary values ​​at the same collection time point is also excluded as outlier data. After excluding outlier data, the valid data corresponding to the updated data is obtained.

[0050] The obtained valid data is format-converted according to the standard format configured in the preprocessing rules to obtain the corresponding preprocessed data. The preprocessing rules also configure standard formats corresponding to each value. These standard formats can be used to convert the format of each value in the valid data, excluding RFID tags, such as converting the weight of goods measured in "tons" to the weight of goods measured in "kilograms". The converted values ​​are then normalized or divided into different containers to obtain the converted values ​​corresponding to each value. These converted values ​​are then converted to values ​​ranging from [0,1]. The resulting converted values ​​are then combined to form the preprocessed data. Normalization involves converting numerical values ​​into corresponding normalized values ​​using a normalization function. The range of normalized values ​​is [0,1]. Bucketing involves classifying numerical values ​​and assigning a defined value to each category. For example, if transportation routes A, B, C, and D are sorted by transportation time, transportation route B, which has the shortest transportation time, is assigned a defined value of "0.1", transportation route C, which has the second shortest transportation time, is assigned a defined value of "0.3", and so on.

[0051] S130. Perform correlation analysis on the preprocessed data based on the preset feature extraction model and the transportation order to obtain the corresponding correlation feature analysis information.

[0052] Furthermore, correlation analysis is performed on the preprocessed data based on the pre-set feature extraction model and transportation orders. Each transportation order includes a completion flag, the specific value of which records whether the transportation order has been completed. Therefore, completed orders with the completion flag "Completed" can be filtered out from the completed orders. Completed orders contain loss frequencies. Correlation analysis is performed based on the loss frequencies of completed orders and the corresponding preprocessed data to obtain correlation feature analysis information. The loss frequency is the frequency information obtained by statistically analyzing the losses of the same batch of transported goods after completion. For example, if the total loss frequency of a transportation order is 6, and the order includes 100 transported goods, then the corresponding loss frequency is 6 / 100 = 0.06.

[0053] In a specific embodiment, step S130 includes the following sub-steps: extracting corresponding key feature information from the preprocessed data according to a preset feature extraction model; performing correlation analysis on the key feature information according to the transportation order to obtain the correlation features corresponding to each transportation type as the corresponding correlation feature analysis information.

[0054] First, completed orders are screened and retrieved. Then, key feature information is extracted from the preprocessed data corresponding to these orders using a feature extraction model. Completed orders are categorized based on their transportation type, which can be based on the type of goods and the transportation distance. For example, if the goods are "fresh produce" and the transportation distance is "short-distance transportation" (within 500km), a transportation type of "fresh produce + short-distance transportation" can be formed. Similarly, if the goods are "express parcels" and the transportation distance is "long-distance transportation" (over 1500km), a transportation type of "express parcels + long-distance transportation" can be formed. Through classification, the loss frequency of completed orders corresponding to each transportation type is obtained and correlated with the corresponding key feature information to obtain the correlation features for each transportation type. Correlation features are those strongly correlated with the transportation type. These correlation features are used as the corresponding correlation feature analysis information. For example, for a certain transportation type, correlation analysis shows a strong correlation between transportation route A+, high transportation temperature, and high loss frequency. Therefore, the key features corresponding to this transportation type include the transportation route and transportation temperature.

[0055] In a specific embodiment, the step of extracting the corresponding key feature information from the preprocessed data according to the preset feature extraction model includes: obtaining the feature value corresponding to each feature item in the feature extraction model; classifying the feature value corresponding to each feature item according to the transportation order to obtain the classification feature information corresponding to each transportation type; and statistically analyzing the classification feature information of each transportation type according to the statistical rules of the feature extraction model to obtain the corresponding feature statistical value as key feature information.

[0056] The feature extraction model includes multiple feature terms, with the number of feature terms being less than the number of data terms in the preprocessed data. It can obtain the feature values ​​corresponding to each feature term from the preprocessed data corresponding to completed orders. Furthermore, the feature values ​​corresponding to each feature term are classified according to the transportation order to obtain the feature values ​​corresponding to each transportation type and each feature term, serving as the classification feature information for each transportation type.

[0057] Furthermore, the classification feature information of each transportation type is statistically analyzed according to the statistical rules in the feature extraction model to obtain the corresponding feature statistics as the key feature information. Specifically, the statistical rules include multiple statistical intervals, such as the low value interval [0, 0.2], the medium value interval (0.20, 0.40], the medium-high value interval (0.40, 0.60], the high value interval (0.60, 0.80], and the extremely high value interval (0.80, 1.0). Based on the statistical rules, the proportion of each feature value in the classification feature information of each transportation type within each statistical interval is statistically analyzed to obtain the feature statistics of each feature item as the corresponding key feature information.

[0058] In a specific embodiment, the step of performing correlation analysis on the key feature information based on the transportation order to obtain the correlation features corresponding to each transportation type as the corresponding correlation feature analysis information includes: constructing variable pairs corresponding to each transportation type based on the transportation order and key feature information; performing correlation analysis on the variable pairs of each transportation type to obtain the correlation degree of each variable pair; and filtering the variable pairs contained in each transportation type based on the correlation degree of the variable pairs of each transportation type to obtain the corresponding correlation variable pairs as the correlation features of each transportation type.

[0059] Based on transportation orders and key feature information, variable pairs corresponding to each transportation type are constructed. Specifically, based on the feature statistics of a certain feature item in the key feature information of each transportation type, the loss frequency of completed orders for that transportation type in the transportation orders is binned, thus constructing variable pairs. For example, for the feature statistics of the transportation temperature feature item, the low-value range can be obtained, and the completed orders corresponding to the low-value range can be calculated. For instance, if the feature value of the transportation temperature feature item belongs to the low-value range, the corresponding completed orders account for 18% of all completed orders. The loss frequency of the completed orders corresponding to the low-value range is binned to obtain the corresponding high loss ratio. The total number of the top 25% of completed orders with the highest loss frequency among all completed orders can be counted as the total number of high-frequency loss orders. Then, the number of the top 25% of completed orders with the highest loss frequency corresponding to the low-value range can be obtained as the number of high-frequency loss orders corresponding to the low-value range. Dividing the number of high-frequency loss orders corresponding to the low-value range by the total number of high-frequency loss orders yields the corresponding high loss ratio. Based on the above method, the proportion of high loss corresponding to each statistical interval can be obtained sequentially. By combining the proportion of orders for each statistical interval with the proportion of high loss corresponding to each statistical interval, a set of variable pairs corresponding to the feature can be constructed. Based on this technical method, multiple sets of variable pairs corresponding to each transportation type can be constructed respectively, so multiple sets of variable pairs can be constructed for each transportation type.

[0060] Correlation analysis is performed on the variable pairs for each transportation type to obtain the correlation degree corresponding to each variable pair for that transportation type. Correlation analysis can be achieved using the Spearman rank correlation coefficient. Specifically, the values ​​of the two variables in a variable pair can be sorted "from low to high." For example, for the characteristic item of transportation temperature, the order percentage (set as X value) and the high loss percentage (set as Y value) can be sorted separately. If there are five statistical intervals, five ranking levels corresponding to the X value and five ranking levels corresponding to the Y value can be obtained, with larger values ​​corresponding to higher ranking levels.

[0061] For example, the order percentages in the low value range [0, 0.2], medium value range (0.20, 0.40], medium-high value range (0.40, 0.60], high value range (0.60, 0.80], and extremely high value range (0.80, 1.0] are 18%, 22%, 25%, 15%, and 20%, respectively. The corresponding high loss percentages for these five statistical ranges are 12%, 16%, 21%, 21%, and 30%, respectively. If the percentages for two statistical ranges are equal, the levels are averaged. The sorting results are shown in Table 1.

[0062] Table 1

[0063]

[0064] Further calculation yields the correlation degree ρ, where:

[0065] (3);

[0066] Where, d 2 The summation result is 17.5, where n is the total number of statistical intervals. According to Table 1, n=5, so the final calculated correlation coefficient is 0.125. The correlation coefficient ρ ranges from -1 to 1; a correlation coefficient close to "-1" indicates a high negative correlation, while a correlation coefficient close to "+1" indicates a high positive correlation.

[0067] Based on the above method, the correlation degree of each variable pair in each transportation type can be obtained. Then, the variable pairs contained in each transportation type can be screened according to the correlation degree of the variable pairs, so as to obtain the corresponding related variables as the correlation features of each transportation type.

[0068] Specifically, a filtering interval can be set, and variable pairs within that interval can be selected as correlation variables. For example, the filtering interval can be set to [-1, -0.7] & [0.7, 1]. If the correlation degree of a variable pair is within this interval, the corresponding feature item can be obtained and retained; if the correlation degree is not within this interval, it can be filtered out. The feature items obtained from the variable pairs of each transportation type after filtering can be used as correlation variables. One or more correlation variables corresponding to each transportation type constitute the correlation features of the transportation type.

[0069] S140. Based on the association feature analysis information, configure the association analysis model corresponding to each transportation type.

[0070] Furthermore, based on the correlation feature analysis information, the correlation analysis model corresponding to each transportation type is configured. Since the correlation features of each transportation type are different, a correlation analysis model can be configured and generated for each transportation type.

[0071] In a specific embodiment, step S140 includes the following sub-steps: configuring input nodes for the initial analysis model based on the correlation features of each transportation type in the correlation feature analysis information; configuring weight layers for the corresponding initial analysis model with output nodes based on the correlation degree of the correlation variable pairs in the correlation features of each transportation type, thereby obtaining the correlation analysis model corresponding to each transportation type.

[0072] Specifically, the input nodes of the initial analysis model can be configured based on the correlation features of each transportation type in the correlation feature analysis information. For example, if the correlation features of a certain transportation type contain three feature items, then three input nodes can be configured for the initial analysis model accordingly. Each input node inputs the value corresponding to one feature item in the correlation features. The initial analysis model is configured with multiple intermediate layers, each containing multiple intermediate nodes. Each intermediate node in the first intermediate layer establishes a correlation relationship with the configured input nodes. The initial analysis model also includes an output node, which is used to analyze and output a corresponding loss frequency prediction value. Each intermediate node in the last intermediate layer establishes a correlation relationship with the output node.

[0073] Furthermore, based on the correlation degree of each pair of related variables contained in the correlation features of each transportation type, a weight layer is configured for the initial analysis model with configured input nodes. The weight layer can be configured between the input nodes and the first intermediate layer. The weight layer is used to quantify the importance of the input values ​​at each input node. For example, the correlation degree of each pair of related variables contained in the correlation features can be obtained, and the absolute value of the correlation degree can be used as the weight coefficient. The weight layer is configured based on the corresponding weight coefficients, so each weight coefficient in the weight layer is associated with an input node of the feature term corresponding to that weight coefficient. After the complete weight layer configuration, a complete correlation analysis model can be generated. The configured correlation analysis model can be a deep learning model, such as a Transformer model or an LSTM model.

[0074] The obtained correlation analysis model needs to be trained before use to improve its accuracy. For example, a training order set can be constructed by selecting historical transportation orders corresponding to specific transportation types for the correlation analysis model. Feature values ​​of the corresponding feature items of each historical transportation order in the training order set and the input nodes of the correlation analysis model are obtained. These feature values ​​are then input into the correlation analysis model. The corresponding output values ​​are obtained from the output nodes and compared with the loss frequencies recorded in the historical transportation orders to obtain the loss value. Backward learning is then performed based on this loss value to adjust the correlation coefficients between nodes in the correlation analysis model, thereby optimizing the model. Each historical transportation order can be used to train the correlation analysis model once. By sequentially performing the above operations on each historical transportation order in the training order set, iterative training of the correlation analysis model can be achieved, ultimately resulting in a mature correlation analysis model.

[0075] Once trained, the association analysis model can predict the loss frequency of transportation orders. If all data in a transportation order is known (data collected by data acquisition devices has been recorded, or prediction can be made in real time while acquiring data collected by data acquisition devices), but the actual loss frequency of a transportation order is not recorded before it is completed, then the values ​​corresponding to the input nodes of a specific association analysis model can be obtained based on the known data in the transportation order, and input into the specific association analysis model for prediction. Thus, the output node outputs the predicted loss frequency, which can be used by transportation managers to flexibly adjust transportation plans before the transportation order is completed.

[0076] S150. Perform trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information.

[0077] Furthermore, quality trend analysis can be performed based on the obtained correlation analysis model. First, trend analysis can be performed on the correlation feature values ​​of each transportation type in the preprocessed data. These correlation feature values ​​are extracted from the data of incomplete orders for each transportation type. The correlation feature analysis information includes the correlation features of each transportation type. Based on these correlation features, the feature values ​​corresponding to the correlation features of each transportation type can be obtained from the preprocessed data as correlation feature values. Trend analysis is then performed on each correlation feature value to obtain the correlation feature trend analysis information.

[0078] In a specific embodiment, step S150 includes the following sub-steps: obtaining the basic analysis features corresponding to each transportation type and the associated feature analysis information in the preprocessed data; performing time-series processing on the basic analysis features of each transportation type to obtain time-series feature information corresponding to each transportation type; fitting each key feature value in the time-series feature information of each transportation type to obtain the corresponding feature fitting curve; and extracting the associated trend feature value corresponding to the future time point from the feature fitting curve of each transportation type as the corresponding associated feature trend analysis information.

[0079] Specifically, based on the correlation features of transportation types in the correlation feature analysis information, the feature values ​​corresponding to the correlation features of each transportation type can be obtained from the preprocessed data as basic analysis features. Since the acquisition time of the feature values ​​in the basic analysis features is different, the feature values ​​contained in the basic analysis features can be sorted chronologically according to the acquisition time. For example, the feature values ​​corresponding to transportation orders on the same day can be averaged and used as the key feature value corresponding to a certain feature item on that day. By obtaining the key feature values ​​of each feature item in the basic analysis features on each date, chronological sorting can be achieved, and chronological feature information corresponding to the basic analysis features can be obtained.

[0080] The process involves fitting data to the associated feature values ​​in the time-series feature information. Specifically, taking multiple key feature values ​​of a certain feature item as an example, the acquisition time of these key feature values ​​is used as the horizontal axis, and the specific values ​​of the key feature values ​​are used as the vertical axis, resulting in a scatter plot in a two-dimensional coordinate system. Based on the distribution of the scatter points in the two-dimensional coordinate system, multiple preset curve functions are fitted to obtain the fitting function corresponding to each curve function. The degree of overlap between each data point in the fitting function and the scatter points is calculated, and the function curve corresponding to the fitting function with the highest degree of overlap is obtained as the feature fitting curve.

[0081] Each feature item in the correlation features of the transportation type can be fitted with a corresponding feature fitting curve. Based on the obtained feature fitting curves, the correlation trend feature values ​​corresponding to future time points can be extracted. Since the horizontal axis value of the feature fitting curve can be arbitrarily set, using the current date t0 as the horizontal axis, the vertical axis values ​​corresponding to the horizontal axis of the feature fitting curve at several future time points, such as t0+12 (12 hours later), t0+24 (24 hours later), t0+48 (48 hours later), and t0+72 (72 hours later), can be obtained. By obtaining the vertical axis values ​​of multiple set future time points in the feature fitting curve, the correlation trend feature value corresponding to that feature fitting curve can be obtained. The correlation trend feature value can contain one or more correlation trend values. By obtaining the correlation trend feature values ​​corresponding to each transportation type, the correlation feature trend analysis information can be obtained.

[0082] S160. Perform quality trend analysis on the correlation feature trend analysis information according to the correlation analysis model to obtain quality early warning analysis information.

[0083] The correlation analysis model constructed based on the above steps can perform quality trend analysis on correlation feature trend analysis information, thereby obtaining corresponding quality early warning analysis information.

[0084] In a specific embodiment, step S160 includes the following sub-steps: inputting the correlation trend feature values ​​of the same future time point in the correlation feature trend analysis information of each transportation type into the corresponding correlation analysis model for quality trend analysis to obtain the loss frequency corresponding to each future time point; obtaining the loss frequency corresponding to each transportation type and each future time point and combining them to obtain the corresponding quality early warning analysis information.

[0085] Specifically, from the correlation trend analysis information of various transportation types, the correlation trend feature value corresponding to the input node in the correlation analysis model of the same future time point can be obtained, and then input into the correlation analysis model of that transportation type for quality trend analysis. For example, the correlation trend feature value corresponding to the input node in the correlation analysis model of the transportation type at the future time point t0+48 (48 hours later) can be obtained as input, thereby obtaining the loss frequency corresponding to the future time point t0+48 (48 hours later). The loss frequency obtained at this time represents the loss frequency obtained by predicting the quality of freight transportation in the following 48 hours. A loss frequency can be obtained for each transportation type and each future time point; by combining the loss frequencies corresponding to each transportation type and each future time point, quality early warning analysis information can be obtained; by analyzing the increase / decrease of the loss frequency of transportation types in the future period, early warning analysis of transportation quality can be achieved.

[0086] S170. Optimize the associated feature analysis information according to the preset quality optimization strategy to obtain quality optimization suggestions for each transportation type.

[0087] Furthermore, the correlation feature analysis information can be optimized based on quality optimization strategies. This allows for the generation of quality optimization suggestions for each transportation type. For example, if the correlation features of a certain transportation type include two features: transportation route and transportation temperature, and it is determined that the high loss proportion is highest in the statistical interval of "transportation route A" (e.g., transportation route A corresponding to the low value interval), then quality optimization suggestions for adjusting the transportation route can be generated based on the quality optimization strategy. Similarly, if it is determined that the high loss proportion is highest in the statistical interval of "extremely high value interval" of the transportation temperature feature, then quality optimization suggestions for reducing the transportation temperature can be generated based on the quality optimization strategy. By obtaining these quality optimization suggestions, managers can easily adjust transportation strategies in a timely manner, thereby improving the quality of cargo transportation.

[0088] The quality intelligent control method based on big data analysis of logistics networks disclosed in the above embodiments includes: collecting and integrating initial data collected by data acquisition equipment to obtain integrated data and preprocessing it; performing correlation analysis on the preprocessed data in conjunction with transportation orders to obtain correlation feature analysis information and configuring a corresponding correlation analysis model; performing trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data to obtain correlation feature trend analysis information and performing quality trend analysis to obtain quality early warning analysis information; and performing optimization analysis on the correlation feature analysis information according to quality optimization strategies to obtain quality optimization suggestions. This quality intelligent control method can integrate and collect initial data from multiple stages of cargo transportation and perform correlation analysis and quality trend analysis in conjunction with transportation orders, thereby obtaining quality early warning analysis information reflecting transportation quality trends and quality optimization suggestions for optimizing the transportation process for different transportation types, significantly improving the efficiency of quality control in the logistics transportation process.

[0089] This invention also provides a quality intelligent control system based on big data analysis of logistics networks, such as... Figure 3As shown, the quality intelligent control system 100 based on logistics network big data analysis is configured on a quality intelligent control platform. The quality intelligent control platform communicates with data acquisition devices set up in the logistics network to realize data information transmission. The quality intelligent control system 100 based on logistics network big data analysis is used to execute any embodiment of the aforementioned quality intelligent control method based on logistics network big data analysis. Specifically, the aforementioned quality intelligent control system 100 based on logistics network big data analysis specifically includes an integrated data acquisition unit 110, a preprocessing data acquisition unit 120, a correlation analysis unit 130, a model configuration unit 140, a correlation feature trend analysis unit 150, a quality trend analysis information acquisition unit 160, and a quality optimization suggestion acquisition unit 170.

[0090] The integrated data acquisition unit 110 is used to acquire the initial data collected by the data acquisition device, and to collect and integrate the initial data according to the pre-stored transportation orders to obtain the corresponding integrated data.

[0091] The preprocessing data acquisition unit 120 is used to preprocess the integrated data according to preset preprocessing rules to obtain corresponding preprocessed data.

[0092] The association analysis unit 130 is used to perform association analysis on key feature information based on a preset feature extraction model and the transportation order to obtain corresponding association feature analysis information.

[0093] The model configuration unit 140 is used to configure the correlation analysis model corresponding to each transportation type based on the correlation feature analysis information.

[0094] The correlation feature trend analysis unit 150 is used to perform trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information.

[0095] The quality trend analysis information acquisition unit 160 is used to perform quality trend analysis on the correlation feature trend analysis information according to the correlation analysis model to obtain quality early warning analysis information.

[0096] The quality optimization suggestion acquisition unit 170 is used to perform optimization analysis on the associated feature analysis information according to the preset quality optimization strategy in order to obtain quality optimization suggestions corresponding to each transportation type.

[0097] The quality intelligent control system based on logistics network big data analysis provided in this embodiment of the invention executes the aforementioned quality intelligent control method based on logistics network big data analysis. It aggregates and integrates the initial data collected by the data acquisition device to obtain integrated data and performs preprocessing. It then performs correlation analysis on the preprocessed data in conjunction with transportation orders to obtain correlation feature analysis information and configures corresponding correlation analysis models. Furthermore, it performs trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data to obtain correlation feature trend analysis information and performs quality trend analysis to obtain quality early warning analysis information. Finally, it optimizes the correlation feature analysis information according to quality optimization strategies to obtain quality optimization suggestions. This quality intelligent control method can integrate and aggregate initial data from multiple stages of cargo transportation and perform correlation analysis and quality trend analysis in conjunction with transportation orders, thereby obtaining quality early warning analysis information reflecting transportation quality trends and quality optimization suggestions for optimizing the transportation process for different transportation types, significantly improving the efficiency of quality control in the logistics transportation process.

[0098] The aforementioned control-based configuration unit can be implemented as a computer program, which can be used in, for example... Figure 4 It runs on the computer device shown.

[0099] Please see Figure 4 , Figure 4 This is a schematic block diagram of a computer device provided in an embodiment of the present invention. This computer device can be a quality intelligent control platform used to execute a quality intelligent control method based on big data analysis of logistics networks for data analysis and logistics quality control.

[0100] See Figure 4 The computer device 500 includes a processor 502, a memory, and a communication interface 505 connected via a communication bus 501. The memory may include a storage medium 503 and internal memory 504.

[0101] The storage medium 503 can store the operating system 5031 and the computer program 5032. When the computer program 5032 is executed, it enables the processor 502 to execute a quality intelligent control method based on big data analysis of the logistics network. The storage medium 503 can be a volatile storage medium or a non-volatile storage medium.

[0102] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0103] The internal memory 504 provides an environment for the computer program 5032 in the storage medium 503 to run. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a quality intelligent control method based on big data analysis of logistics networks.

[0104] This communication interface 505 is used for network communication, such as providing data transmission. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device 500 to which the present invention is applied. The specific computer device 500 may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0105] The processor 502 is used to run the computer program 5032 stored in the memory to implement the corresponding functions in the above-mentioned quality intelligent control method based on big data analysis of logistics networks.

[0106] Those skilled in the art will understand that Figure 4 The embodiments of the computer device shown do not constitute a limitation on the specific configuration of the computer device. In other embodiments, the computer device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. For example, in some embodiments, the computer device may include only memory and a processor. In such embodiments, the structure and function of the memory and processor are different from those shown. Figure 4 The embodiments shown are consistent and will not be repeated here.

[0107] It should be understood that, in this embodiment of the invention, the processor 502 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, 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, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0108] In another embodiment of the invention, a computer-readable storage medium is provided. This computer-readable storage medium may be volatile or non-volatile. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps included in the aforementioned quality intelligent control method based on big data analysis of logistics networks.

[0109] Those skilled in the art will readily understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0110] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Units with the same function may be grouped into one unit. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0111] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0112] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0113] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned computer-readable storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), magnetic disks, or optical disks.

[0114] The above description is merely a specific embodiment 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 quality intelligent control method based on logistics network big data analysis, characterized in that, The method is applied to a quality intelligent control platform, which communicates with data acquisition devices set up in the logistics network to achieve data information transmission. The method includes: The initial data collected by the data acquisition device is obtained, and the initial data is aggregated and integrated according to the pre-stored transportation orders to obtain the corresponding integrated data; The integrated data is preprocessed according to preset preprocessing rules to obtain corresponding preprocessed data; Based on the preset feature extraction model and the transportation order, the preprocessed data is subjected to correlation analysis to obtain the corresponding correlation feature analysis information; Based on the association feature analysis information, an association analysis model corresponding to each transportation type is configured; Trend analysis is performed on the correlation feature values ​​corresponding to each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information. Based on the correlation analysis model, quality trend analysis is performed on the correlation feature trend analysis information to obtain quality early warning analysis information; The associated feature analysis information is optimized and analyzed according to the preset quality optimization strategy to obtain quality optimization suggestions for each type of transportation. The step of performing correlation analysis on the preprocessed data based on a preset feature extraction model and the transportation order to obtain corresponding correlation feature analysis information includes: The corresponding key feature information is extracted from the preprocessed data according to the preset feature extraction model; Based on the transportation order, the key feature information is analyzed to obtain the associated features corresponding to each transportation type as the corresponding associated feature analysis information. The step of performing correlation analysis on the key feature information based on the transportation order to obtain the correlation features corresponding to each transportation type as the corresponding correlation feature analysis information includes: Constructing variable pairs corresponding to each transportation type based on the transportation orders and key feature information includes: obtaining the high loss ratio corresponding to each feature item in each statistical interval based on the feature items in the key feature information of each transportation type; combining the order ratio of each feature item in each statistical interval with the high loss ratio corresponding to each statistical interval to construct a set of variable pairs corresponding to the feature items; and obtaining multiple sets of variable pairs corresponding to the transportation type. Correlation analysis was performed on the variable pairs for each transportation type to obtain the correlation degree of each variable pair; Based on the correlation between the variable pairs of each transportation type, the variable pairs contained in each transportation type are filtered to obtain the corresponding correlated variable pairs as the correlation features of each transportation type; The step of configuring the correlation analysis model corresponding to each transportation type based on the correlation feature analysis information includes: The input nodes of the initial analysis model are configured based on the association features of each transportation type in the association feature analysis information. Based on the correlation degree of the correlation variables in the correlation characteristics of each transportation type, the corresponding initial analysis model for output node configuration is weighted and configured to obtain the correlation analysis model corresponding to each transportation type. 2.The quality intelligent control method based on logistics network big data analysis of claim 1, wherein, The step of extracting corresponding key feature information from the preprocessed data according to a preset feature extraction model includes: Obtain the feature values ​​corresponding to each feature term in the feature extraction model; The feature values ​​corresponding to each feature item are classified according to the transportation order to obtain the classification feature information corresponding to each transportation type; According to the statistical rules of the feature extraction model, the classification feature information of each transportation type is statistically analyzed to obtain the corresponding feature statistical values ​​as key feature information. 3.The quality intelligent control method based on logistics network big data analysis according to claim 1 or 2, characterized in that, The step of performing trend analysis on the correlation feature values ​​corresponding to each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information includes: Obtain the basic analytical features corresponding to each transportation type and the associated feature analysis information in the preprocessed data; The basic analytical characteristics of each type of transportation are organized in a time series to obtain the time series characteristic information corresponding to each type of transportation. By fitting the key feature values ​​in the temporal feature information of each transportation type, the corresponding feature fitting curves are obtained; The correlation trend feature values ​​corresponding to future time points are extracted from the feature fitting curves of each transportation type as the corresponding correlation feature trend analysis information. 4.The quality intelligent control method based on logistics network big data analysis of claim 3, wherein, The step of performing quality trend analysis on the correlation feature trend analysis information based on the correlation analysis model to obtain quality early warning analysis information includes: Input the correlation trend feature values ​​of the same future time point in the correlation feature trend analysis information of each transportation type into the corresponding correlation analysis model to perform quality trend analysis, so as to obtain the loss frequency corresponding to each future time point; By obtaining and combining the loss frequencies corresponding to each transportation type and each future time point, corresponding quality early warning analysis information can be obtained.

5. A quality intelligent control system based on big data analysis of logistics networks, characterized in that, The system is configured on a quality intelligent control platform, which communicates with data acquisition devices set up in the logistics network to achieve data transmission. The system is used to execute the quality intelligent control method based on big data analysis of the logistics network as described in any one of claims 1-4. The system includes: An integrated data acquisition unit is used to acquire the initial data collected by the data acquisition device, and to collect and integrate the initial data according to the pre-stored transportation orders to obtain the corresponding integrated data. A preprocessing data acquisition unit is used to preprocess the integrated data according to preset preprocessing rules to obtain corresponding preprocessed data; The correlation analysis unit is used to perform correlation analysis on key feature information based on the preset feature extraction model and the transportation order to obtain the corresponding correlation feature analysis information; The model configuration unit is used to configure the correlation analysis model corresponding to each transportation type based on the correlation feature analysis information. The correlation feature trend analysis unit is used to perform trend analysis on the correlation feature values ​​of each transportation type in the preprocessed data and the correlation feature analysis information to obtain the corresponding correlation feature trend analysis information. The quality trend analysis information acquisition unit is used to perform quality trend analysis on the correlation feature trend analysis information according to the correlation analysis model to obtain quality early warning analysis information. The quality optimization suggestion acquisition unit is used to perform optimization analysis on the associated feature analysis information according to the preset quality optimization strategy in order to obtain quality optimization suggestions corresponding to each transportation type.

6. A computer device, characterized in that, The device includes 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; When a processor executes a program stored in a memory, it implements the steps of the quality intelligent control method based on big data analysis of logistics networks as described in any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the quality intelligent control method based on big data analysis of logistics networks as described in any one of claims 1-4.

Images