High-speed toll station passing efficiency evaluation system and method based on multi-source time sequence data

Abstract

The invention relates to the technical field of intelligent traffic, and discloses a highway toll station traffic efficiency evaluation system and method based on multi-source time sequence data, and the system comprises a data analysis module, a preprocessing module, a feature processing module, a dual-model evaluation module and a grading evaluation module. The method is applied to the system. The method focuses on expressway toll station passing efficiency evaluation to form a multi-source time sequence data driven dual-index evaluation system, realizes real-time accurate evaluation of the toll station operation state, provides scientific and datamation decision basis for operation management, and effectively solves the problems of insufficient data fusion, monotonous evaluation standard and the like of traditional manual monitoring.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and specifically to a system and method for evaluating the traffic efficiency of a highway toll station based on multi-source time series data. Background Art

[0002] With the continued growth of traffic volume and the increasing complexity of road networks, high-density traffic and the enormous demands of network operations and management are placing higher demands on expressway toll collection operators and managers. Efficiently monitoring the operating status of toll stations and establishing an objective and accurate traffic evaluation system have become key to driving the digital transformation of road network operations and improving traffic efficiency and service quality.

[0003] At present, traditional methods of evaluating the operating status of toll stations mainly rely on the construction of monitoring centers by road section management units and manual monitoring tasks, which have significant technical bottlenecks: First, the data fusion capability is insufficient, relying only on a single video data source or individual auxiliary data of the road section, making it difficult to effectively integrate multi-dimensional information such as equipment status, traffic data, and transaction records, and unable to fully reflect the actual operation of the toll station; second, the evaluation criteria lack objectivity, over-reliance on manual experience and judgment, and the failure to establish a quantitative and scientific scoring system, resulting in subjective bias in the evaluation results; third, the monitoring timeliness and coverage are limited, making it difficult to achieve 7×24 hours of real-time monitoring of all scenarios, and the ability to trace back and analyze historical operation data is weak, which cannot meet the needs of rapid fault diagnosis and long-term operation optimization. Therefore, there is an urgent need for a technology that can realize real-time and accurate evaluation of the operating status of toll stations. Summary of the Invention

[0004] The purpose of this application is to provide a high-speed toll station traffic efficiency evaluation system and method based on multi-source time series data to solve the technical problems raised in the above background technology.

[0005] To achieve the above objectives, this application discloses the following technical solutions: In a first aspect, the present application discloses a high-speed toll station traffic efficiency evaluation system based on multi-source time series data, the system comprising: The data analysis module is configured to collect multi-dimensional time series data from multiple toll stations, and transmit, analyze, and monitor the quality of the collected time series data to obtain standardized structured data; A preprocessing module is configured to: perform timestamp alignment and device operation status conversion processing on the structured data to obtain structured time series data with device operation status identification; A feature processing module is configured to: extract a pressure parameter feature vector and a service parameter feature vector based on the structured time series data; The dual-model evaluation module is configured to: input the pressure parameter feature vector and the service parameter feature vector respectively based on the constructed pressure index model and service index model, and output a pressure index prediction value and a service index prediction value; The hierarchical evaluation module is configured to: output a corresponding traffic pressure evaluation result based on the pressure index prediction value according to a preset hierarchical mapping relationship, and output a corresponding service quality evaluation result based on the service index prediction value.

[0006] Preferably, the data parsing module includes: a data transmission service unit, configured to: transmit the multi-dimensional time series data through a hybrid transmission processing mechanism, wherein the hybrid transmission processing mechanism includes message queue asynchronous transmission and streaming data real-time processing, and the multi-dimensional time series data is unstructured or semi-structured data; The file parsing service unit is configured to: convert the multi-dimensional time series data into structured data through regular expressions and heterogeneous data format parsing tools; The data quality monitoring unit is configured to: evaluate the integrity of the structured data through a data volume statistical monitoring benchmark model and non-empty rate analysis of key business fields, evaluate the accuracy of the structured data through a business rule engine and an outlier detection algorithm, evaluate the consistency of the structured data through a cross-data source alignment verification mechanism, evaluate the timeliness of the structured data through an end-to-end delay indicator and a graded alarm mechanism, and evaluate the effectiveness of the structured data through data usage frequency and demand matching analysis.

[0007] Preferably, the multi-dimensional time series data includes vehicle transaction flow, real-time vehicle monitoring data, equipment monitoring heartbeat data and lane log data, wherein: The vehicle transaction flow includes lane code, universal unique identifier, transaction time, license plate number, toll vehicle type, entrance information, door frame identification and exit information; The real-time vehicle monitoring data includes lane code, license plate number, lane entry time, lane exit time, vehicle passing time, average vehicle speed, transaction time and special situation type; The device monitoring heartbeat data includes lane code, upload time, storage time, device ID, status code, response delay, and fault flag; The lane log data includes lane code, software version, upload time, storage time, data version, special situation number and special situation type.

[0008] Preferably, the preprocessing module includes: The timestamp alignment unit is configured to align all data sources using a time window of S seconds and fill missing values ​​with regular linear interpolation. The filling formula is: in, is the missing value of the target time point t to be filled, is the observation value at the previous valid time point t-1, is the observation value at the next valid time point t+1, is the time interval between the effective time point t+1 and the effective time point t-1, is the time interval between the target time point t and the effective time point t-1, and ; The device status conversion unit is configured to convert the device heartbeat data into device operating status data. If a device has fewer than a dynamic threshold x heartbeat data records within a time window of S seconds, it is marked as a device failure. The conversion formula is: in, The device fault marking result. When the value is 1, it means the device is judged to be in a fault state. When the value is 0, it means the device is judged to be in a normal state. The number of valid heartbeat data recorded at time point t. When the value is 1, it means that there is a valid heartbeat record at the corresponding time point. When the value is 0, there is no valid heartbeat record at the corresponding time point. It is the total number of valid heartbeat data records within the time window of S seconds.

[0009] Preferably, the calculation formula of the pressure index model is: in, is the predicted pressure index value of the i-th sample output by the model at time t; is the dimension of the pressure parameter eigenvector, is the value of the kth pressure parameter eigenvector, is the kth pressure parameter eigenvector weight, is the pressure bias term; The calculation formula of the service index model is: in, is the predicted service index value of the i-th sample output by the model at time t; is the dimension of the service parameter feature vector, is the value of the kth service parameter feature vector, is the weight of the kth service parameter feature vector, This is the service bias item.

[0010] Preferably, the pressure parameter characteristic vector Including the idle time of lane i in period t , Idle times , real-time traffic flow , traffic flow saturation , total number of devices , Number of devices with abnormal heartbeats , equipment failure ratio , Failure rate sliding window statistics and moving average of traffic flow index .

[0011] Preferably, the service parameter feature vector Including the actual average travel time of lane i in time period t , theoretical benchmark travel time , travel time deviation , vehicle speed fluctuation rate , continuous fault-free time , transaction success rate smoothing value and the efficiency ratio of lane charging mode of toll station j in time period t .

[0012] Preferably, the dual-model evaluation module is further configured to: perform model training and optimization using mean square error and L2 regularization as loss functions, and update model parameters using back propagation algorithm and gradient descent algorithm; wherein: The loss function of the pressure index model is: The loss function of the service index model is: The formula for updating model parameters is: in, is the training sample size; is the labeled pressure index value of the i-th sample in the training dataset at time t; is the label service index value of the i-th sample in the training data set at time t; is the regularization parameter of the pressure index model; is the regularization parameter of the service index model; is the learning rate; is the loss function right gradient; is the loss function right gradient.

[0013] Preferably, the hierarchical mapping relationship includes: When the pressure index prediction value is in the range of [0,5), the corresponding traffic pressure evaluation result is unimpeded; when the pressure index prediction value is in the range of [5,7), the corresponding traffic pressure evaluation result is basically unimpeded; when the pressure index prediction value is in the range of [7,8), the corresponding traffic pressure evaluation result is slightly congested; when the pressure index prediction value is in the range of [8,9), the corresponding traffic pressure evaluation result is moderately congested; when the pressure index prediction value is in the range of [9,10], the corresponding traffic pressure evaluation result is severely congested. When the service index prediction value is in the range of [0,2], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (2,4], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (4,6], the corresponding service quality evaluation result is fair; when the service index prediction value is in the range of (6,8], the corresponding service quality evaluation result is good; when the service index prediction value is in the range of (8,10], the corresponding service quality evaluation result is excellent.

[0014] In a second aspect, the present application discloses a method for evaluating the traffic efficiency of a highway toll station based on multi-source time series data, which is applied to the above-mentioned system for evaluating the traffic efficiency of a highway toll station based on multi-source time series data. The method comprises: Data analysis step: Collect multi-dimensional time series data from multiple toll booths, and transmit, analyze, and monitor the quality of the collected time series data to obtain standardized structured data; Preprocessing step: performing timestamp alignment and device operation status conversion processing on the structured data to obtain structured time series data with device operation status identification; Feature processing step: extracting pressure parameter feature vectors and service parameter feature vectors based on the structured time series data; Dual-model evaluation step: based on the constructed pressure index model and service index model, respectively input the pressure parameter feature vector and the service parameter feature vector, and output the pressure index prediction value and the service index prediction value; Hierarchical evaluation step: according to a preset hierarchical mapping relationship, a corresponding traffic pressure evaluation result is output based on the pressure index prediction value, and a corresponding service quality evaluation result is output based on the service index prediction value.

[0015] Beneficial effects: The highway toll station traffic efficiency evaluation system and method based on multi-source time series data of the present application collects multi-dimensional time series data from multiple toll stations, and obtains standardized structured data through transmission, analysis and quality monitoring, which can comprehensively and truly reflect comprehensive operating information such as toll station equipment status, traffic flow changes, and transaction status; secondly, by extracting pressure and service parameter feature vectors and combining dual-model evaluation to output quantitative pressure index and service index, and then outputting evaluation results through hierarchical mapping relationship, a scientific and objective quantitative evaluation system is constructed to reduce evaluation bias; at the same time, based on real-time data processing, timestamp alignment processing and real-time output capabilities, full-scene real-time monitoring can be achieved, and the standardized structured data facilitates historical operation data retrospective analysis, meeting the needs of rapid fault diagnosis and long-term operation optimization, and improving the coverage and depth of monitoring; in addition, through the output pressure index prediction value, service index prediction value and performance level evaluation results, accurate and dynamic evaluation results are provided to highway operation and management units, promoting the transformation of road network operation services from manual monitoring to data-based and intelligent monitoring, and providing scientific decision-making support for improving traffic efficiency and service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 A structural block diagram of a high-speed toll station traffic efficiency evaluation system based on multi-source time series data provided in an embodiment of the present application; Figure 2 A flowchart of a method for evaluating the traffic efficiency of a high-speed toll station based on multi-source time series data provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following is a clear and complete description of the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0019] In this document, the term "comprising" is intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0020] In a first aspect, this embodiment provides a Figure 1 The highway toll station traffic efficiency evaluation system based on multi-source time series data shown in the figure includes: The data analysis module is configured to collect multi-dimensional time series data from multiple toll stations, and transmit, analyze, and monitor the quality of the collected time series data to obtain standardized structured data; A preprocessing module is configured to: perform timestamp alignment and device operation status conversion processing on the structured data to obtain structured time series data with device operation status identification; A feature processing module is configured to: extract a pressure parameter feature vector and a service parameter feature vector based on the structured time series data; The dual-model evaluation module is configured to: input the pressure parameter feature vector and the service parameter feature vector respectively based on the constructed pressure index model and service index model, and output a pressure index prediction value and a service index prediction value; The hierarchical evaluation module is configured to: output a corresponding traffic pressure evaluation result based on the pressure index prediction value according to a preset hierarchical mapping relationship, and output a corresponding service quality evaluation result based on the service index prediction value.

[0021] Based on this, the highway toll station traffic efficiency evaluation system based on multi-source time series data of this embodiment integrates multi-dimensional time series data through the data analysis module, solves the limitation of traditional evaluation relying on a single data source, and realizes the comprehensiveness of data; ensures the time consistency and integrity of data through the preprocessing module, laying the foundation for subsequent analysis; outputs quantitative indexes through the feature processing module and the dual-model evaluation module, replaces traditional manual experience judgment, and constructs an objective dual-indicator evaluation system; finally, outputs intuitive evaluation results through the hierarchical evaluation module, realizes real-time and accurate evaluation of the operating status of the toll station, effectively solves the problems of insufficient data integration, subjective evaluation, and poor timeliness of traditional evaluation, and provides data-based decision support for operation management.

[0022] In this embodiment, the data parsing module includes: a data transmission service unit, configured to: transmit the multi-dimensional time series data through a hybrid transmission processing mechanism, wherein the hybrid transmission processing mechanism includes message queue asynchronous transmission and streaming data real-time processing (such as Apache Flink), and the multi-dimensional time series data is unstructured or semi-structured data; A file parsing service unit is configured to: convert the multi-dimensional time series data into structured data using a regular expression and a heterogeneous data format parsing tool; wherein the heterogeneous data format parsing tool includes an XML parser or a JSON parser; The data quality monitoring unit is configured to: evaluate the integrity of the structured data through a data volume statistical monitoring benchmark model and non-empty rate analysis of key business fields, evaluate the accuracy of the structured data through a business rule engine and an outlier detection algorithm, evaluate the consistency of the structured data through a cross-data source alignment verification mechanism, evaluate the timeliness of the structured data through an end-to-end delay indicator and a graded alarm mechanism, and evaluate the effectiveness of the structured data through data usage frequency and demand matching analysis.

[0023] Through the setting of the above-mentioned data parsing module, the three units of the data parsing module work together: the data transmission service unit adopts a hybrid transmission mechanism, which not only ensures the stability of unstructured / semi-structured data transmission, but also realizes real-time processing, solving the problem of delay or instability of traditional data transmission; the file parsing service unit standardizes data through regular expressions and XML / JSON parsers, providing a unified format for subsequent processing, overcoming the defects of traditional data formats that are chaotic and difficult to integrate; the data quality monitoring unit evaluates data quality from five dimensions to ensure that the data is complete, accurate, consistent, timely and effective, avoiding evaluation deviations caused by data quality issues, and providing data guarantee for the reliability of the entire system.

[0024] It is feasible that the multi-dimensional time series data includes vehicle transaction flow, real-time vehicle monitoring data, equipment monitoring heartbeat data and lane log data, wherein: The vehicle transaction flow includes lane code, universal unique identifier, transaction time, license plate number, toll vehicle type, entrance information, door frame identification and exit information; The real-time vehicle monitoring data includes lane code, license plate number, lane entry time, lane exit time, vehicle passing time, average vehicle speed, transaction time and special situation type; The device monitoring heartbeat data includes lane code, upload time, storage time, device ID, status code, response delay, and fault flag; The lane log data includes lane code, software version, upload time, storage time, data version, special situation number and special situation type.

[0025] Through the above-mentioned design of multi-dimensional time series data, the types and key fields of multi-dimensional time series data are clarified, the data sources and core information are accurately defined, the comprehensiveness and pertinence of data collection are ensured, and reliable and standardized original data are provided for subsequent data analysis, preprocessing and feature extraction, avoiding analysis deviations caused by missing or unclear data fields, and further ensuring the effectiveness of the entire evaluation system.

[0026] In this embodiment, the preprocessing module includes: The timestamp alignment unit is configured to align all data sources using a time window of S seconds and fill missing values ​​with regular linear interpolation. The filling formula is: in, is the missing value of the target time point t to be filled, is the observation value at the previous valid time point t-1, is the observation value at the next valid time point t+1, is the time interval between the effective time point t+1 and the effective time point t-1, is the time interval between the target time point t and the effective time point t-1, and ; The device status conversion unit is configured to convert the device heartbeat data into device operating status data. If a device has fewer than a dynamic threshold x heartbeat data records within a time window of S seconds, it is marked as a device failure. The conversion formula is: in, The device fault marking result. When the value is 1, it means the device is judged to be in a fault state. When the value is 0, it means the device is judged to be in a normal state. The number of valid heartbeat data recorded at time point t. When the value is 1, it means that there is a valid heartbeat record at the corresponding time point. When the value is 0, there is no valid heartbeat record at the corresponding time point. It is the total number of valid heartbeat data records within the time window of S seconds.

[0027] Based on the settings of the above-mentioned preprocessing module, the timestamp alignment unit aligns data with a time window of S seconds, fills missing values ​​with a linear interpolation formula, and uses the valid observations before and after the target time point to estimate missing values ​​through the time interval ratio, ensuring the time consistency of different data sources, filling the data gaps, and avoiding analysis errors caused by time dislocation or data gaps; the device status conversion unit reflects the device activity by comparing the heartbeat frequency with the dynamic threshold, accurately identifies the device fault status, and provides a reliable device status basis for subsequent pressure and service index analysis, solving the problem of fuzzy traditional device status judgment.

[0028] In this embodiment, the calculation formula of the pressure index model is: in, is the predicted pressure index value of the i-th sample output by the model at time t; is the dimension of the pressure parameter eigenvector, is the value of the kth pressure parameter eigenvector, is the kth pressure parameter eigenvector weight, is the pressure bias term.

[0029] The calculation formula of the service index model is: in, is the predicted service index value of the i-th sample output by the model at time t; is the dimension of the service parameter feature vector, is the value of the kth service parameter feature vector, is the weight of the kth service parameter feature vector, This is the service bias item.

[0030] Based on the above-mentioned pressure index model and service index model, by linearly combining feature vectors and quantifying the impact of each feature on pressure / service quality through weights, multi-dimensional features are integrated into a single index, which realizes the transformation of complex multi-dimensional data into intuitive and comparable quantitative indicators, that is, the quantitative evaluation of traffic pressure and service quality is realized, which solves the problem of lack of unified quantitative standards in traditional evaluation and makes the evaluation results more objective and comparable.

[0031] It is feasible that the pressure parameter characteristic vector Including the idle time of lane i in period t , Idle times , real-time traffic flow , traffic flow saturation , total number of devices , Number of devices with abnormal heartbeats , equipment failure ratio , Failure rate sliding window statistics and moving average of traffic flow index .

[0032] Through the design of the above-mentioned pressure parameter characteristic vector, the key factors affecting traffic pressure are comprehensively captured, providing multi-dimensional and targeted input features for the pressure index model, ensuring that the model can accurately characterize the changes in traffic pressure, solving the problem of the single dimension of traditional pressure assessment, and improving the comprehensiveness and accuracy of pressure evaluation.

[0033] Secondly, the service parameter feature vector Including the actual average travel time of lane i in time period t , theoretical benchmark travel time , travel time deviation , vehicle speed fluctuation rate , continuous fault-free time , transaction success rate smoothing value and the efficiency ratio of lane charging mode of toll station j in time period t , where lane charging methods include ETC or MTC.

[0034] Through the design of the above-mentioned service parameter feature vector, the core indicators of service quality are covered, providing multi-dimensional feature support for the service index model, enabling the model to accurately quantify service quality, solving the one-sided problem of traditional service evaluation indicators, and improving the accuracy of service quality evaluation.

[0035] Furthermore, the dual-model evaluation module is further configured to: perform model training and optimization using mean square error and L2 regularization as loss functions, and update model parameters using back propagation algorithm and gradient descent algorithm; wherein: The loss function of the pressure index model is: The loss function of the service index model is: The formula for updating model parameters is: in, is the number of training samples (i.e. the total number of samples participating in model training); is the labeled pressure index value of the i-th sample in the training dataset at time t (i.e., the actual observed pressure index); is the labeled service index value of the i-th sample in the training dataset at time t (the actual observed service index); is the regularization parameter of the pressure index model, which is used to reduce the complexity of the model and prevent overfitting; It is the regularization parameter of the service index model, which is used to reduce the complexity of the model and prevent overfitting; and Used for L2 regularization calculation; is the learning rate, which is used to control the step size of parameter updates in the gradient descent algorithm; is the loss function Yes gradient, is the loss function pair The gradient of the parameter and The trend and degree of impact of small changes on the loss function value.

[0036] Based on the above, the loss function for model training and optimization uses mean square error and L2 regularization. The mean square error measures the error by calculating the square difference between the predicted value and the true value to ensure the accuracy of the model prediction; L2 regularization limits the parameters from being too large by weighting the sum of squares of the parameters to prevent the model from overfitting; the parameter update formula iteratively optimizes the parameters in the opposite direction of the loss function gradient to minimize the loss; thereby improving the model's prediction accuracy and generalization ability, avoiding the actual application deviation caused by the model overfitting the training data, and ensuring the stable output of the dual model in different scenarios.

[0037] In this embodiment, the hierarchical mapping relationship includes: When the pressure index prediction value is in the range of [0,5), the corresponding traffic pressure evaluation result is unimpeded; when the pressure index prediction value is in the range of [5,7), the corresponding traffic pressure evaluation result is basically unimpeded; when the pressure index prediction value is in the range of [7,8), the corresponding traffic pressure evaluation result is slightly congested; when the pressure index prediction value is in the range of [8,9), the corresponding traffic pressure evaluation result is moderately congested; when the pressure index prediction value is in the range of [9,10], the corresponding traffic pressure evaluation result is severely congested. When the service index prediction value is in the range of [0,2], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (2,4], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (4,6], the corresponding service quality evaluation result is fair; when the service index prediction value is in the range of (6,8], the corresponding service quality evaluation result is good; when the service index prediction value is in the range of (8,10], the corresponding service quality evaluation result is excellent.

[0038] Based on the design of the above-mentioned hierarchical mapping relationship, the quantitative index is mapped to a specific level through a preset interval, so that the abstract quantitative index can be converted into an easy-to-understand evaluation conclusion, enabling operators to quickly grasp the traffic pressure and service quality status of the toll station, solving the problem that traditional evaluation results are obscure and difficult to directly apply, and improving decision-making efficiency.

[0039] In summary, the highway toll station traffic efficiency evaluation system based on multi-source time series data in this embodiment focuses on the highway toll station traffic efficiency evaluation, and uses a dual-indicator evaluation system driven by multi-source time series data to achieve real-time and accurate evaluation of the toll station operating status, providing a scientific and data-based decision-making basis for operation management, and effectively solving problems such as insufficient data fusion and monotonous evaluation standards in traditional manual monitoring.

[0040] In a second aspect, this embodiment provides a Figure 2 The method for evaluating the traffic efficiency of a toll station based on multi-source time series data is applied to the above-mentioned system for evaluating the traffic efficiency of a toll station based on multi-source time series data. The method includes the following steps: Through the processor architecture composed of ARM processor and FPGA module, the ARM processor performs system management, edge computing task scheduling and northbound interface docking, while the FPGA module performs protocol conversion and real-time data stream processing; In the protocol conversion of the FPGA module, different IoT protocols are efficiently converted through parallel processing. The IoT protocols include several of Zigbee, LoRa, Modbus, MQTT, CoAP and Matter protocols. A hybrid task scheduling algorithm based on genetic algorithm and Petri net is used to dynamically allocate cross-protocol task resources and optimize task execution paths; Use a unified security framework module that integrates protocol-independent key management, cross-protocol security mechanism integration, and zero-trust architecture to unify the management of symmetric and asymmetric encryption; Based on the edge computing module, local data analysis is performed and task priorities are dynamically adjusted based on the analysis results.

[0041] It should be noted that the highway toll station traffic efficiency evaluation method based on multi-source time series data of this embodiment corresponds to the aforementioned highway toll station traffic efficiency evaluation system based on multi-source time series data. Therefore, the parts that are not specifically described in the highway toll station traffic efficiency evaluation method based on multi-source time series data of this embodiment (including but not limited to specific implementation technical means and technical effects) can be referred to the relevant records in the aforementioned highway toll station traffic efficiency evaluation system based on multi-source time series data, and this text will not go into details here.

[0042] In the embodiments provided herein, it should be understood that the embodiments described herein can be implemented using hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA module), a processor, a controller, a microcontroller, a microprocessor, or other electronic units designed to implement the functionality described herein, or any combination thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the relevant hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media includes any medium that facilitates the transmission of a computer program from one location to another. The storage medium may be any available medium that can be accessed by a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing the desired program code in the form of instructions or data structures and accessible by a computer.

[0043] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A high-speed toll station traffic efficiency evaluation system based on multi-source time series data, characterized by: The system includes: The data analysis module is configured to collect multi-dimensional time series data from multiple toll stations, and transmit, analyze, and monitor the quality of the collected time series data to obtain standardized structured data; A preprocessing module is configured to: perform timestamp alignment and device operation status conversion processing on the structured data to obtain structured time series data with device operation status identification; A feature processing module is configured to: extract a pressure parameter feature vector and a service parameter feature vector based on the structured time series data; The dual-model evaluation module is configured to: input the pressure parameter feature vector and the service parameter feature vector respectively based on the constructed pressure index model and service index model, and output a pressure index prediction value and a service index prediction value; The hierarchical evaluation module is configured to: output a corresponding traffic pressure evaluation result based on the pressure index prediction value according to a preset hierarchical mapping relationship, and output a corresponding service quality evaluation result based on the service index prediction value.

2. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 1 is characterized in that: The data parsing module includes: a data transmission service unit, configured to: transmit the multi-dimensional time series data through a hybrid transmission processing mechanism, wherein the hybrid transmission processing mechanism includes message queue asynchronous transmission and streaming data real-time processing, and the multi-dimensional time series data is unstructured or semi-structured data; The file parsing service unit is configured to: convert the multi-dimensional time series data into structured data through regular expressions and heterogeneous data format parsing tools; The data quality monitoring unit is configured to: evaluate the integrity of the structured data through a data volume statistical monitoring benchmark model and non-empty rate analysis of key business fields, evaluate the accuracy of the structured data through a business rule engine and an outlier detection algorithm, evaluate the consistency of the structured data through a cross-data source alignment verification mechanism, evaluate the timeliness of the structured data through an end-to-end delay indicator and a graded alarm mechanism, and evaluate the effectiveness of the structured data through data usage frequency and demand matching analysis.

3. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 2 is characterized in that: The multi-dimensional time series data includes vehicle transaction flow, real-time vehicle monitoring data, equipment monitoring heartbeat data, and lane log data, among which: The vehicle transaction flow includes lane code, universal unique identifier, transaction time, license plate number, toll vehicle type, entrance information, door frame identification and exit information; The real-time vehicle monitoring data includes lane code, license plate number, lane entry time, lane exit time, vehicle passing time, average vehicle speed, transaction time and special situation type; The device monitoring heartbeat data includes lane code, upload time, storage time, device ID, status code, response delay, and fault flag; The lane log data includes lane code, software version, upload time, storage time, data version, special situation number and special situation type.

4. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 1 is characterized in that: The pre-processing module comprises: The timestamp alignment unit is configured to align all data sources using a time window of S seconds and fill missing values ​​with regular linear interpolation. The filling formula is: in, is the missing value of the target time point t to be filled, is the observation value at the previous valid time point t-1, is the observation value at the next valid time point t+1, is the time interval between the effective time point t+1 and the effective time point t-1, is the time interval between the target time point t and the effective time point t-1, and ; The device status conversion unit is configured to convert the device heartbeat data into device operating status data. If a device has fewer than a dynamic threshold x heartbeat data records within a time window of S seconds, it is marked as a device failure. The conversion formula is: in, The device fault marking result. When the value is 1, it means the device is judged to be in a fault state. When the value is 0, it means the device is judged to be in a normal state. The number of valid heartbeat data recorded at time point t. When the value is 1, it means that there is a valid heartbeat record at the corresponding time point. When the value is 0, there is no valid heartbeat record at the corresponding time point. It is the total number of valid heartbeat data records within the time window of S seconds.

5. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 1 is characterized in that: The calculation formula of the pressure index model is: in, is the predicted pressure index value of the i-th sample output by the model at time t; is the dimension of the pressure parameter eigenvector, is the value of the kth pressure parameter eigenvector, is the kth pressure parameter eigenvector weight, is the pressure bias term; The calculation formula of the service index model is: in, is the predicted service index value of the i-th sample output by the model at time t; is the dimension of the service parameter feature vector, is the value of the kth service parameter feature vector, is the weight of the kth service parameter feature vector, This is the service bias item.

6. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 5 is characterized in that: The pressure parameter characteristic vector Including the idle time of lane i in period t , Idle times , real-time traffic flow , traffic flow saturation , Total number of devices , Number of devices with abnormal heartbeats , equipment failure ratio , Failure rate sliding window statistics and moving average of traffic flow index .

7. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 5 is characterized in that: The service parameter feature vector Including the actual average travel time of lane i in time period t , theoretical benchmark travel time , travel time deviation , vehicle speed fluctuation rate , continuous fault-free time , transaction success rate smoothing value and the efficiency ratio of lane charging mode of toll station j in time period t .

8. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 5 is characterized in that: The dual-model evaluation module is further configured to: perform model training and optimization using mean square error and L2 regularization as loss functions, and update model parameters using back propagation algorithm and gradient descent algorithm; wherein: The loss function of the pressure index model is: The loss function of the service index model is: The formula for updating model parameters is: in, is the number of training samples; is the labeled pressure index value of the i-th sample in the training dataset at time t; is the label service index value of the i-th sample in the training data set at time t; is the regularization parameter of the pressure index model; is the regularization parameter of the service index model; is the learning rate; is the loss function right gradient; is the loss function right gradient.

9. The high-speed toll station traffic efficiency evaluation system based on multi-source time series data according to claim 1 is characterized in that: The hierarchical mapping relationship includes: When the pressure index prediction value is in the range of [0,5), the corresponding traffic pressure evaluation result is unimpeded; when the pressure index prediction value is in the range of [5,7), the corresponding traffic pressure evaluation result is basically unimpeded; when the pressure index prediction value is in the range of [7,8), the corresponding traffic pressure evaluation result is slightly congested; when the pressure index prediction value is in the range of [8,9), the corresponding traffic pressure evaluation result is moderately congested; when the pressure index prediction value is in the range of [9,10], the corresponding traffic pressure evaluation result is severely congested. When the service index prediction value is in the range of [0,2], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (2,4], the corresponding service quality evaluation result is poor; when the service index prediction value is in the range of (4,6], the corresponding service quality evaluation result is fair; when the service index prediction value is in the range of (6,8], the corresponding service quality evaluation result is good; when the service index prediction value is in the range of (8,10], the corresponding service quality evaluation result is excellent.

10. A method for evaluating the traffic efficiency of a highway toll station based on multi-source time series data, applied to the system for evaluating the traffic efficiency of a highway toll station based on multi-source time series data as claimed in any one of claims 1 to 9, characterized in that: The method includes: Data analysis step: Collect multi-dimensional time series data from multiple toll booths, and transmit, analyze, and monitor the quality of the collected time series data to obtain standardized structured data; Preprocessing step: performing timestamp alignment and device operation status conversion processing on the structured data to obtain structured time series data with device operation status identification; Feature processing step: extracting pressure parameter feature vectors and service parameter feature vectors based on the structured time series data; Dual-model evaluation step: based on the constructed pressure index model and service index model, respectively input the pressure parameter feature vector and the service parameter feature vector, and output the pressure index prediction value and the service index prediction value; Hierarchical evaluation step: according to a preset hierarchical mapping relationship, a corresponding traffic pressure evaluation result is output based on the pressure index prediction value, and a corresponding service quality evaluation result is output based on the service index prediction value.

Images