Truck overrun and overload classification and grading management and control system based on big data portrait technology

Abstract

The invention discloses a truck overrun and overload classification and grading management and control system based on a big data portrait technology, and belongs to the technical field of traffic transportation management, and the system specifically comprises the steps: collecting truck dynamic driving data from a multivariate data source, and carrying out the preprocessing; deeply integrating the data by using a multi-modal feature fusion strategy, extracting truck multi-dimensional comprehensive feature information, constructing big data portrait templates of different types of trucks by using data mining and machine learning on the basis of the truck multi-dimensional comprehensive feature information, and accurately outlining the features of various trucks; based on the portraits, whether the truck is overloaded or not is judged in real time in combination with real-time dynamic driving data, accurate grading is performed according to the overloading proportion, the illegal frequency and the harm consequence, differential management and control are performed for trucks of different grades, the problem of overloading and overload supervision of the trucks is solved, and the traffic transportation management efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of transportation management, and specifically to a classification and grading control system for overweight and overloaded trucks based on big data profiling technology. Background Art

[0002] Overloaded trucks not only cause severe damage to highway infrastructure, shorten highway lifespans, and increase maintenance costs, but also significantly impact traffic safety and can easily lead to accidents. Traditional methods for managing overloaded trucks rely primarily on fixed or mobile weighing stations. This approach suffers from limited coverage, the risk of missed inspections, and an inability to monitor truck traffic in real time, making it difficult to achieve accurate and efficient management and control.

[0003] For example, Chinese patent application publication number CN116644304A discloses a similarity-based method for extracting and classifying abnormal truck bearing features. The method includes: collecting basic data, including time-domain acceleration signals from truck bearings; preprocessing the basic data and dividing it into a training set and a test set; converting the time-domain data in both the training and test sets into frequency-domain data; establishing a twin network feature extraction model; and including a feature calculation network and a twin network within the resulting twin network feature extraction model. The feature calculation network and the twin network have identical structures and share parameters. Feature extraction is performed using the twin network feature extraction model. This technical solution can intelligently and accurately identify and analyze abnormal truck bearings, helping to identify early-stage truck bearing failures with high accuracy and real-time performance.

[0004] For example, a Chinese patent with authorization announcement number CN112905578B discloses a method for identifying stop points on a truck's GPS trajectory, including: cleaning the GPS trajectory data of the truck based on its unique vehicle identification ID and chaotic time series, i.e., filtering the GPS trajectory data to eliminate duplicate and invalid data; classifying and clustering the GPS trajectory data with different unique vehicle identification IDs in the same data file based on the cleaned GPS trajectory data; reordering the time series of the data based on the GPS trajectory data classified by the vehicle ID; calculating the density value based on the processed GPS trajectory data and according to the actual research requirements on the stay time of the stay points to determine the stay points in the truck's GPS trajectory data; calculating the stay time of the vehicle at each stay point based on the determined stay point data, and filtering the stay points according to the stay time limit required by the actual research, and finally storing them as a local TXT text file.

[0005] The above existing technologies all have the following problems: narrow application scenarios, single data dimensions; and lack of systematic management and control measures. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention proposes a classification and grading control system for overweight and overloaded trucks based on big data portrait technology, which collects and pre-processes dynamic driving data of trucks from multiple data sources; uses a multimodal feature fusion strategy to deeply integrate data, extracts multi-dimensional comprehensive feature information of trucks, and uses data mining and machine learning to construct big data portrait templates of different types of trucks based on the multi-dimensional comprehensive feature information of trucks, accurately outlining the characteristics of various types of trucks; based on the portrait, combined with real-time driving data, it is judged in real time whether the truck is overweight or overloaded, and accurately graded according to the overweight and overload ratio, violation frequency and harmful consequences, and implements differentiated control for trucks of different grades, which solves the problem of supervising overweight and overloaded trucks and improves the efficiency of transportation management.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The truck overload and overweight classification and grading control system based on big data profiling technology includes: data analysis module, profiling module, judgment and grading module, and control and execution module;

[0009] The data analysis module includes a data acquisition unit and a fusion unit. The data acquisition unit is used to collect dynamic driving data of trucks through sensors. The fusion unit is configured with a multimodal feature fusion strategy. The multimodal feature fusion strategy is used to deeply fuse the pre-processed dynamic driving data of trucks to extract multi-dimensional comprehensive feature information of trucks.

[0010] The portrait construction module includes a cluster analysis unit and a portrait template generation unit. The cluster analysis unit is configured with an improved clustering strategy, which is used to determine whether the current point in the fused truck data is a core point and perform clustering; the portrait template generation unit is used to generate a big data portrait template for each truck group;

[0011] The judgment and classification module is used to combine the real-time collected dynamic driving data of trucks and big data portraits to comprehensively judge whether the trucks are overloaded and overweight, and classify them according to the preset classification rules;

[0012] The control and execution module includes a measure formulation unit, which is used to formulate differentiated control measures according to the classification results.

[0013] Specifically, the judgment and classification module includes a behavior judgment unit, a classification rule setting unit, and a classification processing unit;

[0014] The behavior judgment unit is used to judge the overloaded or over-limited behavior of the truck by using the big data portrait template and the real-time dynamic driving data of the truck;

[0015] The classification rule setting unit is used to set classification rules based on the proportion of overload and overlimit, the frequency of historical violations, and the harmful consequences caused;

[0016] The grading processing unit is used to perform grading processing on overloaded trucks according to grading rules.

[0017] Specifically, the fusion unit adopts a multimodal feature fusion strategy, and the specific steps of the multimodal feature fusion strategy include:

[0018] A1: Obtain pre-processed truck dynamic driving data from distributed data storage nodes and accurately divide the pre-processed truck dynamic driving data into five modal groups based on data source and deep feature attributes;

[0019] A2: Using a distributed computing framework, divide the data of each modality group into N data blocks and assign the task of normalizing the data of each modality group to different computing nodes.

[0020] A3: On each computing node assigned to the task, use the normalization formula to calculate the data block for each node to obtain the normalized dynamic driving data of the truck. , and aggregate the results back to the master node of the distributed computing framework, where represents the normalized dynamic driving data of the truck in the jth data block of the i-th modal group, and i ∈ [ 1 , 5 ], j ∈ [ 1 , N ] ;

[0021] A4: Arrange the normalized data of each modal group according to the feature dimension to form the normalized feature vector of each modal group. ,in, represents the normalized dynamic driving data of the truck in the Nth data block of the i-th modal group.

[0022] Specifically, the specific steps of the multimodal feature fusion strategy also include:

[0023] A5: Normalize the eigenvectors according to each modal group , by calculating the correlation between the two modal groups , and generate a multidimensional correlation matrix ,in, represents the normalized dynamic driving data of the nth truck in the ith modal group, n represents the number of dynamic driving data of trucks, and , Indicates the number of truck dynamic driving data in each data block, represents the correlation between the e-th modal group and the r-th modal group, and , r and e represent two different modal groups.

[0024] Specifically, the specific steps of the multimodal feature fusion strategy also include:

[0025] A6: Based on multidimensional correlation matrix , introduces the attention mechanism, combines the output of the attention mechanism with the business priority rules, calculates the real-time weights of different modalities through the improved Softmax function, and forms a dynamic weight vector W = [ Soft max( A ′ 1 , λ , u , v ), … , Soft max( A ′ 5 , λ , u , v )] ,in, A ′ i , i ∈ [ 1 , 5 ] represents the importance score of the adjusted i-th modal group, represents the regularization parameter, u represents the bias term, and v represents the weight adjustment parameter;

[0026] A7: Perform weighted summation on the feature vectors of each mode according to the dynamic weight vector to obtain the final fused truck data, which is then stored in the feature extraction input buffer.

[0027] Specifically, the specific steps of A6 include:

[0028] A6.1: Obtaining a Multidimensional Incidence Matrix , and introduce the attention mechanism;

[0029] A6.2: The elements in are used as the input of the attention mechanism, through , get the attention score vector A = [ A 1 , … , A 5 ] ,in, A i , i ∈ [ 1 , 5 ] represents the importance score of the i-th modal group, represents the attention mechanism function;

[0030] A6.3: Define priority rules for different modalities based on business needs, and obtain business priority score vectors based on the priority rules for different modalities. P = [ P 1 , … , P 5 ] The priority rule is to prioritize based on the scoring results of the reliability, importance, and real-time factors of the modality, where: P i , i ∈ [ 1 , 5 ] represents the priority score vector of the i-th modality group.

[0031] Specifically, the specific steps of A6 further include:

[0032] A6.4: According to the formula , get the adjusted score vector A ′ = [ A ′ 1 , … , A ′ 5 ] ,in, A ′ i , i ∈ [ 1 , 5 ] represents the importance score of the i-th modal group after adjustment;

[0033] A6.5: Add the adjusted score vector A ′ = [ A ′ 1 , … , A ′ 5 ] Input into the improved Softmax function to calculate the real-time weights of different modes and generate a dynamic weight vector W = [ Soft max( A ′ 1 , λ , u , v ), … , Soft max( A ′ 5 , λ , u , v )] ,in, represents the improved Softmax function, represents the regularization parameter.

[0034] Specifically, the cluster analysis unit adopts an improved clustering strategy, and the specific steps of the improved clustering strategy include:

[0035] B1: Extract the final fused truck data obtained from A7 from the buffer , according to the clustering requirements, set the core point judgment threshold h, where, represents the qth final fused truck data, q represents the number of final fused truck data;

[0036] B2: For Each data point in z g , g ∈ [ 1 , q ] , calculate the density around each data point ;

[0037] B3: The density around each data point Compare with h;

[0038] like , then determine the data point As the core point;

[0039] B4: Taking the core point as the center, the surrounding data points are classified into the same cluster according to the set h;

[0040] B5: For the formed clusters, continue to look for clusters that are adjacent to the formed clusters and satisfy data points and will satisfy The data points are classified into the cluster until no more data points that meet the conditions are found;

[0041] B6: Extract characteristic information of each truck group from the clustering results, and fill the extracted characteristic information into a preset big data portrait template to generate a big data portrait template for each truck group. The characteristic information of the truck group includes driving route, speed distribution, and stop time.

[0042] Specifically, the five modal groups in A1 include: basic vehicle attribute data, transportation behavior characteristic data, violation record data, weather condition data and cargo information; the basic vehicle attribute data includes vehicle model, age, load tonnage, and affiliated company; the transportation behavior characteristic data includes transportation route, transportation frequency, commonly loaded cargo type, and average driving speed; the violation record data includes the number of times, location, and severity of overloading; the weather condition data includes rainy days and strong winds; and the cargo information includes special cargo attributes.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention proposes a classification and grading control system for overweight and overloaded trucks based on big data profiling technology, and optimizes and improves the architecture, operating steps and processes. The system has the advantages of simple processes, low investment and operating costs, and low production costs.

[0045] 2. The present invention proposes a classification and grading control system for overweight and overloaded trucks based on big data profiling technology. By collecting dynamic driving data of trucks from multiple sources, it not only ensures that the data is real-time and accurate, but also provides a basis for subsequent analysis after pre-processing; it uses a multimodal feature fusion strategy to deeply mine data, integrates multi-dimensional comprehensive features such as vehicle basic attributes, transportation behavior, violations of laws and regulations, weather conditions, cargo information, etc., to comprehensively outline the overall picture of the truck, effectively improving the ability to accurately understand the operating status of the truck.

[0046] 3. The present invention proposes a classification and grading control system for overweight and overloaded trucks based on big data portrait technology. It uses data mining and machine learning to build a big data portrait template, and combines real-time driving data to make overweight and overload judgments and classifications, making the control more targeted. It can accurately divide the overweight and overloaded trucks into mild, moderate and severe categories based on the degree of overweight and overload, historical violations, and harmful consequences, and then formulate differentiated control measures for different levels, forming a complete closed loop from source prevention, process supervision to post-event punishment, effectively ensuring road transportation safety, standardizing logistics market order, and reducing safety risks while improving traffic management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is the architecture diagram of the truck overload classification and grading control system based on big data profiling technology of the present invention;

[0048] Figure 2 This is a principle flow chart of the truck overload classification and grading control system based on big data profiling technology of the present invention;

[0049] Figure 3 This is a flowchart of the algorithm implementation of the fusion unit of the truck overload classification and grading control system based on big data profiling technology of the present invention;

[0050] Figure 4 This is a flowchart of the algorithm implementation of the cluster analysis unit of the truck overload classification and grading control system based on big data profiling technology in the present invention. DETAILED DESCRIPTION

[0051] Example 1

[0052] See also Figure 1 and Figure 2 The present invention provides an embodiment of a truck overload classification and grading control system based on big data profiling technology, comprising:

[0053] Data analysis module, portrait construction module, judgment and classification module, control and execution module;

[0054] The data analysis module is used to collect dynamic driving data of trucks from multiple data sources and pre-process the dynamic driving data of trucks to ensure the accuracy, completeness and consistency of the data;

[0055] The portrait construction module is used to use data mining algorithms and machine learning models to construct big data portrait templates of different types of trucks based on the extracted multi-dimensional comprehensive feature information of trucks;

[0056] The judgment and classification module is used to combine real-time collected dynamic driving data of trucks with big data portraits to comprehensively judge whether the trucks are overloaded or overweight, and classify them according to preset classification rules;

[0057] The management, control and execution module is used to formulate and implement differentiated management and control measures for overloaded trucks of different classifications and grades.

[0058] The data analysis module includes: data acquisition unit, preprocessing unit, and fusion unit;

[0059] A data acquisition unit, used to collect dynamic driving data of the truck through sensors;

[0060] A preprocessing unit is used to preprocess the collected dynamic driving data of trucks, including deduplication, missing value filling, and outlier processing, so as to improve the data quality;

[0061] The fusion unit is used to deeply fuse the preprocessed dynamic driving data of the truck using a multimodal feature fusion strategy to extract multi-dimensional comprehensive feature information of the truck.

[0062] The portrait construction module includes: a cluster analysis unit and a portrait template generation unit;

[0063] Cluster analysis unit, used to perform cluster analysis on the multi-dimensional comprehensive feature information of trucks and identify different types of truck groups;

[0064] The portrait template generation unit is used to generate a representative big data portrait template for each truck group, including feature vectors and typical feature descriptions.

[0065] The judgment and classification module includes: behavior judgment unit, classification rule setting unit, and classification processing unit;

[0066] The behavior judgment unit is used to judge the overloaded or over-limit behavior of trucks by using big data portrait templates and real-time dynamic driving data of trucks;

[0067] The classification rule setting unit is used to set classification rules based on factors such as the proportion of overload and overlimit, the frequency of historical violations, and the harmful consequences caused;

[0068] The grading processing unit is used to grade overloaded trucks according to grading rules, providing a basis for subsequent control measures.

[0069] The control and execution module includes: measure formulation unit, measure execution unit, and evaluation and feedback unit;

[0070] The measures formulation unit is used to formulate differentiated control measures based on the classification results, such as warnings, fines, and revocation of operating licenses;

[0071] The measures execution unit is used to implement the formulated differentiated control measures, including notifying truck drivers and enforcing penalties;

[0072] The evaluation and feedback unit is used to evaluate the effectiveness of the implementation of control measures and adjust and optimize the measures based on the evaluation results.

[0073] In summary, the overall implementation process of the truck overload and overload classification and grading control system based on big data profiling technology includes:

[0074] Step S1: Collecting dynamic driving data of trucks from multiple data sources, including weighing data from highway toll stations, inspection data from overweight control stations, Beidou satellite navigation system or GPS trajectory data for trucks, vehicle management office registration information, logistics company transportation order information, and dynamic data such as axle weight, gross weight, speed, and tire pressure during the truck's driving process, to ensure the real-time and accuracy of the data, and pre-processing the collected dynamic driving data of trucks;

[0075] Step S2: A multimodal feature fusion strategy is used to deeply fuse the pre-processed truck dynamic driving data to extract multi-dimensional comprehensive feature information of the truck. The multi-dimensional comprehensive feature information of the truck includes basic vehicle attributes, such as vehicle model, age, load tonnage, and company; transportation behavior characteristics, such as transportation route, transportation frequency, common cargo type, and average driving speed; violation records, including the number, location, and severity of overload and overlimit violations; real-time weather conditions, such as the need to reduce speed on slippery roads in rainy days and the impact of strong winds on the stability of high-box trucks; and cargo information, such as special cargo attributes such as fragile goods, flammable and explosive goods.

[0076] Step S3: Using a data mining algorithm and a machine learning model, the multi-dimensional comprehensive feature information of the trucks extracted in step S2 is used as input, and cluster analysis is performed based on similar features to construct big data portrait templates of different types of trucks. Each portrait template includes a set of representative feature vectors that describe the typical characteristics of that type of truck. The data mining algorithm and machine learning model are prior art in this field and do not constitute the inventive solution of this application, and are not described in detail here.

[0077] Step S4: Based on the truck big data portrait and the real-time collected dynamic driving data of the truck, a comprehensive judgment is made on whether the truck has overloaded or exceeded the limit. For trucks that are determined to be overloaded or exceeded, they are classified according to the preset classification rules based on factors such as the overload ratio, the frequency of historical violations, and the harmful consequences caused. For example, a truck that is overloaded or exceeded by less than 30% and is the first violation is classified as a light overloaded truck; a truck that is overloaded or exceeded by 30% or more and has violated the limit 2-3 times within a year is classified as a moderate overloaded truck; and a truck that is overloaded or exceeded by more than 50% or has repeatedly caused serious traffic congestion or road damage is classified as a severe overloaded truck.

[0078] Furthermore, the specific steps of step S4 include:

[0079] (1) Data collection and integration:

[0080] Collect big data portrait information of trucks, including historical driving data, load records, and vehicle model information; at the same time, collect real-time driving data of trucks, such as current load, speed, location, etc.

[0081] (2) Judgment of over-limit and overload:

[0082] Compare the actual load of the truck with the approved load, and calculate the overload rate by (actual load - approved load) / approved load × 100%. Use the overload rate to determine whether there is any overloading behavior;

[0083] (3) Setting of classification rules:

[0084] Setting classification rules, which can be based on factors such as the proportion of overloads and limits, the frequency of historical violations, and the harmful consequences caused;

[0085] (4) Comprehensive assessment and grading:

[0086] Based on real-time collected dynamic driving data and truck big data portrait information, the system comprehensively evaluates truck overload and over-limit behaviors, and classifies trucks into different levels according to the set classification rules;

[0087] (5) Result output and application:

[0088] Output the grading results and take corresponding measures according to different levels, such as warnings, fines, and temporary vehicle seizures. The grading results will be used for subsequent supervision and management to improve the safety and compliance of truck transportation.

[0089] Step S5: Formulate and implement differentiated control measures for overloaded trucks of different classifications and grades output in step S4;

[0090] For trucks that are slightly overloaded or exceeding the limit, an early warning message will be sent to the driver through the on-board terminal, reminding him to go to the nearest designated unloading point to unload the goods. The violation information will be recorded and the affiliated company will be notified to strengthen safety education.

[0091] For moderately overloaded trucks, in addition to the aforementioned warnings and notifications, their driving trajectory will be tracked in real time, and they will be directed to nearby overload control stations for processing. Fines, point deductions, and other penalties will be imposed in accordance with the law, and the companies will be required to make corrections within a specified period of time.

[0092] For severely overloaded trucks, a joint law enforcement mechanism will be immediately activated to intercept them, force them to unload their cargo, seize their vehicles, and impose penalties on the drivers and their companies, including heavy fines, suspension of business, and revocation of relevant qualifications. The companies will also be included in the key supervision list, and the frequency of daily inspections will be increased.

[0093] Example 2

[0094] See also Figure 3 In this embodiment, the fusion unit adopts a multimodal feature fusion strategy. The specific steps of the multimodal feature fusion strategy include:

[0095] A1: Obtain pre-processed truck dynamic driving data from distributed data storage nodes and accurately divide the pre-processed truck dynamic driving data into five modal groups based on data source and deep feature attributes;

[0096] Among them, data sources, such as from different sensor types and different acquisition devices; deep feature attributes, such as speed-related features and load-related features.

[0097] Furthermore, the specific steps for accurately dividing the pre-processed truck dynamic driving data into five modal groups include:

[0098] (1) Data feature combing: Comprehensively comb the pre-processed truck dynamic driving data to identify all potential data features;

[0099] (2) Formulate classification rules for modal groups: Based on the five known modal groups and the specific data categories they contain, formulate precise classification rules:

[0100] For the vehicle basic attribute data modal group: set rules. As long as the data features directly describe the inherent characteristics of the vehicle itself, such as information clearly marked as vehicle model, age, load tonnage, and company, it will be classified into this modal group. For example, in a data record: Dongfeng Tianlong heavy truck, age 5 years, load tonnage 30 tons, company A Logistics, all elements in it should be classified into this modal group.

[0101] For the transport behavior characteristic data modal group: If the data reflects the behavioral patterns and dynamic conditions of trucks during operation, such as the transport route is a series of geographic coordinates or road section names, the transport frequency is the number of trips within a certain time period, the common cargo type is the cargo name or category, and the average driving speed is the speed value, all of these data should be classified into this modal group. For example, if the transport frequency this month is 15 times, the common cargo is steel, the transport route is from City B to City C, and the average driving speed is 60 km / h, these data should be placed in this modal group.

[0102] Regarding the violation record data modal group: When data involves past violations of trucks, specifically the number of violations related to overloading and exceeding limits, the location of occurrence, and the severity, where the severity is expressed in terms of fines, points deducted, or the rating by law enforcement agencies, it is classified into the violation record data modal group. For example, on the D highway section, if there were three violations and the severity was medium, this information would be classified into this modal group.

[0103] For the weather data modal group: Data comes from the meteorological monitoring system and describes the weather phenomena encountered by trucks during driving, such as rainy days, strong winds, and other weather tags and the corresponding time and location ranges. This data is classified into this modal group. For example, on July 10, 2023, the road section from City B to City C encountered rainy days, and this data is classified into this modal group.

[0104] For the cargo information modal group: any data related to the special properties and categories of the cargo itself, such as descriptions of special cargo attributes such as fragile goods, flammable and explosive goods, or annotations related to special goods in the detailed cargo manifest, should be classified into the cargo information modal group. For example, if the cargo included fragile porcelain, this information would be placed in this modal group.

[0105] (3) Data classification and preliminary grouping:

[0106] According to the established classification rules, each pre-processed truck dynamic driving data is traversed. Through programming methods, such as using the Pandas library in Python for data processing, conditional judgment statements are applied to determine the modal group to which each data feature should belong. For example, for a comprehensive data record containing multiple features, each feature is checked in turn. If it meets the classification rules for vehicle basic attribute data, it is added to the temporary storage structure of vehicle basic attribute data; if it meets the rules for transportation behavior feature data, it is stored in the corresponding storage location of transportation behavior feature data, and so on, completing the preliminary grouping work.

[0107] (4) Group verification and adjustment:

[0108] After the preliminary grouping is completed, each modal group is verified. On the one hand, the data in each modal group is checked for consistency and logic to avoid incorrect classification. For example, in the vehicle basic attribute data modal group, if a feature similar to the current transport speed of 80km / h is found, it obviously does not meet the definition of the modal group and should be reclassified into the transport behavior characteristic data modal group. On the other hand, it is ensured that the five modal groups can fully cover all pre-processed truck dynamic driving data without omissions or repeated classifications. If problems are found, the classification strategy is adjusted in time and the incorrectly classified data is reallocated until the grouping results are accurate.

[0109] A2: Using a distributed computing framework, divide the data of each modality group into N data blocks and assign the task of normalizing the data of each modality group to different computing nodes.

[0110] Furthermore, the specific steps of A2 include:

[0111] (1) In a distributed computing framework, the normalization processing tasks are assigned to different computing nodes. Each computing node is responsible for processing the normalization of one or more data blocks, and each computing node is ensured to have the resources and environment required to perform the normalization processing.

[0112] (2) The divided data blocks are distributed to the corresponding computing nodes through a distributed file system. The distributed file system is a prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0113] A3: On each computing node assigned to a task, a normalization method is used to calculate the data block for which it is responsible, and the results are aggregated back to the master node of the distributed computing framework through the communication and synchronization mechanism of the distributed computing framework. The formula is:

[0114] ;

[0115] in, represents the normalized dynamic driving data of the truck in the jth data block of the i-th modal group, represents the dynamic driving data of the truck in the jth data block of the i-th modal group, express The corresponding minimum value in express The corresponding maximum value in , a and b represent scaling factors, and c represents the coefficient of nonlinear transformation. Represents a positive number, represents a nonlinear function;

[0116] A4: Arrange the normalized data of each modal group according to the feature dimension to form the normalized feature vector of each modal group. ,in, represents the normalized dynamic driving data of the truck in the Nth data block of the i-th modal group;

[0117] A5: Normalize the eigenvectors according to each modal group , calculate the correlation between two modal groups and generate a multidimensional correlation matrix , the formula is:

[0118] ;

[0119] in, represents the normalized dynamic driving data of the nth truck in the ith modal group, n represents the number of dynamic driving data of trucks, and , Indicates the number of truck dynamic driving data in each data block, represents the correlation between the e-th modal group and the r-th modal group, represents the weighting factor of the k-th truck dynamic driving data in the normalized e-th modal group and the r-th modal group, and represent the normalized dynamic driving data of the kth truck in the eth modal group and the rth modal group, respectively. and represent the means of the normalized dynamic driving data of trucks in the e-th mode group and the r-th mode group, respectively. and They represent the normalized offsets of the k-th truck dynamic driving data in the e-th modal group and the r-th modal group, respectively. r and e represent two different modal groups. represents the correlation between the i-th modal group and the i-1-th group;

[0120] A6: Based on multidimensional correlation matrix , introduces the attention mechanism, combines the output of the attention mechanism with the business priority rules, calculates the real-time weights of different modalities through the improved Softmax function, and forms a dynamic weight vector ,in, represents the importance score of the adjusted i-th modal group, represents the regularization parameter, u represents the bias term, and v represents the weight adjustment parameter;

[0121] A7: Perform weighted summation on the feature vectors of each mode according to the dynamic weight vector to obtain the final fused truck data, which is then stored in the feature extraction input buffer.

[0122] In the present invention, it can be seen from step A6 that the dynamic weight vector is , perform weighted summation on the dynamic weight vector: ,right Discretize to get the final fused truck data , represents the weighting coefficient of the dynamic weight vector, Represents the weighted sum of the dynamic weight vectors.

[0123] The specific steps of A6 include:

[0124] A6.1: Obtaining a Multidimensional Incidence Matrix , and introduces the attention mechanism, wherein the attention mechanism is the existing technical content in this field and is not the inventive solution of this application, and will not be described in detail here;

[0125] A6.2: The elements in are used as the input of the attention mechanism, through , get the attention score vector ,in, represents the importance score of the i-th modal group, represents the attention mechanism function;

[0126] A6.3: Define priority rules for different modalities based on business needs, and obtain business priority score vectors based on the priority rules for different modalities. The priority rule is to prioritize based on the scoring results of the reliability, importance, and real-time factors of the modality, where: represents the priority score vector of the i-th modality group;

[0127] A6.4: According to the formula , get the adjusted score vector ,in, represents the importance score of the i-th modal group after adjustment;

[0128] A6.5: Add the adjusted score vector Input into the improved Softmax function to calculate the real-time weights of different modes and generate a dynamic weight vector , the calculation formula of the dynamic weight vector is:

[0129] ;

[0130] in, represents the improved Softmax function, represents the regularization parameter, represents the regularization term of the i-th modal group, represents the bias term of the i-th modal group, represents the weight adjustment parameter of the i-th modal group, Represents the exponential function.

[0131] Example 3

[0132] See also Figure 4 In this embodiment, the cluster analysis unit adopts an improved clustering strategy. The specific steps of the improved clustering strategy include:

[0133] B1: Extract the final fused truck data obtained from A7 from the buffer , according to the clustering requirements, set the core point judgment threshold h, where, represents the qth final fused truck data, q represents the number of final fused truck data;

[0134] B2: For Each data point in , calculate the density around each data point ;

[0135] In the present invention, the specific steps of calculating the surrounding density include:

[0136] (1) Setting parameters: First, you need to set the two key parameters of the DBSCAN algorithm: neighborhood radius and minimum number of points. , wherein, the DBSCAN algorithm is the existing technical content in this field and is not the inventive solution of this application, and is not described here in detail;

[0137] (2) Calculate the neighborhood: For each data point in the dataset, calculate the number of data points in its neighborhood. This is usually achieved by traversing the dataset and checking whether the distance between each data point and other data points is less than or equal to the neighborhood radius.

[0138] (3) Determine the core point: If the neighborhood of a data point contains at least data points, then the data point is considered a core point, which means that the density around the data point is high enough to form the core of a cluster.

[0139] B3: The surrounding density of each data point Compare with h;

[0140] like , then determine the data point As the core point;

[0141] B4: Taking the core point as the center, the surrounding data points are classified into the same cluster according to the set h;

[0142] B5: For the formed clusters, continue to look for clusters that are adjacent to the formed clusters and satisfy data points and will satisfy The data points are classified into the cluster until no more data points that meet the conditions are found;

[0143] B6: Extract characteristic information of each truck group from the clustering results, and fill the extracted characteristic information into a preset big data portrait template to generate a big data portrait template for each truck group. The characteristic information of the truck group includes driving route, speed distribution, and stop time.

[0144] Furthermore, the specific steps of B6 include:

[0145] (1) Analyze the clustering results and identify different truck groups, i.e., clusters;

[0146] (2) For each truck group, extract its characteristic information, which may include:

[0147] Driving route: the main driving path or area of ​​the truck group;

[0148] Travel time: The travel time distribution of truck groups, such as peak hours and off-peak hours;

[0149] Dwell time: the dwell time of truck groups at different locations;

[0150] Cargo type: the type of cargo transported by the truck group;

[0151] Speed ​​distribution: the speed distribution of the truck group;

[0152] (3) Design a preset big data portrait template, which should contain fields or areas that can display the characteristic information of the truck group;

[0153] (4) Fill the extracted feature information into the corresponding fields or areas according to the format of the preset big data portrait template to ensure the accuracy and completeness of the information so as to generate a meaningful big data portrait;

[0154] (5) Generate a big data portrait template for each truck group based on the entered feature information;

[0155] (6) Verify the generated big data portrait template to ensure that it can accurately reflect the characteristic information of the truck group.

[0156] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific embodiments. The above-mentioned specific embodiments are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention. These are all protected by the present invention.

[0157] If the technical solution disclosed herein involves personal information, the product using the technical solution disclosed herein has clearly informed the individual of the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using the technical solution disclosed herein has obtained the individual's separate consent before processing the sensitive personal information and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, a clear and prominent sign is set to inform that the individual has entered the personal information collection scope and that personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that they agree to the collection of their personal information; or on the personal information processing device, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload their personal information. The personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A classification and grading control system for overloaded trucks based on big data profiling technology, characterized by: include: Data analysis module, portrait construction module, judgment and classification module, control and execution module; The data analysis module includes a data acquisition unit and a fusion unit. The data acquisition unit is used to collect dynamic driving data of trucks through sensors. The fusion unit is configured with a multimodal feature fusion strategy. The multimodal feature fusion strategy is used to deeply fuse the pre-processed dynamic driving data of trucks to extract multi-dimensional comprehensive feature information of trucks. The portrait construction module includes a cluster analysis unit and a portrait template generation unit. The cluster analysis unit is configured with an improved clustering strategy, and the improved clustering strategy is used to determine whether the current point in the fused truck data is a core point and perform clustering; The portrait template generating unit is used to generate a big data portrait template for each truck group; The judgment and classification module is used to combine the real-time collected dynamic driving data of the truck and the big data portrait to comprehensively judge whether the truck has overloaded or overloaded behavior, and classify it according to the preset classification rules; The control and execution module includes a measure formulation unit, which is used to formulate differentiated control measures according to the classification results.

2. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 1 is characterized in that: The judgment and classification module includes a behavior judgment unit, a classification rule setting unit, and a classification processing unit; The behavior judgment unit is used to judge the overload behavior of the truck by using the big data portrait template and the real-time dynamic driving data of the truck; The classification rule setting unit is used to set the classification rules according to the overload and overlimit ratio, the frequency of historical violations, and the harmful consequences caused; The grading processing unit is used to perform grading processing on overloaded trucks according to grading rules.

3. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 2 is characterized in that: The fusion unit adopts a multimodal feature fusion strategy, and the specific steps of the multimodal feature fusion strategy include: A1: Obtain pre-processed truck dynamic driving data from the distributed data storage node, and accurately divide the pre-processed truck dynamic driving data into five modal groups based on the data source and deep feature attributes; A2: Using the distributed computing framework, the data of each modality group is divided into N data blocks, and the task of normalizing the data of each modality group is assigned to different computing nodes; A3: On each computing node assigned to the task, use the normalization formula to calculate the data block for each node to obtain the normalized dynamic driving data of the truck. , and summarize the results back to the master node of the distributed computing framework, where represents the normalized dynamic driving data of the truck in the jth data block of the ith modal group, and ; A4: Arrange the normalized data of each modal group according to the feature dimension to form the normalized feature vector of each modal group. ,in, Represents the normalized dynamic driving data of the truck in the Nth data block of the ith modal group.

4. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 3 is characterized in that: The specific steps of the multimodal feature fusion strategy also include: A5: Normalize the eigenvectors according to each modal group , by calculating the correlation between the two modal groups , and generate a multidimensional incidence matrix ,in, represents the normalized n-th truck dynamic driving data in the ith modal group, n represents the number of truck dynamic driving data, and , Indicates the number of truck dynamic driving data in each data block, represents the correlation between the e-th mode group and the r-th mode group, and , r and e represent two different modal groups.

5. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 4 is characterized in that: The specific steps of the multimodal feature fusion strategy also include: A6: Based on multidimensional correlation matrix , introduce the attention mechanism, and combine the output of the attention mechanism with the business priority rules, calculate the real-time weights of different modes through the improved Softmax function, and form a dynamic weight vector ,in, represents the importance score of the adjusted i-th modal group, represents the regularization parameter, u represents the bias term, and v represents the weight adjustment parameter; A7: Perform weighted summation on the feature vectors of each mode according to the dynamic weight vector to obtain the final fused truck data, and store it in the feature extraction input buffer.

6. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 5 is characterized in that: The specific steps of A6 include: A6.1: Obtaining a multidimensional incidence matrix , and introduce the attention mechanism; A6.2: The elements in are used as the input of the attention mechanism, through , get the attention score vector ,in, represents the importance score of the i-th modal group, represents the attention mechanism function; A6.3: Define the priority rules for different modalities according to business needs, and obtain the business priority score vector according to the priority rules for different modalities , the priority rule is to prioritize based on the scoring results of the reliability, importance, and real-time factors of the modality, where, Represents the priority score vector of the i-th modality group.

7. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 6 is characterized in that: The specific steps of A6 also include: A6.4: According to the formula , and get the adjusted score vector ,in, represents the importance score of the i-th modal group after adjustment; A6.5: The adjusted score vector Input into the improved Softmax function, calculate the real-time weights of different modes, and generate a dynamic weight vector ,in, represents the improved Softmax function, represents the regularization parameter.

8. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 7 is characterized in that: The cluster analysis unit adopts an improved clustering strategy, and the specific steps of the improved clustering strategy include: B1: Extract the final fused truck data obtained from A7 from the buffer , according to the clustering requirements, set the judgment threshold h of the core point, where, represents the qth final fused truck data, q represents the number of final fused truck data; B2: For Each data point in , calculate the density around each data point ; B3: The density around each data point Compare with h; like , then determine the data point As the core point; B4: Taking the core point as the center, the surrounding data points are classified into the same cluster according to the set h; B5: For the formed clusters, continue to look for clusters that are adjacent to the formed clusters and satisfy of data points and will satisfy The data points are classified into this cluster until no more data points that meet the conditions are found; B6: Extract characteristic information of each truck group from the clustering results, and fill the extracted characteristic information into a preset big data portrait template to generate a big data portrait template for each truck group. The characteristic information of the truck group includes driving route, speed distribution, and stop time.

9. The classification and grading control system for overloaded trucks based on big data portrait technology as claimed in claim 8 is characterized in that: The five modal groups in A1 include: basic vehicle attribute data, transportation behavior characteristic data, violation record data, weather condition data and cargo information; the basic vehicle attribute data includes vehicle model, age, load tonnage, and company to which it belongs; the transportation behavior characteristic data includes transportation route, transportation frequency, commonly loaded cargo type, and average driving speed; the violation record data includes the number of times, locations, and severity of overloading; the weather condition data includes rainy days and strong winds; and the cargo information includes special cargo attributes.

Images