A dynamic traffic optimization management method, device, equipment and medium

Abstract

The present application relates to the technical field of traffic informationization. Specifically relates to a kind of dynamic traffic optimization management method, device, equipment and medium, its method includes using sensing credibility information to the traffic deviation degree information is weighted fusion processing, obtains fusion deviation degree information, obtains candidate signal control scheme set according to fusion deviation degree information and preset phase configuration record information, for each candidate signal control scheme in candidate signal control scheme set, according to traffic state alignment sequence and the timing parameter information corresponding to candidate signal control scheme Queuing Growth Estimation processing is executed, obtain scheme evaluation information, according to scheme evaluation information and preset constraint rule information, determine target signal control scheme from candidate signal control scheme set, according to target signal control scheme to target intersection executes signal timing control, obtains target timing parameter information.The present application has the effect of improving traffic operation state analysis accuracy and control strategy adaptability in traffic management process.

Description

Technical Field

[0001] This invention relates to the technical field of traffic information technology, and in particular to a dynamic traffic optimization management method, device, equipment, and medium. Background Technology

[0002] With the increase in urban motor vehicle ownership and the increasing complexity of road network structures, intersections have become a key bottleneck in traffic operation. Signal control, as the main means of organizing traffic at intersections, is widely used to improve traffic efficiency and reduce the risk of congestion spreading. Existing intelligent traffic management solutions typically rely on traffic detection equipment, video monitoring, or floating cars to obtain road network operation data and adjust intersection signal timing parameters accordingly to adapt to changes in traffic demand at different times and in different directions.

[0003] However, existing technologies still face several challenges in engineering implementation. First, traffic data from different sources vary in sampling frequency, timestamp accuracy, transmission delay, and missing noise. Without unified data quality verification and timing consistency processing, biased judgments of intersection operational status can easily occur, affecting the stability of signal control adjustments. Second, some solutions rely on a limited number of indicators or fixed thresholds for status determination and strategy selection, making it difficult to adapt to natural fluctuations and sudden disturbances in the road network under different scenarios. This can lead to "over-adjustment" or "lagging" of adjustment strategies. Third, some solutions primarily use rule-driven or experience-based parameter settings in timing decisions, lacking sufficient assessment of post-adjustment queuing evolution trends and intersection saturation risks. This may cause queue overflow, increased upstream and downstream interference, and negatively impact the effectiveness of intersection coordinated control.

[0004] Therefore, for intersection signal control under multi-source data conditions, how to form a more robust judgment of the operating status while ensuring data reliability, and improve the adaptability to queuing and congestion evolution in the process of strategy selection and timing adjustment, remains a direction that existing intelligent traffic management technologies need to further improve. Summary of the Invention

[0005] To improve the accuracy of traffic operation status analysis and the adaptability of control strategies in the traffic management process, this application provides a dynamic traffic optimization management method, device, equipment and medium.

[0006] The above-mentioned objective of this application is achieved through the following technical solution: A dynamic traffic optimization management method, the dynamic traffic optimization management method comprising: Traffic status information of the target intersection and adjacent road segments is acquired, and the traffic status information is time-aligned to obtain a traffic status aligned sequence. Acquire historical traffic status information corresponding to the target intersection, perform statistical calculations on the historical traffic status information according to the preset statistical window, and obtain historical baseline statistical parameter information. Based on the traffic status alignment sequence and historical baseline statistical parameters, the traffic deviation information is calculated according to the preset deviation calculation rules. Based on the missing rate, fluctuation amplitude and cross-source consistency in the traffic state alignment sequence, sensor credibility information is generated. The traffic deviation information is then weighted and fused using the sensor credibility information to obtain fused deviation information. Obtain the preset phase configuration record information of the target intersection, and generate a set of candidate signal control schemes based on the fused deviation information and the preset phase configuration record information; For each candidate signal control scheme in the candidate signal control scheme set, queue growth estimation is performed based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme to obtain scheme evaluation information; Based on the scheme evaluation information and preset constraint rule information, the target signal control scheme is determined from the candidate signal control scheme set, and the signal timing control is performed on the target intersection according to the target signal control scheme to obtain the target timing parameter information.

[0007] By adopting the above technical solutions, time alignment processing can be performed on traffic status collection information, mitigating the status judgment bias caused by differences in sampling frequency, inconsistent timestamps, and transmission delays of multi-source data. This improves the stability and consistency of signal timing adjustments. Historical traffic status collection information can be used to form historical baseline statistical parameters, and traffic deviation information can be obtained according to deviation calculation rules. This makes the quantification of traffic status changes closer to the normal level of different time periods, thereby reducing the risk of misjudgment caused by fixed thresholds. Missing rate, fluctuation amplitude, and cross-source consistency can be introduced to generate sensor reliability information and weighted fusion of traffic deviation information, reducing the impact of missing noise and abnormal fluctuations on decision-making, thereby improving the reliability of strategy selection. Under the constraints of preset phase configuration record information, a set of candidate signal control schemes can be generated and scheme evaluation information can be obtained through queue growth estimation. Then, the target signal control scheme can be determined by combining preset constraint rules, thereby improving the adaptability of timing decisions to queue evolution and congestion spread risks.

[0008] In a preferred embodiment, this application can be further configured as follows: the step of calculating traffic deviation information according to a preset deviation calculation rule based on traffic state alignment sequence and historical baseline statistical parameter information includes; Select traffic status information corresponding to the target sampling time from the traffic status alignment sequence; Select the historical baseline statistical parameter information corresponding to the target sampling time from the historical baseline statistical parameter information; Based on the deviation calculation rule information, the deviation calculation process is performed on the traffic state collection information and the historical baseline statistical parameter information to obtain the single index deviation sequence. Based on the deviation calculation rule information, normalization and aggregation processing is performed on the single indicator deviation sequence to obtain the passage deviation information.

[0009] By adopting the above technical solution, traffic state collection information and historical baseline statistical parameters can be selected correspondingly at the target sampling time, so that the observed value and the reference value at the same time are consistent, thereby reducing the impact of sampling frequency differences and time alignment errors on deviation calculation. According to the deviation calculation rules, the traffic state collection information and historical baseline statistical parameters can be processed to calculate the deviation, and the change of traffic state relative to the baseline can be transformed into a single-index deviation sequence, thereby improving the consistency of quantitative expression of traffic state changes. The single-index deviation sequence can be normalized and converged to output traffic deviation information under a unified scale, thereby reducing the comparison bias caused by different dimensions or different value ranges, which facilitates subsequent continuous analysis and control decisions of traffic state.

[0010] In a preferred embodiment, this application can be further configured as follows: the step of generating sensor reliability information based on the missing rate, fluctuation amplitude, and cross-source consistency in the traffic state alignment sequence, and using the sensor reliability information to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information includes: Within a preset statistical window, the traffic status aligned sequence is statistically processed according to the data source to obtain the missing rate information and fluctuation range information corresponding to each data source. Within a preset statistical window, traffic status information collected from different data sources at the same sampling time is compared for consistency to obtain cross-source consistency information corresponding to each data source. The missing rate information, fluctuation range information and cross-source consistency information are combined and processed to obtain the sensing reliability information corresponding to each data source. The traffic deviation information is weighted and fused using the sensor reliability information to obtain the fused deviation information.

[0011] By adopting the above technical solutions, statistical processing of traffic state alignment sequences according to data sources can be performed within a preset statistical window to obtain missing rate information and fluctuation amplitude information. This enables the data sources to have quantifiable quality characterization in terms of missing data and fluctuation characteristics, thereby reducing the interference caused by missing data and abnormal fluctuations on traffic status analysis. It also enables consistency comparison processing of traffic state collection information from different data sources at the same sampling time to obtain cross-source consistency information, providing a constrained description of the degree of difference between different data sources, thereby reducing the impact of single data source deviation on the judgment results. Furthermore, it enables combined operation processing of missing rate information, fluctuation amplitude information, and cross-source consistency information to obtain sensor credibility information, providing a basis for the weight allocation of each data source, thereby improving the interpretability and stability of the subsequent fusion process. Finally, it enables weighted fusion processing of traffic deviation information using sensor credibility information to obtain fusion deviation information, thereby improving the adaptability of the fusion results to the quality differences of multi-source data.

[0012] In a preferred embodiment, this application can be further configured as follows: obtaining the preset phase configuration record information of the target intersection, and generating a set of candidate signal control schemes based on the fused deviation information and the preset phase configuration record information, includes: Phase sequence information is determined based on preset phase configuration record information; The phase green light ratio information corresponding to the phase sequence information is determined based on the fusion deviation information and the preset phase configuration record information; the phase offset information corresponding to the phase sequence information is determined based on the fusion deviation information and the preset phase configuration record information. The phase sequence information, phase green light ratio information, and phase offset information are combined and processed to obtain a set of candidate signal control schemes.

[0013] By adopting the above technical solutions, phase sequence information can be determined based on preset phase configuration record information, so that the phase organization of signal control meets the existing phase settings and switching constraints of the target intersection, thereby reducing the risk of control conflicts caused by phase setting mismatch. It can determine the phase green light ratio information corresponding to the phase sequence information based on the fusion deviation information, so that the green light time allocation can be adjusted in linkage with the traffic status, thereby reducing the adjustment lag caused by fixed ratio parameters. It can determine the phase offset information by combining the fusion deviation information and the preset phase configuration record information, so that the adjustment of the phase start position within the cycle is subject to the phase configuration constraint, thereby reducing the arbitrariness of offset adjustment. It can combine the phase sequence information, phase green light ratio information and phase offset information to form a set of candidate signal control schemes, thereby providing a comparable scheme space for subsequent scheme evaluation and screening.

[0014] In a preferred embodiment, this application can be further configured as follows: for each candidate signal control scheme in the candidate signal control scheme set, queue growth estimation processing is performed based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme to obtain scheme evaluation information, including: Obtain information from the preset estimation window; Based on the traffic status alignment sequence, the arrival flow rate and initial queue length are extracted within the time range corresponding to the preset estimation window information. The effective release time value for each phase is determined based on the timing parameter information corresponding to the candidate signal control scheme. The number of vehicles that can be released is calculated based on the effective release time value and the preset release capacity parameter information corresponding to the preset estimation window information. Based on the arrival flow rate, initial queue length, and number of vehicles that can be released, a queue recursive calculation is performed to obtain the predicted queue length change sequence. The estimated queue growth is calculated based on the predicted queue length change sequence, and the estimated queue growth is written into the scheme evaluation information.

[0015] By adopting the above technical solutions, it is possible to obtain preset estimation window information and limit the estimation time range, so that the queuing evolution assessment has a consistent time scale, thereby avoiding the incomparability of results caused by different assessment intervals. It is possible to extract the arrival flow value and the initial queue length value from the traffic state alignment sequence within the time range, so that the queue growth estimation processing has an input basis consistent with the on-site state, thereby reducing the deviation caused by relying solely on empirical parameters. It is possible to determine the effective release time value based on the timing parameter information corresponding to the candidate signal control scheme and calculate the number of vehicles that can be released by combining the preset release capacity parameter information, so that the impact of the candidate signal control scheme on the release capacity can be quantitatively expressed, thereby improving the pertinence of the scheme assessment. It is possible to use the arrival flow value, the initial queue length value, and the number of vehicles that can be released to perform queue recursive calculation processing to obtain the predicted queue length change sequence, and further calculate the queue growth estimate value and write it into the scheme assessment information, so that the candidate signal control schemes have a comparable basis in the queue growth dimension, which facilitates subsequent screening decisions.

[0016] In a preferred embodiment, this application can be further configured such that: obtaining the preset estimation window information further includes: Obtain traffic status data corresponding to the preset closed-loop correction window information, perform deviation calculation processing based on the traffic status data and scheme evaluation information, and update historical baseline statistical parameter information and sensor reliability information.

[0017] By adopting the above technical solutions, traffic status information can be acquired within the time range corresponding to the preset closed-loop correction window information, enabling correction processing to be based on actual operational data. This reduces the risk of cumulative deviation caused by long-term fixed parameters. It can also perform deviation calculation processing on traffic status information and scheme evaluation information, quantifying the difference between scheme evaluation results and actual conditions, thus providing a consistent correction basis for parameter updates. Furthermore, it can use the deviation calculation processing results to update historical baseline statistical parameter information, allowing historical baseline statistical parameter information to be adjusted in a rolling manner according to changes in intersection traffic characteristics, thereby reducing the impact of baseline mismatch on traffic deviation information. Finally, it can synchronously update sensor reliability information, allowing the weight allocation of different data sources to be corrected according to changes in missing rate, fluctuation amplitude, and cross-source consistency, thereby maintaining the adaptability of fused deviation information to changes in data quality.

[0018] In a preferred embodiment, this application can be further configured as follows: based on scheme evaluation information and preset constraint rule information, a target signal control scheme is determined from a set of candidate signal control schemes; signal timing control is performed on the target intersection according to the target signal control scheme to obtain target timing parameter information, including: Perform constraint verification processing on the scheme evaluation information and preset constraint rule information corresponding to each candidate signal control scheme in the candidate signal control scheme set to obtain constraint verification result information; The information on the constraint verification results indicates the candidate signal control schemes that have passed the constraint verification, thus obtaining a subset of candidate signal control schemes that have passed the verification. The subset of candidate signal control schemes that have passed the verification are sorted according to the preset sorting rules to obtain the target sequence number information; Select the candidate signal control scheme corresponding to the target sequence number information from the subset of candidate signal control schemes that have passed the verification as the target signal control scheme; write the timing parameter information corresponding to the target signal control scheme into the signal control parameter table to obtain the target timing parameter information.

[0019] By adopting the above technical solution, constraint verification processing can be performed on the scheme evaluation information and preset constraint rule information corresponding to each candidate signal control scheme in the candidate signal control scheme set. This allows candidate signal control schemes that do not meet the constraints such as cycle length, phase organization, or timing boundary to be identified before entering the decision-making process, thereby reducing the risk of candidate signal control schemes that do not meet the constraints entering the execution. Based on the constraint verification results, a subset of candidate signal control schemes that have passed the verification can be selected, making the scope of subsequent sorting processing clearer and improving the standardization of the scheme selection process. The subset of candidate signal control schemes that have passed the verification can be sorted according to the preset sorting rule information to obtain target sequence number information, so that the candidate signal control schemes form a comparable sorting result under the same evaluation caliber, thereby reducing the inconsistency caused by manual experience selection. The candidate signal control scheme corresponding to the target sequence number information can be selected as the target signal control scheme, and the timing parameter information corresponding to the target signal control scheme can be written into the signal control parameter table to form the target timing parameter information, so that the issuance and execution of signal timing control parameters have a clear data carrier and a consistent parameter source.

[0020] The second objective of this invention is achieved through the following technical solution: A dynamic traffic optimization management device, the dynamic traffic optimization management device comprising: The traffic status acquisition and time alignment module is used to acquire traffic status information of the target intersection and adjacent road segments, perform time alignment processing on the traffic status acquisition information, and obtain a traffic status alignment sequence. The historical baseline statistical parameter calculation module is used to acquire historical traffic status information corresponding to the target intersection, perform statistical calculations on the historical traffic status information according to the preset statistical window, and obtain historical baseline statistical parameter information. The traffic deviation calculation module is used to calculate traffic deviation information according to preset deviation calculation rules based on traffic state alignment sequence and historical baseline statistical parameter information. The credibility generation and deviation generation modules are used to generate sensor credibility information based on the missing rate, fluctuation amplitude and cross-source consistency in the traffic state alignment sequence. The sensor credibility information is used to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information. The candidate signal control scheme set generation module is used to obtain the preset phase configuration record information of the target intersection and generate a candidate signal control scheme set based on the fused deviation information and the preset phase configuration record information. The queue growth estimation and scheme evaluation module is used to perform queue growth estimation processing on each candidate signal control scheme in the candidate signal control scheme set based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme, and obtain scheme evaluation information. The signal timing control module is used to determine the target signal control scheme from the candidate signal control scheme set based on the scheme evaluation information and preset constraint rule information, and to perform signal timing control on the target intersection according to the target signal control scheme to obtain the target timing parameter information.

[0021] By adopting the above technical solutions, time alignment processing can be performed on traffic state collection information, mitigating the state judgment bias caused by differences in sampling frequency, inconsistent timestamps, and transmission delays of multi-source data. This improves the stability and consistency of signal timing adjustments. Historical traffic state collection information can be used to form historical baseline statistical parameters, and traffic deviation information can be obtained according to deviation calculation rules. This makes the quantification of traffic state changes closer to the normal levels of different time periods, thereby reducing the risk of misjudgment caused by fixed thresholds. Missing rate, fluctuation amplitude, and cross-source consistency can be introduced to generate sensor reliability information and weighted fusion of traffic deviation information, reducing the impact of missing noise and abnormal fluctuations on decision-making, thus improving the reliability of strategy selection. Under the constraints of preset phase configuration record information, a set of candidate signal control schemes can be generated, and scheme evaluation information can be obtained through queue growth estimation. Then, combined with preset constraint rules, the target signal control scheme can be determined, thereby improving the adaptability of timing decisions to queue evolution and congestion spread risks. The above-mentioned objective three of this application is achieved through the following technical solution: A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned dynamic traffic optimization management method.

[0022] The fourth objective of this application is achieved through the following technical solution: A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned dynamic traffic optimization management method.

[0023] In summary, this application includes at least one of the following beneficial technical effects: 1. It can perform time alignment processing on traffic status collection information, mitigating the status judgment bias caused by differences in sampling frequency, inconsistent timestamps, and transmission delays of multi-source data, thereby improving the stability and consistency of signal timing adjustment. It can use historical traffic status collection information to form historical baseline statistical parameter information and obtain traffic deviation information according to deviation calculation rules, making the quantification of traffic status changes closer to the normal level of different time periods, thereby reducing the risk of misjudgment caused by fixed thresholds. It can introduce missing rate, fluctuation amplitude, and cross-source consistency to generate sensor credibility information and weightedly fuse traffic deviation information, reducing the impact of missing noise and abnormal fluctuations on decision-making, thereby improving the reliability of strategy selection. It can generate a set of candidate signal control schemes under the constraints of preset phase configuration record information and obtain scheme evaluation information through queue growth estimation, and then determine the target signal control scheme by combining preset constraint rules, thereby improving the adaptability of timing decisions to queue evolution and congestion spread risks. Attached Figure Description

[0024] Figure 1 This is a flowchart of a dynamic traffic optimization management method according to an embodiment of this application.

[0025] Figure 2 This is a flowchart illustrating the implementation of step S30 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 3 This is a flowchart illustrating the implementation of step S40 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 4 This is a flowchart illustrating the implementation of step S50 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 5 This is a flowchart illustrating the implementation of step S60 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 6 This is a flowchart illustrating the implementation of step S601 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 7 This is a flowchart illustrating the implementation of step S70 in a dynamic traffic optimization management method according to an embodiment of this application. Figure 8 This is a schematic diagram of a dynamic traffic optimization management device according to one embodiment of this application. Detailed Implementation

[0026] The present application will be further described in detail below with reference to the accompanying drawings.

[0027] In one embodiment, such as Figure 1 As shown, this application discloses a dynamic traffic optimization management method, which specifically includes the following steps: S10: Obtain traffic status information for the target intersection and adjacent road segments, perform time alignment processing on the traffic status information, and obtain a traffic status aligned sequence.

[0028] In this embodiment, traffic status acquisition information refers to data records related to road traffic status formed at the target intersection and adjacent road segments. Traffic status acquisition information includes acquisition time information and traffic status numerical information. Acquisition time information refers to the timestamp corresponding to the traffic status numerical information. Traffic status numerical information refers to the traffic status quantity value formed at the time corresponding to the acquisition time information. Time alignment processing refers to the process of converting traffic status numerical information corresponding to different acquisition time information to a unified time reference. Alignment time axis refers to the time sequence generated according to a preset sampling period. Alignment time refers to any time point on the alignment time axis. Alignment assignment processing refers to the process of assigning values ​​to match traffic status numerical information to the alignment time. Traffic status alignment sequence refers to the sequence of traffic status numerical information arranged in chronological order according to the alignment time axis.

[0029] Specifically, when acquiring traffic status information for the target intersection and adjacent road segments, traffic status information is read from the corresponding acquisition terminals of the target intersection and adjacent road segments, and the consistency of the acquisition time information format within the traffic status information is verified. Then, an aligned time axis covering the target statistical period is generated according to a preset sampling period. The acquisition time information of each traffic status acquisition piece within the target statistical period is then mapped to the aligned time axis. Alignment assignment processing is performed for each alignment moment. The alignment assignment processing is completed using either a nearest neighbor method or an interpolation method. The nearest neighbor method involves selecting the traffic status value information corresponding to the most recent acquisition time information before the alignment moment and then... The result is assigned to the alignment time. The interpolation method is to select two adjacent collection time information on both sides of the alignment time, perform interpolation calculation on the corresponding traffic state numerical information according to the time interval ratio, and assign the interpolation calculation result to the alignment time. For example, when two traffic state collection information are formed between adjacent road segments at 10:00:05 and 10:00:15 and the alignment time is 10:00:10, linear interpolation calculation is performed on the two traffic state numerical information according to the time interval ratio from 10:00:05 to 10:00:15 to obtain the alignment value of 10:00:10. Then, the alignment values ​​of each alignment time are arranged in chronological order to form a traffic state alignment sequence.

[0030] S20: Obtain historical traffic status information corresponding to the target intersection, perform statistical calculations on the historical traffic status information according to the preset statistical window, and obtain historical baseline statistical parameter information.

[0031] In this embodiment, the preset statistical window refers to the time span parameter used to limit the statistical range of historical traffic status collection information, and the historical baseline statistical parameter information refers to the parameter record obtained by statistical analysis of historical traffic status collection information within the preset statistical window. The historical baseline statistical parameter information includes the baseline statistical parameter value corresponding to the time period identifier.

[0032] Specifically, when acquiring historical traffic status information corresponding to the target intersection, the historical traffic status information associated with the target intersection is read in chronological order. Based on the collection time information, historical traffic status information falling into a preset statistical window is filtered. Then, statistical calculations are performed on the filtered traffic status numerical information. The statistical calculations are performed separately according to time period identifiers. These time period identifiers are generated by preset segmentation rules and used to distinguish statistical objects from different time periods. For example, when dividing the daytime into morning peak time period identifiers and off-peak time period identifiers, the historical traffic status information is assigned to the corresponding time period identifier according to the collection time information. Baseline statistical parameter values ​​are then calculated for the traffic status numerical information under each time period identifier. The baseline statistical parameter values ​​are calculated using at least one of the mean calculation method and the quantile calculation method. The mean calculation method involves summing the traffic status numerical information under the same time period identifier within the preset statistical window and dividing by the sample size to obtain the mean. The quantile calculation method involves sorting the traffic status numerical information under the same time period identifier by value and selecting the corresponding value as the quantile value according to the preset quantile position. Subsequently, the baseline statistical parameter values ​​corresponding to each time period identifier are aggregated to form historical baseline statistical parameter information.

[0033] S30: Calculate traffic deviation information according to the traffic status alignment sequence and historical baseline statistical parameter information, and according to the preset deviation calculation rules.

[0034] In this embodiment, the deviation calculation rule information refers to the rule record used to convert the traffic state alignment sequence and historical baseline statistical parameter information into traffic deviation information. The traffic deviation information refers to the calculation result information used to characterize the degree of deviation of the traffic state alignment sequence from the historical baseline statistical parameter information.

[0035] Specifically, when calculating traffic deviation information, firstly, based on the data collection time information, the baseline statistical parameter value corresponding to the current alignment time of the traffic state alignment sequence is selected from the historical baseline statistical parameter information. Then, the alignment value corresponding to the current alignment time is read from the traffic state alignment sequence. Next, the deviation amount is calculated by performing deviation calculation on the alignment value and the baseline statistical parameter value according to the deviation calculation rules. The deviation amount calculation adopts either the difference calculation method or the ratio calculation method. The difference calculation method is to subtract the baseline statistical parameter value from the alignment value to obtain the deviation amount. The ratio calculation method is to divide the alignment value by the baseline statistical parameter value to obtain the deviation amount. Subsequently, the deviation amount is scale-converted according to the deviation calculation rules. The scale conversion adopts either the standardization conversion method or the interval conversion method. The standardization conversion method is to divide the deviation amount by a preset scale parameter to obtain the conversion value. The interval conversion method is to map the deviation amount to the corresponding interval number according to a preset interval threshold. Finally, the conversion values ​​or interval numbers obtained at each alignment time are arranged in chronological order to form traffic deviation information.

[0036] S40: Generate sensor credibility information based on the missing rate, fluctuation amplitude and cross-source consistency in the traffic state alignment sequence, and use the sensor credibility information to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information.

[0037] In this embodiment, the missing rate refers to the ratio of the number of alignment moments in the traffic state alignment sequence that did not form an alignment value within a preset statistical period to the total number of alignment moments. The fluctuation range refers to the statistical result of the change range of the alignment value in the traffic state alignment sequence within the preset statistical period. Cross-source consistency refers to the degree of difference between the alignment values ​​formed by different data sources at the same alignment moment. The sensor reliability information refers to the weighted record used to characterize the reliability of the data source, calculated from the missing rate, fluctuation range, and cross-source consistency. The fusion deviation information refers to the deviation result information obtained by performing weighted fusion processing on the traffic deviation information.

[0038] Specifically, when generating sensor reliability information, firstly, within a preset statistical period, the traffic state alignment sequence is processed for missing values ​​according to the data source. The moments with empty alignment values ​​in the alignment time are counted as missing moments, and the proportion of missing moments is calculated to obtain the missing rate. Then, the traffic state alignment sequence is processed for fluctuation according to the data source. The fluctuation statistical processing is performed by calculating the absolute value of the difference between the alignment values ​​of adjacent alignment moments and taking the average or maximum value within the preset statistical period to obtain the fluctuation amplitude. Then, at the same alignment moment, the alignment values ​​corresponding to different data sources are processed for difference calculation. The difference calculation processing is performed by calculating the absolute difference between the alignment values ​​of different data sources and taking the average or maximum value within the preset statistical period to obtain the difference value corresponding to cross-source consistency. Then, the missing rate, fluctuation amplitude, and difference value are converted into sensor reliability information according to a preset mapping relationship. The preset mapping relationship is completed by comparing the missing rate, fluctuation amplitude, and difference value with preset thresholds respectively and assigning greater reliability weights to cases that meet the threshold conditions. Finally, the traffic deviation information is multiplied by the weights corresponding to the sensor reliability information according to the data source, and the multiplication results are summed to obtain the fused deviation information.

[0039] S50: Obtain the preset phase configuration record information of the target intersection, and generate a set of candidate signal control schemes based on the fused deviation information and the preset phase configuration record information.

[0040] In this embodiment, the preset phase configuration record information refers to the phase configuration parameter record corresponding to the target intersection, the candidate signal control scheme set refers to the scheme set formed by the aggregation of multiple candidate signal control schemes, and the candidate signal control scheme refers to the signal control scheme record formed by the combination of phase sequence information, phase green light ratio information and phase offset information.

[0041] Specifically, when acquiring preset phase configuration record information, the phase configuration parameter record corresponding to the target intersection is read, and the phase number set and phase switching order constraint are extracted from the phase configuration parameter record. Then, the phase sequence information is determined based on the preset phase configuration record information. The phase sequence information is obtained by arranging the phase number set according to the phase switching order constraint. Then, the phase green light ratio information is generated based on the fusion deviation information. The phase green light ratio information is obtained by matching the fusion deviation information with the preset ratio segmentation rule. The preset ratio segmentation rule is used to map the fusion deviation information into several ratio intervals and set different ratio intervals. The green light time ratio is determined by assigning a higher green light time ratio to the corresponding phase when the fusion deviation information falls into a higher range, and a lower green light time ratio to the corresponding phase when the fusion deviation information falls into a lower range. Then, phase offset information is generated based on the preset phase configuration record information. The phase offset information is generated by reading the initial value of the parameter corresponding to the phase offset in the phase configuration parameter record and generating multiple offset values ​​according to the preset offset adjustment step size. Finally, the phase sequence information, phase green light ratio information and phase offset information are combined according to the preset combination rules to form multiple candidate signal control schemes and then aggregated to obtain a candidate signal control scheme set.

[0042] S60: For each candidate signal control scheme in the candidate signal control scheme set, perform queue growth estimation processing based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme to obtain scheme evaluation information.

[0043] In this embodiment, timing parameter information refers to the signal timing parameter record corresponding to the candidate signal control scheme, queue growth estimation processing refers to the process of differentially extrapolating the number of arriving vehicles and the number of vehicles that can be released within the estimation time range, and scheme evaluation information refers to the evaluation record used to record the queue growth estimation processing result.

[0044] Specifically, when performing queue growth estimation processing on each candidate signal control scheme, the initial queue length value corresponding to the estimated starting alignment time is first read from the traffic state alignment sequence, and the arrival vehicle value corresponding to each alignment time within the estimation time range is read. The arrival vehicle value is obtained by converting the increment of traffic state value information of adjacent alignment times or by directly reading the arrival count value given in the traffic state value information. Then, the cycle length value and the green light duration value of each phase within the estimation time range are read from the timing parameter information corresponding to the candidate signal control scheme, and the release time mark corresponding to each alignment time is obtained by expanding according to the cycle length value within the estimation time range. Then, the number of vehicles that can be released at each alignment time is determined according to the release time mark. The number of vehicles that can be released is determined by the green light duration of the corresponding alignment time. The time interval is allocated to the alignment time and multiplied by the preset unit release rate parameter. Then, the queue length is used as the initial value for recursion calculation for each alignment time. The queue length prediction value is obtained by adding the queue length value of the previous alignment time to the number of vehicles arriving at the current alignment time and subtracting the number of vehicles that can be released at the current alignment time. Queue length prediction values ​​less than zero are treated as zero. For example, if the number of vehicles arriving at a certain alignment time within the estimated time range is 8, the number of vehicles that can be released is 5, and the queue length value of the previous alignment time is 20, the queue length prediction value of the current alignment time is 23. Finally, the sequence of queue length prediction values ​​and the queue growth statistics within the estimated time range are written into the scheme evaluation information and associated with the corresponding candidate traffic control scheme.

[0045] S70: Based on the scheme evaluation information and preset constraint rule information, determine the target signal control scheme from the candidate signal control scheme set, perform signal timing control on the target intersection according to the target signal control scheme, and obtain the target timing parameter information.

[0046] In this embodiment, the preset constraint rule information refers to the rule record used to limit the available range of candidate signal control schemes, the target signal control scheme refers to the signal control scheme record selected from the candidate signal control scheme set for actual execution, and the target timing parameter information refers to the timing parameter record corresponding to the target signal control scheme and used for issuance and execution.

[0047] Specifically, when determining the target signal control scheme, the preset constraint rule information is first read, and each candidate signal control scheme in the candidate signal control scheme set is associated with its corresponding scheme evaluation information. Then, constraint verification processing is performed on each candidate signal control scheme in sequence. The constraint verification processing is completed by comparing the cycle length value, phase green light duration value, and phase switching sequence in the timing parameter information corresponding to the candidate signal control scheme with the constraints in the preset constraint rule information item by item. During the comparison, candidate signal control schemes that do not meet the constraints are marked as unusable and removed from the candidate set, while candidate signal control schemes that meet the constraints are retained as usable candidate signal control schemes. Then, the scheme evaluation information corresponding to the usable candidate signal control schemes is sorted. The process involves statistically analyzing the predicted queue length sequence in the scheme evaluation information using preset sorting rules to generate sorting key values. These preset sorting rules define the calculation method and direction of the sorting key values. Subsequently, the first candidate signal control scheme from the available candidate signal control schemes is selected as the target signal control scheme. Finally, the timing parameter information corresponding to the target signal control scheme is written into the signal control parameter table to generate target timing parameter information. Then, signal timing control is performed on the target intersection according to the target timing parameter information. This is achieved by writing the cycle length value and the green light duration value of each phase from the signal control parameter table into the parameter area of ​​the signal control equipment and switching to the corresponding phase at the phase switching time.

[0048] In one embodiment, such as Figure 2 As shown, in step S30, traffic deviation information is calculated according to the traffic state alignment sequence and historical baseline statistical parameter information, and according to the preset deviation calculation rule information, including: S301: Select traffic status information corresponding to the target sampling time from the traffic status alignment sequence.

[0049] In this embodiment, the target sampling time is the sampling time to be processed in the traffic state alignment sequence, and the traffic state collection information is a combination record of traffic state numerical information and collection time information corresponding to the target sampling time.

[0050] Specifically, the traffic state alignment sequence is traversed in chronological order, and the sampling time encountered is taken as the target sampling time. Records whose collection time information matches the target sampling time are retrieved from the traffic state alignment sequence, and the traffic state value information is read as the traffic state collection information corresponding to the target sampling time. When there are multiple data source records at the position corresponding to the target sampling time in the traffic state alignment sequence, the records of each data source are read to form a set of traffic state collection information corresponding to the same target sampling time. When no traffic state value information is formed at the position corresponding to the target sampling time in the traffic state alignment sequence, the traffic state collection information corresponding to the target sampling time is recorded as empty, and the collection time information is retained for subsequent processing.

[0051] S302: Select the historical baseline statistical parameter information corresponding to the target sampling time from the historical baseline statistical parameter information.

[0052] Specifically, the acquisition time information corresponding to the target sampling time is read, and the acquisition time information is converted into a time period identifier according to the preset segmentation rules. Then, the baseline statistical parameter value record with the same time period identifier is retrieved in the historical baseline statistical parameter information, and the baseline statistical parameter value record is read as the historical baseline statistical parameter information corresponding to the target sampling time. When multiple baseline statistical parameter value records with the same time period identifier are retrieved, one of them is selected as the historical baseline statistical parameter information corresponding to the target sampling time according to the record storage order.

[0053] S303: Based on the deviation calculation rule information, the deviation calculation process is performed on the traffic status collection information and the historical baseline statistical parameter information to obtain the single index deviation sequence.

[0054] Specifically, first, traffic status numerical information corresponding to the same indicator name is selected from the traffic status collection information and sorted according to the collection time information. Then, baseline statistical parameter values ​​corresponding to the indicator name and matching the time period identifier of the collection time information are selected from the historical baseline statistical parameter information. The deviation calculation rule information sets the deviation calculation formula as either a difference calculation formula or a ratio calculation formula. The difference calculation formula is written as follows: Where xi represents the traffic status value corresponding to the i-th data collection time, b represents the baseline statistical parameter value, and di represents the deviation corresponding to the i-th data collection time. The ratio calculation formula is written as: When the baseline statistical parameter value meets the preset zero value condition, the baseline statistical parameter value is replaced with the preset replacement value b0 according to the deviation calculation rule information, and then the division operation is performed and written as... Then, for each collection time information, the corresponding xi and b or b0 are substituted sequentially to perform a deviation calculation and obtain the corresponding di. Finally, the di corresponding to each collection time information is arranged in the order of the collection time information to form a single index deviation sequence. For example, when the baseline statistical parameter value is 20 and the traffic status values ​​corresponding to the three collection time information are 30, 28 and 25 respectively, the difference calculation formula yields deviations of 10, 8 and 5 respectively, which are then arranged in the order of time to obtain a single index deviation sequence.

[0055] S304: Based on the deviation calculation rule information, perform normalization and aggregation processing on the single index deviation sequence to obtain the passage deviation information.

[0056] Specifically, first, the normalized parameter information corresponding to the indicator name is read according to the deviation calculation rule information. The normalized parameter information includes the normalization lower limit value L and the normalization upper limit value U. Then, normalization calculation is performed on each deviation value di in the single indicator deviation sequence. The normalization calculation adopts the linear normalization formula. Where ni represents the normalized value corresponding to the i-th deviation, and for cases where UL meets the preset zero value condition, U or L is replaced with a preset substitution value according to the deviation calculation rule information before performing a division operation. Then, boundary clipping is performed on the normalized result, treating ni values ​​less than 0 as 0 and ni values ​​greater than 1 as 1. Subsequently, the weight value w corresponding to the indicator name is read according to the deviation calculation rule information, and a convergence formula is used. The single-indicator aggregation value pi corresponding to the i-th data collection time is obtained. When the deviation calculation rule information sets multiple indicator names to participate in the aggregation simultaneously, the single-indicator aggregation values ​​corresponding to each indicator name under the same data collection time are summed and the formula is applied. The convergence deviation value corresponding to the collection time information is obtained. Finally, the convergence deviation value or the P corresponding to each collection time information is arranged in the order of the collection time information to form the passage deviation information.

[0057] In one embodiment, such as Figure 3 As shown, in step S40, sensor reliability information is generated based on the missing rate, fluctuation amplitude, and cross-source consistency in the traffic state alignment sequence. This sensor reliability information is then used to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information, including: S401: Within the preset statistical window, perform statistical processing on the traffic status aligned sequence according to the data source to obtain the missing rate information and fluctuation range information corresponding to each data source.

[0058] In this embodiment, the data source refers to the source identifier of the acquisition terminal used to form the traffic state alignment sequence, the missing rate information is used to characterize the proportion of traffic state alignment sequences within the preset statistical window that do not form alignment values ​​at the alignment time, and the fluctuation amplitude information is used to characterize the amplitude of the change in alignment values ​​of traffic state alignment sequences within the preset statistical window between adjacent alignment times.

[0059] Specifically, within a preset statistical window, the set of alignment times and the set of alignment values ​​corresponding to the data source in the traffic status alignment sequence are extracted according to the data source. The total number of alignment times within the preset statistical window is denoted as N. Alignment times with empty alignment values ​​are counted as missing alignment times, and the number of missing alignment times is denoted as M. The missing rate value corresponding to the data source is obtained according to the missing rate calculation formula r=M / N and written into the missing rate information. Then, after removing missing alignment times, the remaining alignment values ​​are arranged in chronological order and recorded as... When K is greater than or equal to 2, the alignment values ​​at adjacent alignment times are calculated by taking the absolute value of the difference and then applying the formula. Obtain the sequence of adjacent changes Then calculate according to the fluctuation range formula. The fluctuation amplitude value corresponding to the data source is obtained and written into the fluctuation amplitude information. When K is less than 2, the fluctuation amplitude value is written into the fluctuation amplitude information according to the preset default value. Finally, the statistical processing of each data source in the traffic status alignment sequence is repeated and aggregated to form the missing rate information and fluctuation amplitude information corresponding to each data source.

[0060] S402: Within a preset statistical window, perform consistency comparison processing on traffic status information collected from different data sources at the same sampling time to obtain cross-source consistency information corresponding to each data source.

[0061] In this embodiment, cross-source consistency information is used to characterize the degree of difference between traffic status collection information from different data sources at the same sampling time.

[0062] Specifically, within a preset statistical window, each sampling time of the traffic state alignment sequence is iterated sequentially. At each sampling time, the alignment value set is extracted from the traffic state data collected from different data sources, and alignment values ​​with empty alignment values ​​are removed. Let m be the number of elements in the alignment value set. When m is greater than or equal to 2, a reference value for the alignment value set is calculated. Where xk represents the alignment value of the k-th data source at sampling time, then the deviation value between the alignment value of each data source at sampling time and the reference value is calculated and written as . Then, the deviation values ​​obtained at each sampling time within the preset statistical window are summarized and calculated according to the data source. The summary calculation uses the mean formula. Where T represents the number of sampling times involved in the calculation. This represents the deviation value of data source k at sampling time t. Finally, cross-source consistency information is generated according to the magnitude of the deviation value. Ek is used as the cross-source difference quantity corresponding to data source k and written into the cross-source consistency information. When m is less than 2, the calculation of the sampling time is skipped and not included in T.

[0063] S403: Perform combined operations on the missing rate information, fluctuation range information, and cross-source consistency information to obtain the sensing reliability information corresponding to each data source.

[0064] Specifically, during the combined operation, the missing rate value r from the missing rate information, the fluctuation amplitude value A from the fluctuation amplitude information, and the cross-source difference E from the cross-source consistency information are read from each data source. Then, the missing rate value r, the fluctuation amplitude value A, and the cross-source difference E are converted into dimensionless values ​​according to the preset scaling parameters. The conversion formulas are written as r'=r and A'=A / S. A E'=E / S E S A With S E The preset scale parameters are then used, and then based on the preset combined weights wr and w A w E Perform a weighted summation on the dimensionless values ​​and write it as Finally, according to the sensor reliability calculation formula... Obtain the sensor confidence value c corresponding to the data source and write it into the sensor confidence information, where S A or S E When the preset zero value condition is met, the corresponding scale parameter is replaced with the preset substitution value before the division operation is performed.

[0065] S404: Use the sensor reliability information to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information.

[0066] Specifically, during the weighted fusion process, at each sampling time, the traffic deviation value corresponding to different data sources in the traffic deviation information is read and recorded as pk, where k is the data source number. Simultaneously, the sensor confidence value corresponding to the data source number in the sensor confidence information is read and recorded as ck. Then, the weighted fusion formula is applied. Calculate the fusion deviation value P corresponding to the sampling time, where m is the number of data sources participating in the fusion at the sampling time. Write the fusion deviation value P into the fusion deviation information. Then, repeatedly perform weighted fusion processing on each sampling time within a preset statistical window and arrange the fusion deviation values ​​in the order of the sampling time to form fusion deviation information. Where m is the number of data sources participating in the fusion at the sampling time, the fusion deviation value P is calculated. When the preset zero value condition is met, the fusion deviation value P is written into the fusion deviation information according to the preset default value.

[0067] In one embodiment, such as Figure 4As shown, in step S50, the preset phase configuration record information of the target intersection is obtained, and a set of candidate signal control schemes is generated based on the fused deviation information and the preset phase configuration record information, including: S501: Determine the phase sequence information based on the preset phase configuration record information.

[0068] Specifically, the phase number set and phase switching order record are read from the preset phase configuration record information. The phase switching order record is used to represent the sequential relationship between phase numbers. Then, the phase number set is sorted according to the phase switching order record to generate a sorted result sequence. The sorting process is completed by arranging the preceding phase numbers in the phase switching order record in sequence and inserting the corresponding subsequent phase number after each preceding phase number. When there are multiple optional subsequent phase numbers in the phase switching order record, the subsequent phase number is selected according to the priority order field in the preset phase configuration record information and inserted into the sorted result sequence. Finally, the sorted result sequence is written into the phase sequence information.

[0069] S502: Determine the phase green light ratio information corresponding to the phase sequence information based on the fusion deviation information, and determine the phase offset information corresponding to the phase sequence information based on the fusion deviation information and the preset phase configuration record information.

[0070] In this embodiment, the phase green light ratio information refers to the set of green light time proportion parameters corresponding one-to-one with each phase number in the phase sequence information, and the phase offset information refers to the set of initial offset parameters within the period corresponding one-to-one with each phase number in the phase sequence information.

[0071] Specifically, the fusion deviation value P corresponding to the target control cycle is read, and the set of default ratio parameters corresponding to the phase sequence information is read from the preset phase configuration record information and recorded as follows. And the set of upper and lower limits of the ratio parameters, denoted as Based on the fusion deviation value P, the adjustment coefficient k is determined according to the mapping relationship of the ratio adjustment coefficient in the preset phase configuration record information. The temporary ratio is calculated for each phase number j in the phase sequence information. And for the temporary ratio r' j Performing upper and lower limit constraint processing yields Then assign r'' to all phase numbers j Normalization was performed to obtain And Write the phase green light allocation information. For example, if the phase sequence information contains three phase numbers and the default allocation parameter set is 0.30, 0.30, and 0.40, and the fusion deviation value P matches the adjustment coefficient k of 1.10, the temporary allocation is 0.33, 0.33, and 0.44. After upper and lower limit constraint processing and normalization processing, write the phase green light allocation information. Then, read the default offset parameter set corresponding to the phase sequence information from the preset phase configuration record information and record it as... And the offset step size parameter and the offset upper and lower limit parameters are denoted as: and Based on the fusion deviation value P, the offset step number s is determined according to the offset step mapping relationship in the preset phase configuration record information. For each phase number j in the phase sequence information, the offset candidate value is calculated. And perform upper and lower limit constraints on the offset candidate value o'j to obtain ,Will Write the phase offset information.

[0072] S503: Combine the phase sequence information, phase green light ratio information and phase offset information to obtain a set of candidate signal control schemes.

[0073] In this embodiment, the combination processing refers to the process of binding the green light ratio value corresponding to the phase number in the phase green light ratio information and the phase offset value corresponding to the phase number in the phase offset information according to the phase number order in the phase sequence information to form a candidate signal control scheme record. The candidate signal control scheme set refers to the set record formed by the collection of at least one candidate signal control scheme record.

[0074] Specifically, the phase sequence information is read to obtain the phase number sequence. The green light ratio value corresponding to the phase number in the phase green light ratio information is selected sequentially according to the phase number sequence, and the phase offset value corresponding to the phase number in the phase offset information is selected. The phase number, green light ratio value, and phase offset value are associated and written according to the same phase number to form a candidate signal control scheme record. The candidate signal control scheme record is written into the candidate signal control scheme set. When there are multiple sets of phase number sequence records in the preset phase configuration record information, or multiple sets of green light ratio value sets in the phase green light ratio information, or multiple sets of phase offset value sets in the phase offset information, the association writing process is repeatedly performed on different phase number sequence sequences and corresponding green light ratio value sets and corresponding phase offset value sets according to the combination order given by the preset phase configuration record information, and written into the candidate signal control scheme set one by one to obtain the candidate signal control scheme set.

[0075] In one embodiment, such as Figure 5As shown, in step S60, for each candidate signal control scheme in the candidate signal control scheme set, queue growth estimation processing is performed based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme to obtain scheme evaluation information, including: S601: Obtain preset estimation window information.

[0076] In this embodiment, the preset closed-loop correction window information refers to the time range parameter record used to carry out correction processing.

[0077] Specifically, the estimation window parameter record corresponding to the target intersection in the preset configuration table is read. The estimation window parameter record includes the estimation start time parameter, the estimation duration parameter, and the estimation step size parameter. The estimation time range is determined based on the estimation start time parameter and the estimation duration parameter, and an estimation time sequence covering the estimation time range is generated based on the estimation step size parameter. The estimation start time parameter, the estimation duration parameter, the estimation step size parameter, the estimation time range, and the estimation time sequence are written into the preset estimation window information.

[0078] S602: Extract arrival flow rate and initial queue length value within the time range corresponding to the preset estimation window information based on the traffic status alignment sequence.

[0079] Specifically, the estimation start time and estimation end time are determined based on the preset estimation window information. The set of alignment times falling within the estimation start time to estimation end time is located in the traffic state alignment sequence. Then, at the first alignment time in the alignment time set, the traffic state value information corresponding to the queue length is read and the reading result is recorded as the initial queue length value. At the same time, the traffic state value information corresponding to the arrival flow is read one by one along the time order of the alignment time set and the reading results are arranged according to the alignment time to form the arrival flow value sequence. When the traffic state alignment sequence records the number of vehicles entering the target intersection in each sampling period according to the sampling period, the number of vehicles corresponding to each alignment time is directly used as the arrival flow value. When the traffic state alignment sequence records the vehicle arrival rate according to the sampling period, the vehicle arrival rate is multiplied by the sampling period duration to obtain the arrival flow value. The arrival flow value and the initial queue length value are written into the input data record of the subsequent queue growth estimation processing.

[0080] S603: Determine the effective release time value for each phase based on the timing parameter information corresponding to the candidate signal control scheme, and calculate the number of vehicles that can be released based on the effective release time value and the preset release capacity parameter information, according to the preset estimation window information.

[0081] Specifically, the estimation start and end times are determined based on the preset estimation window information. The cycle length value, the phase number sequence corresponding to the phase sequence information, and the phase green light duration and phase offset value corresponding to each phase number sequence are read from the timing parameter information corresponding to the candidate signal control scheme. Then, using the estimation start time as the time starting point and combining it with the phase offset value, the first cycle's phase start time sequence within the estimation time range is determined. Next, the phase start time sequence for each cycle is expanded along the estimation time range according to the cycle length value. Finally, the green light time period appearing for each phase number within the estimation time range is calculated and estimated. The overlap duration within the time range is calculated, and the cumulative overlap duration yields the effective release time value corresponding to the phase number. The overlap duration is obtained by subtracting the larger of the start time of the green light period and the estimated end time from the smaller of the end time and the estimated end time of the green light period, treating results less than zero as zero. Then, preset release capacity parameters are read, and the unit-time release capacity value corresponding to the phase number is obtained. The unit-time release capacity value represents the number of vehicles allowed to pass through the phase number within a unit of time. Based on the effective release time value and the unit-time release capacity value, the number of vehicles that can be released for the phase number is calculated and written as... Where tj represents the effective release time value corresponding to the phase number, qj represents the release capacity value per unit time corresponding to the phase number, and nj represents the number of vehicles that can be released corresponding to the phase number. Finally, the nj corresponding to each phase number in the phase number sequence is summed to obtain the number of vehicles that can be released corresponding to the preset estimation window information, which is recorded as the input quantity for queue growth estimation processing.

[0082] S604: Perform queue recursive calculation based on the arrival flow value, initial queue length value, and number of vehicles that can be released to obtain the predicted queue length change sequence.

[0083] Specifically, the estimated time sequence is determined based on the preset estimation window information, and the queue length value corresponding to the first estimated time in the estimated time sequence is set as the initial queue length value. Then, queue recursion calculation is performed once for each subsequent estimated time according to the chronological order of the estimated time sequence. Within the i-th estimation step, the number of arriving vehicles corresponding to the estimation step in the arrival flow value sequence is read and recorded as ai. At the same time, the number of vehicles that can be released is allocated according to the estimated time sequence as the number of vehicles that can be released corresponding to the estimation step and recorded as si. The allocation is done by either dividing the number of vehicles that can be released equally according to the number of estimation steps or by allocating according to the green light time period coverage ratio corresponding to the timing parameter information. Then, a recursive formula is used to obtain si. Calculate the predicted queue length Li corresponding to the end time of the estimated step, where This is the predicted queue length value corresponding to the end time of the previous estimation step. This is used to treat predicted queue length values ​​less than zero as zero, and finally, the Li values ​​obtained from each estimated step are arranged in chronological order of the estimated time sequence to form a predicted queue length change sequence.

[0084] S605: Calculate the queue growth estimate based on the predicted queue length change sequence, and write the queue growth estimate into the scheme evaluation information.

[0085] Specifically, the predicted queue length value corresponding to the estimated start time is read from the predicted queue length change sequence and recorded as L0, and the predicted queue length value corresponding to the estimated end time is read and recorded as LT. Then, the queue growth estimate is calculated using the formula. Calculate the estimated queue growth and the estimated queue growth The candidate signal control scheme identifier corresponding to the candidate signal control scheme is associated with and written into the scheme evaluation information. When the predicted queue length change sequence contains predicted queue length values ​​with multiple estimated steps, the peak value calculation formula is also used. Calculate the peak value of the predicted queue length change sequence and and Include this in the scheme evaluation information.

[0086] In one embodiment, such as Figure 6 As shown, in step S601, which involves obtaining the preset estimation window information, the method further includes: S6011: Obtain traffic status data corresponding to the preset closed-loop correction window information, perform deviation calculation processing based on the traffic status data and scheme evaluation information, and update historical baseline statistical parameter information and sensor reliability information.

[0087] In this embodiment, the preset closed-loop correction window information includes correction start time parameters, correction duration parameters, and correction step size parameters. Deviation calculation processing refers to the process of performing difference calculations on the observed values ​​extracted from traffic state collection information and the evaluation values ​​in the scheme evaluation information to form a deviation value. Updating historical baseline statistical parameter information refers to the process of performing rolling merging calculations on the baseline statistical parameter values ​​in the historical baseline statistical parameter information and writing them back. Updating sensor reliability information refers to the process of performing rolling merging calculations on the sensor reliability values ​​in the sensor reliability information and writing them back.

[0088] Specifically, the start and end times of the correction are determined based on the preset closed-loop correction window information. Traffic state data is collected between the start and end times. The traffic state data is sorted by collection time, and the traffic state values ​​corresponding to the start and end times of the correction are selected. Then, the evaluation values ​​corresponding to the target signal control scheme identifier are read from the scheme evaluation information, and the evaluation values ​​matching the correction time range are selected. The deviation calculation formula is then applied. Calculate the deviation value Where y represents the observed value selected or statistically obtained from traffic condition data collection, and x represents the evaluation value read from the scheme evaluation information, then statistical calculations are performed on the traffic condition data collection within the correction time range to obtain the correction statistical parameter value. The statistical calculations are performed according to the statistical caliber used in the historical baseline statistical parameter information and output the correction statistical parameter value with the same dimensions as the baseline statistical parameter value. The correction statistical parameter value and the baseline statistical parameter value in the historical baseline statistical parameter information are then updated with preset weight parameters. Perform merge calculations and write as ,in This represents the baseline statistical parameter values ​​before the update. Indicates the value of the correction statistical parameter. This indicates the updated baseline statistical parameter values, and... After updating by writing historical baseline statistical parameters, the missing rate, fluctuation range, and cross-source consistency information of the traffic status data are calculated according to the data source within the correction time range, and then processed by combination operation to obtain the corrected sensor confidence value. Then correct the sensor confidence value The sensor confidence value in the sensor confidence information is updated with a preset weight parameter. Perform merge calculations and write as ,in This indicates the sensor confidence value before the update. This indicates the updated sensor confidence value, and will The update is completed by writing the sensor reliability information.

[0089] In one embodiment, such as Figure 7 As shown, in step S70, based on the scheme evaluation information and preset constraint rule information, the target signal control scheme is determined from the candidate signal control scheme set, and signal timing control is performed on the target intersection according to the target signal control scheme to obtain the target timing parameter information, including: S701: Perform constraint verification processing on the scheme evaluation information and preset constraint rule information corresponding to each candidate signal control scheme in the candidate signal control scheme set to obtain constraint verification result information.

[0090] Specifically, the system reads and parses preset constraint rule information to obtain a set of constraint items. This set includes at least cycle length constraints, phase sequence constraints, phase green light allocation constraints, and phase offset constraints. Then, it sequentially selects each candidate signal control scheme from the candidate signal control scheme set and reads the timing parameter information corresponding to the candidate scheme. Simultaneously, it reads the scheme evaluation information corresponding to the candidate signal control scheme. The system compares the cycle length value in the timing parameter information with the value range of the cycle length constraint item, matches the phase sequence information with the allowed order set of the phase sequence constraint item, and assigns each phase in the phase green light allocation information... The green light ratio corresponding to the phase number is compared with the upper and lower limits of the phase green light ratio constraint item. The phase offset value corresponding to each phase number in the phase offset information is compared with the upper and lower limits of the phase offset constraint item. When any comparison result does not meet the preset constraint rule information, the constraint verification result information corresponding to the candidate signal control scheme is written as a failed state. When all comparison results meet the preset constraint rule information, the constraint verification result information corresponding to the candidate signal control scheme is written as a passed state. The passed state or failed state is associated with the scheme evaluation information corresponding to the candidate signal control scheme and written into the constraint verification result information.

[0091] S702: Filter the constraint verification result information to indicate the candidate signal control schemes that have passed the constraint verification, and obtain a subset of candidate signal control schemes that have passed the verification.

[0092] Specifically, the constraint verification result information is traversed and the constraint verification status corresponding to each candidate signal control scheme is read sequentially. When the constraint verification status is "passed", the candidate signal control scheme corresponding to the constraint verification status is extracted from the candidate signal control scheme set and written into the candidate signal control scheme subset that has passed the verification. When the constraint verification status is "failed", the candidate signal control scheme corresponding to the constraint verification status is removed from the subsequent processing queue and not written into the candidate signal control scheme subset that has passed the verification. After the traversal is completed, the candidate signal control scheme set that has been written is recorded as the candidate signal control scheme subset that has passed the verification and output.

[0093] S703: Sort the subset of candidate signal control schemes that have passed the verification according to the preset sorting rules to obtain the target sequence number information.

[0094] Specifically, the preset sorting rule information is read and parsed to obtain sorting index items and sorting direction information. The sorting index items correspond to at least one of the estimated queue growth value and the peak record of the predicted queue length change sequence in the scheme evaluation information. Then, for each candidate signal control scheme in the verified candidate signal control scheme subset, the scheme evaluation information corresponding to the candidate signal control scheme is extracted and the index value corresponding to the sorting index item is read. The index values ​​of each candidate signal control scheme are sorted according to the sorting direction information to obtain the sorting result sequence. The sequence number corresponding to the first element of the sorting result sequence is written into the target sequence number information. When the preset sorting rule information sets multiple sorting index items, the first-level sorting index items are sorted first, and the second-level sorting index items are sorted again if the index values ​​are the same, until a unique first-level sequence number is obtained and written into the target sequence number information.

[0095] S704: Select the candidate signal control scheme corresponding to the target sequence number information from the subset of candidate signal control schemes that have passed the verification as the target signal control scheme; write the timing parameter information corresponding to the target signal control scheme into the signal control parameter table to obtain the target timing parameter information.

[0096] Specifically, the target sequence number information is read, and a candidate signal control scheme with a sequence number matching the target sequence number information is located in the subset of candidate signal control schemes that have passed the verification. The located candidate signal control scheme is written into the target signal control scheme record to form the target signal control scheme. Then, the timing parameter information is extracted from the target signal control scheme and written into the signal control parameter table in the order of the fields. The field order is consistent with the field order in the timing parameter information. The fields must at least include the period length value, the phase number order corresponding to the phase sequence information, the green light duration value corresponding to the phase green light ratio information, and the phase offset value corresponding to the phase offset information. During writing, each field is formatted and the field value is converted into a numerical format supported by the signal control parameter table. Finally, the signal control parameter table after writing is output as the target timing parameter information.

[0097] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0098] In one embodiment, a dynamic traffic optimization management device is provided, which corresponds one-to-one with the dynamic traffic optimization management method described in the above embodiments. For example... Figure 8 As shown, the dynamic traffic optimization management device includes a traffic status acquisition and time alignment module, a historical baseline statistical parameter calculation module, a traffic deviation calculation module, a credibility generation and deviation generation module, a candidate signal control scheme set generation module, a queue growth estimation and scheme evaluation module, and a signal timing control module.

[0099] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A dynamic traffic optimization management method, characterized in that, The dynamic traffic optimization management method includes: Traffic status information of the target intersection and adjacent road segments is acquired, and the traffic status information is time-aligned to obtain a traffic status aligned sequence. Acquire historical traffic status information corresponding to the target intersection, perform statistical calculations on the historical traffic status information according to the preset statistical window, and obtain historical baseline statistical parameter information. Based on the traffic status alignment sequence and historical baseline statistical parameters, the traffic deviation information is calculated according to the preset deviation calculation rules. Based on the missing rate, fluctuation amplitude and cross-source consistency in the traffic state alignment sequence, sensor credibility information is generated. The traffic deviation information is then weighted and fused using the sensor credibility information to obtain fused deviation information. Obtain the preset phase configuration record information of the target intersection, and generate a set of candidate signal control schemes based on the fused deviation information and the preset phase configuration record information; For each candidate signal control scheme in the candidate signal control scheme set, queue growth estimation is performed based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme to obtain scheme evaluation information; Based on the scheme evaluation information and preset constraint rule information, the target signal control scheme is determined from the candidate signal control scheme set, and the signal timing control is performed on the target intersection according to the target signal control scheme to obtain the target timing parameter information.

2. The dynamic traffic optimization management method according to claim 1, characterized in that, The step of calculating traffic deviation information based on traffic state alignment sequence and historical baseline statistical parameter information, according to preset deviation calculation rules, includes: Select traffic status information corresponding to the target sampling time from the traffic status alignment sequence; Select the historical baseline statistical parameter information corresponding to the target sampling time from the historical baseline statistical parameter information; Based on the deviation calculation rule information, the deviation calculation process is performed on the traffic state collection information and the historical baseline statistical parameter information to obtain the single index deviation sequence. Based on the deviation calculation rule information, normalization and aggregation processing is performed on the single indicator deviation sequence to obtain the passage deviation information.

3. The dynamic traffic optimization management method according to claim 1, characterized in that, The process involves generating sensor reliability information based on the missing rate, fluctuation amplitude, and cross-source consistency in the traffic state alignment sequence. This sensor reliability information is then used to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information, including: Within a preset statistical window, the traffic status aligned sequence is statistically processed according to the data source to obtain the missing rate information and fluctuation range information corresponding to each data source. Within a preset statistical window, traffic status information collected from different data sources at the same sampling time is compared for consistency to obtain cross-source consistency information corresponding to each data source. The missing rate information, fluctuation range information and cross-source consistency information are combined and processed to obtain the sensing reliability information corresponding to each data source. The traffic deviation information is weighted and fused using the sensor reliability information to obtain the fused deviation information.

4. The dynamic traffic optimization management method according to claim 1, characterized in that, The process of obtaining the preset phase configuration record information of the target intersection and generating a set of candidate traffic control schemes based on the fused deviation information and the preset phase configuration record information includes: Phase sequence information is determined based on preset phase configuration record information; The phase green light ratio information corresponding to the phase sequence information is determined based on the fusion deviation information and the preset phase configuration record information; the phase offset information corresponding to the phase sequence information is determined based on the fusion deviation information and the preset phase configuration record information. The phase sequence information, phase green light ratio information, and phase offset information are combined and processed to obtain a set of candidate signal control schemes.

5. The dynamic traffic optimization management method according to claim 1, characterized in that, The process involves performing queue growth estimation on each candidate signal control scheme in the candidate signal control scheme set, based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme, to obtain scheme evaluation information, including: Obtain information from the preset estimation window; Based on the traffic status alignment sequence, the arrival flow rate and initial queue length are extracted within the time range corresponding to the preset estimation window information. The effective release time value for each phase is determined based on the timing parameter information corresponding to the candidate signal control scheme. The number of vehicles that can be released is calculated based on the effective release time value and the preset release capacity parameter information corresponding to the preset estimation window information. Based on the arrival flow rate, initial queue length, and number of vehicles that can be released, a queue recursive calculation is performed to obtain the predicted queue length change sequence. The estimated queue growth is calculated based on the predicted queue length change sequence, and the estimated queue growth is written into the scheme evaluation information.

6. The dynamic traffic optimization management method according to claim 5, characterized in that, The step of obtaining the preset estimation window information also includes: Obtain traffic status data corresponding to the preset closed-loop correction window information, perform deviation calculation processing based on the traffic status data and scheme evaluation information, and update historical baseline statistical parameter information and sensor reliability information.

7. The dynamic traffic optimization management method according to claim 1, characterized in that, Based on the scheme evaluation information and preset constraint rule information, a target signal control scheme is determined from the candidate signal control scheme set, and signal timing control is performed on the target intersection according to the target signal control scheme to obtain target timing parameter information, including: Perform constraint verification processing on the scheme evaluation information and preset constraint rule information corresponding to each candidate signal control scheme in the candidate signal control scheme set to obtain constraint verification result information; The information on the constraint verification results indicates the candidate signal control schemes that have passed the constraint verification, thus obtaining a subset of candidate signal control schemes that have passed the verification. The subset of candidate signal control schemes that have passed the verification are sorted according to the preset sorting rules to obtain the target sequence number information; Select the candidate signal control scheme corresponding to the target sequence number information from the subset of candidate signal control schemes that have passed the verification as the target signal control scheme; write the timing parameter information corresponding to the target signal control scheme into the signal control parameter table to obtain the target timing parameter information.

8. A dynamic traffic optimization management device, characterized in that, The dynamic traffic optimization management device includes: The traffic status acquisition and time alignment module is used to acquire traffic status information of the target intersection and adjacent road segments, perform time alignment processing on the traffic status acquisition information, and obtain a traffic status alignment sequence. The historical baseline statistical parameter calculation module is used to acquire historical traffic status information corresponding to the target intersection, perform statistical calculations on the historical traffic status information according to the preset statistical window, and obtain historical baseline statistical parameter information. The traffic deviation calculation module is used to calculate traffic deviation information according to preset deviation calculation rules based on traffic state alignment sequence and historical baseline statistical parameter information. The credibility generation and deviation generation modules are used to generate sensor credibility information based on the missing rate, fluctuation amplitude and cross-source consistency in the traffic state alignment sequence. The sensor credibility information is used to perform weighted fusion processing on the traffic deviation information to obtain fused deviation information. The candidate signal control scheme set generation module is used to obtain the preset phase configuration record information of the target intersection and generate a candidate signal control scheme set based on the fused deviation information and the preset phase configuration record information. The queue growth estimation and scheme evaluation module is used to perform queue growth estimation processing on each candidate signal control scheme in the candidate signal control scheme set based on the traffic state alignment sequence and the timing parameter information corresponding to the candidate signal control scheme, and obtain scheme evaluation information. The signal timing control module is used to determine the target signal control scheme from the candidate signal control scheme set based on the scheme evaluation information and preset constraint rule information, and to perform signal timing control on the target intersection according to the target signal control scheme to obtain the target timing parameter information.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the dynamic traffic optimization management method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the dynamic traffic optimization management method as described in any one of claims 1 to 7.

Images