A traffic signal control effect evaluation method

Abstract

The application provides a traffic signal control effect evaluation method, and relates to the technical field of data processing, and the method comprises the following steps: clustering analysis is performed on multiple artificial intervention events according to the operation data of the traffic signal in a preset historical time period and the artificial intervention demand offset information of each artificial intervention event, a plurality of artificial intervention mode clusters are obtained, and an intervention regularity index is calculated; green wave coordination cascade damage quantification is performed according to the operation records and the corresponding actual release records of each artificial intervention event in the plurality of artificial intervention mode clusters, and a green wave coordination damage index is obtained; and the evaluation result of the traffic signal control effect in the preset historical time period is determined according to the artificial intervention demand offset information of the multiple artificial intervention events, the intervention regularity index and the green wave coordination damage index. The application improves the authenticity and fineness of the evaluation of the traffic control signal.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method for evaluating the effectiveness of traffic signal control. Background Technology

[0002] With the accelerating pace of urbanization, urban road traffic congestion has become increasingly prominent. Traffic signal control systems, as a crucial component of intelligent urban traffic management, play a key role in optimizing intersection efficiency and alleviating traffic pressure. In practice, most urban intersections still widely employ fixed-time-period timing schemes based on historical traffic flow patterns, supplemented by arterial green wave coordination control strategies, aiming to achieve continuous vehicle passage and reduce the number of stops and delays. While signal control systems typically operate automatically according to preset schemes, traffic management personnel often need to intervene in abnormal situations such as sudden traffic surges, emergencies, or special events. They must manually implement measures such as forced phase changes, phase locking, and extended green light periods to address instantaneous traffic pressure.

[0003] Currently, the evaluation of traffic signal control effectiveness mainly relies on macro-level traffic operation indicators such as saturation, queue length, average delay time, and floating car speed. It typically focuses on the system's performance under automatic control, treating human intervention as an abnormal or atypical operating condition, which is often discarded or ignored as noise during data processing. Alternatively, it records the frequency and duration of human intervention, but this often remains at a simple statistical level, lacking in-depth analysis of the underlying changes in traffic demand and failing to establish a correlation between human intervention events and the degree of matching with the original timing scheme.

[0004] However, simply excluding periods of manual intervention in existing technologies leads to distorted evaluation results, failing to accurately reflect the adaptability of signal control systems under critical traffic demand conditions and masking structural flaws in automatic timing schemes when dealing with actual traffic fluctuations. Focusing solely on intervention frequency or duration while ignoring phase-level operational details makes it impossible to pinpoint specific problems. Furthermore, the difficulty in distinguishing between occasional external disturbances and recurring timing scheme defects easily leads to misjudgments or omissions, resulting in evaluation results that are disconnected from actual optimization needs. Summary of the Invention

[0005] The purpose of this application is to provide a method for evaluating the effectiveness of traffic signal control, thereby addressing the shortcomings of the prior art and improving the accuracy and precision of the evaluation of traffic control signals.

[0006] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: In a first aspect, one embodiment of this application provides a method for evaluating the effectiveness of traffic signal control, the method comprising: Based on the traffic signal operation data within a preset historical time period, identify the operation records of multiple manual intervention events within the preset historical time period and the corresponding actual release records; Based on the operation records and corresponding actual release records of each manual intervention event, the resource offset is calculated to obtain the manual intervention requirement offset information for each manual intervention event. Based on the human intervention demand offset information of each of the aforementioned human intervention events, cluster analysis is performed on multiple human intervention events to obtain multiple human intervention pattern clusters; Based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters, calculate the intervention regularity index; Based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters, the green wave coordination cascade damage quantification is performed to obtain the green wave coordination damage index; Based on the human intervention demand offset information of multiple human intervention events, the intervention regularity index, and the green wave coordination damage index, the evaluation result of the traffic signal control effect within the preset historical time period is determined.

[0007] Optionally, the step of calculating the resource offset based on the operation record and corresponding actual release record of each manual intervention event to obtain the manual intervention requirement offset information for each manual intervention event includes: Based on the operation records of each of the aforementioned artificial intervention events, construct the micro-behavioral feature vector of the artificial intervention event; Based on the actual release records, obtain the actual release phase sequence corresponding to the manual intervention event; Based on the micro-behavioral feature vectors of each of the aforementioned human intervention events and the corresponding actual release phase sequences, resource offsets are calculated to obtain the human intervention demand offset information for each of the aforementioned human intervention events.

[0008] Optionally, the artificial intervention demand offset information includes: a phase offset vector, the intention to intervene, and an average stress index; The resource offset calculation based on the micro-behavioral feature vectors of each of the aforementioned human intervention events and the corresponding actual release phase sequences yields the human intervention demand offset information for each of the aforementioned human intervention events, including: Based on the micro-behavioral feature vector and the actual release phase sequence, determine the phase shift intensity of each of the artificial intervention events; The phase shift vector of each artificial intervention event is determined based on the phase shift intensity of each artificial intervention event; Based on the phase offset vector, manual operation type, and sequence of intervention phase changes of each of the aforementioned manual intervention events, the intention of the manual intervention for each of the aforementioned manual intervention events is determined; The average stress index of each human intervention event is determined based on the phase offset vector of each human intervention event, the intention of the human intervention, and the duration of the intervention.

[0009] Optionally, based on the human intervention demand offset information of each of the human intervention events, cluster analysis is performed on multiple human intervention events to obtain multiple human intervention pattern clusters, including: Based on the phase offset vector of each of the artificial intervention events and the intention of the artificial intervention, vectorization is performed based on the day-time ratio and the original scheme of each of the artificial intervention events to obtain the clustering input feature vector of each of the artificial intervention events; Based on the clustering input feature vectors of each of the aforementioned human intervention events, clustering analysis is performed on each of the aforementioned human intervention events according to preset classification rules to obtain multiple clusters of the aforementioned human intervention patterns.

[0010] Optionally, the step of calculating intervention regularity indicators based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters includes: Based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster, the temporal concentration and scheme correlation of each artificial intervention pattern cluster are determined. Based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster and the actual release phase sequence, the phase consistency of each artificial intervention pattern cluster is determined. The regularity indicators are determined based on the temporal concentration, phase consistency, and scheme correlation of each artificial intervention pattern cluster.

[0011] Optionally, the step of quantifying the green wave coordination cascade damage based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters to obtain the green wave coordination damage index includes: Based on the micro-behavioral feature vectors and actual release phase sequences of each of the artificial intervention events in the multiple artificial intervention mode clusters, the coordinated phase center offset of each of the artificial intervention events is determined; Based on the coordinated phase center offset of each manual intervention event and the planned data of upstream and downstream intersections, the arrival deviation and effective green wave loss of each manual intervention event are determined. The coordination recovery time for each of the artificial intervention events is determined based on the micro-behavioral feature vectors of each event, the coordination phase center offset, and the arrival deviation. The green wave coordination damage index is determined based on the micro-behavioral feature vectors of each artificial intervention event and the coordination recovery time.

[0012] Optionally, determining the evaluation result of the traffic signal control effect within the preset historical time period based on the manual intervention demand offset information of multiple manual intervention events, the intervention regularity index, and the green wave coordination damage index includes: The artificial intervention demand offset information, the intervention regularity index, and the green wave coordination damage index of multiple artificial intervention events are normalized to obtain normalized artificial intervention demand offset information, normalized intervention regularity index, and normalized green wave coordination damage index. Based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index, the basic pressure risk, structural risk, and coordination damage risk are determined. The assessment results of the traffic signal control effectiveness are determined based on the aforementioned basic pressure risk, structural risk, and coordination damage risk.

[0013] Optionally, determining the basic pressure risk, structural risk, and coordination damage risk based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index includes: The basic pressure risk is determined based on the normalized human intervention demand offset information; The structural risks are determined based on the normalized artificial intervention demand offset information and the normalized intervention regularity indicators. The coordination damage risk is determined based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index.

[0014] Optionally, determining the assessment result of the traffic signal control effectiveness based on the basic pressure risk, the structural risk, and the coordination damage risk includes: Based on the aforementioned basic pressure risk, structural risk, and coordination damage risk, a signal control health index is determined; The evaluation result of the traffic signal control effect is determined based on the signal control health index.

[0015] Optionally, the step of identifying operation records and corresponding actual release records of multiple manual intervention events within a preset historical time period based on traffic signal operation data within that period includes: Obtain the operational data of traffic signals within the preset historical time period; From the operational data, determine the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information for each manual intervention event; Based on the intervention time information, the manual operation type information, the original intervention plan information, and multiple actual intervention phase information, the operation record for each manual intervention event is determined; Based on the multiple actual intervention phase information, the actual release record of each of the artificial intervention events is determined.

[0016] Secondly, another embodiment of this application provides a traffic signal control effect evaluation device, the device comprising: The identification module is used to identify the operation records of multiple manual intervention events and the corresponding actual release records within the preset historical time period based on the operation data of traffic signals within the preset historical time period. The first calculation module is used to calculate the resource offset based on the operation record and the corresponding actual release record of each manual intervention event, so as to obtain the manual intervention requirement offset information of each manual intervention event. The analysis module is used to perform cluster analysis on multiple human intervention events based on the human intervention demand offset information of each human intervention event, and obtain multiple human intervention pattern clusters. The second calculation module is used to calculate intervention regularity indicators based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters. The quantification module is used to quantify the green wave coordination cascade damage based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters, and to obtain the green wave coordination damage index. The determination module is used to determine the evaluation result of the traffic signal control effect within the preset historical time period based on the manual intervention demand offset information of multiple manual intervention events, the intervention regularity index, and the green wave coordination damage index. Thirdly, another embodiment of this application provides a computer device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the traffic signal control effect evaluation method as described in any of the first aspects above.

[0017] Fourthly, another embodiment of this application provides a storage medium storing a computer program, which, when executed by a processor, performs the steps of the traffic signal control effect evaluation method as described in any of the first aspects above.

[0018] The beneficial effects of this application are: This application provides a method for evaluating the effectiveness of traffic signal control. Based on traffic signal operation data within a preset historical time period, it identifies the operation records and corresponding actual release records of multiple manual intervention events within that period. Based on these records, resource offsets are calculated to obtain the manual intervention demand offset information for each event. Cluster analysis is then performed on these events to obtain multiple manual intervention pattern clusters. Intervention regularity indicators are calculated based on the operation records and actual release records of each event within these clusters. Furthermore, green wave coordination cascade damage quantification is performed to obtain a green wave coordination damage indicator. Finally, based on the manual intervention demand offset information, intervention regularity indicators, and green wave coordination damage indicators, the evaluation results of the traffic signal control effectiveness within the preset historical time period are determined. This application transforms manual control behavior into key evidence for evaluating signal control effectiveness by leveraging the traffic demand characteristics behind the intervention events, thus achieving a shift from passive response to proactive diagnosis. This study constructs a system that starts by identifying human intervention events and their actual release sequences, calculates phase-level resource offsets to reveal structural mismatches between demand and solutions, then extracts recurring intervention patterns through cluster analysis of demand offsets, distinguishes between occasional disturbances and regular defects, and quantifies the cascading damage of such regular interventions to arterial green wave coordination. Finally, it integrates these three types of evidence—demand offsets, regularities, and coordination damage—to form a traceable health assessment result. This improves the accuracy and precision of the assessment and enhances the ability to continuously optimize traffic signals. Attached Figure Description To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 2This is a flowchart illustrating the process of determining the offset information of human intervention demand in a traffic signal control effect evaluation method provided in an embodiment of this application. Figure 3 A flowchart illustrating the process of determining the offset information of human intervention demand in another traffic signal control effect evaluation method provided in this application embodiment; Figure 4 A flowchart illustrating the process of determining a cluster of manual intervention modes in a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 5 A flowchart illustrating the process of determining intervention regularity indicators in a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 6 A flowchart illustrating the process of determining the green wave coordination damage index in a traffic signal control effect evaluation method provided in this application embodiment; Figure 7 A flowchart illustrating the process of determining the evaluation result in a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 8 A flowchart illustrating the determination of indicators in a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 9 A flowchart illustrating the process of determining the evaluation result in another traffic signal control effect evaluation method provided in this application embodiment; Figure 10 A flowchart illustrating the process of determining human intervention events in a traffic signal control effectiveness evaluation method provided in this application embodiment; Figure 11 This is a schematic diagram of the structure of a traffic signal control effect evaluation device provided in an embodiment of this application; Figure 12 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0021] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that the term "comprising" will be used in the embodiments of this application to indicate the presence of the features declared thereafter, but does not exclude the addition of other features.

[0023] With the development of urbanization and intelligent transportation, traffic signal control systems are crucial for improving road traffic efficiency and travel experience. Currently, fixed timing schemes based on historical data are still used, combined with green wave coordination control to reduce vehicle stops. However, traffic flow is characterized by randomness and suddenness, causing preset schemes to often fail to adapt to actual changes. When traffic pressure is too high, traffic police will make temporary adjustments through manual intervention. Existing assessments of traffic signal control effectiveness mostly rely on macroscopic indicators such as saturation and queue length, often excluding periods of manual intervention, making it difficult to accurately assess the effectiveness of traffic signal control. To address this, this application provides a method for evaluating the effectiveness of traffic signal control. Based on the operational data of traffic signals within a preset historical time period, it identifies the operation records of multiple manual intervention events within the preset historical time period and the corresponding actual release records, thereby determining the manual intervention demand offset information, intervention regularity indicators, and green wave coordination damage indicators for the manual intervention time, and determining the evaluation result of the traffic signal control effectiveness within the preset historical time period. The method in this application improves the accuracy of evaluating the effectiveness of traffic signal control.

[0024] To clearly describe the traffic signal control effect evaluation method provided in this application, the method is explained below with reference to several accompanying figures. The traffic signal control effect evaluation method in this application is to evaluate the traffic signal control effect of the same intersection within a preset time period.

[0025] Figure 1 This is a flowchart illustrating a traffic signal control effectiveness evaluation method provided in an embodiment of this application, as shown below. Figure 1 As shown, the method includes: Step 101: Based on the traffic signal operation data within the preset historical time period, identify the operation records of multiple manual intervention events and the corresponding actual release records within the preset historical time period.

[0026] The preset historical time period is the time period for evaluating the effectiveness of traffic signal control, which can be half a month, a month, a quarter, etc., and this application embodiment does not limit this. The operation data of traffic signals may include: the underlying operation log of the traffic signal controller, phase status records, control mode switching records, manual operation records, coordinated control plans, and basic phase schemes of intersections. The underlying operation log of the traffic signal controller records all underlying events such as hardware status, software tasks, communication, faults, and command execution in chronological order. The phase status record is used to record the actual light color status and duration of each phase of the traffic lights at each moment. A phase is a group of non-conflicting traffic flows that have the right of way at the same time. For example, phase A has no conflict between east-west straight traffic and east-west left / right turns, while the north-south direction is all red. The control mode switching record is the time point and triggering reason for switching between different operation modes. Control modes may include automatic control, sensor control, coordinated control, manual control, etc. The manual operation record is the log of intervention commands actively issued by the operator, which may come from the intersection control panel, the central platform, or a handheld terminal. The coordination control plan predefines coordination schemes for intersections, including common cycles, phase differences, green light ratios, coordinated phases, and time slot schedules. The basic phase scheme for an intersection is a fixed, unchanging underlying configuration, defining the number and order of phases, minimum / maximum green light time, yellow light time, all-red light time, and conflict tables. Manual intervention events include forced phase switching, phase locking, early termination, and phase insertion. Forced phase switching immediately terminates the currently running phase and forces a switch to a specified target phase; phase locking fixes a phase in a green light state, preventing automatic switching; early termination occurs when a phase is still within its green light period, but is manually terminated earlier than a preset end time; phase insertion temporarily inserts a phase not in the original scheme into the preset phase sequence. Actual traffic release records are the actual traffic release results executed after a manual intervention event occurs.

[0027] Optionally, the operation data of traffic signals within a preset historical time period are identified to obtain the manual intervention events within the preset historical time period, and the operation records of multiple manual intervention events and the corresponding actual release records are identified.

[0028] Step 102: Based on the operation records and corresponding actual release records of each manual intervention event, calculate the resource offset to obtain the manual intervention requirement offset information for each manual intervention event.

[0029] Among them, the manual intervention demand offset information is used to evaluate the offset information between the actual execution and the original plan during the manual intervention time.

[0030] Optionally, the operation record of each manual intervention event is aligned with its timestamp. Based on the start and end times and green light duration in the corresponding actual release record within that time period, the manual intervention requirement offset information for each manual intervention event is determined by comparing the start and end times and green light duration of the corresponding phase in the actual release record.

[0031] Step 103: Based on the human intervention demand offset information of each human intervention event, perform cluster analysis on multiple human intervention events to obtain multiple human intervention pattern clusters.

[0032] Cluster analysis involves grouping multiple human intervention events according to their characteristics, ensuring similarity within groups and significant differences between groups. The intervention behavior categories formed after clustering human intervention pattern clusters represent intervention scenarios with similar human intervention demand shift information.

[0033] Optionally, based on the features in the human intervention demand offset information of each human intervention event, cluster analysis is performed on multiple human intervention events to obtain multiple human intervention pattern clusters. The features in the human intervention demand offset information can be features related to the human intervention event, such as human intervention intention features or phase offset features.

[0034] Step 104: Calculate the intervention regularity index based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters.

[0035] Among them, the intervention regularity index is used to measure whether the human intervention events at the intersection are stable and predictable within a preset historical time period.

[0036] Optionally, based on each manual intervention event, the manual operation type, trigger time, target phase, and expected parameters are determined, and the actual execution results are extracted from the corresponding actual release records to determine the regularity characteristic quantity of each manual intervention event. The regularity characteristic quantities of all manual intervention events within the same manual intervention pattern cluster are summarized to obtain the regularity characteristic quantity of each manual intervention pattern cluster. The intervention regularity index is determined based on the regularity characteristic quantity of each manual intervention pattern cluster.

[0037] Step 105: Based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters, perform green wave coordination cascade damage quantification to obtain the green wave coordination damage index.

[0038] Among them, the Green Wave Coordination Cascade Damage Quantification refers to the degree of disruption caused by human intervention events to the designed green wave band, which is a coordinated system involving multiple intersections. The Green Wave Coordination Damage Index is used to measure the negative impact of intervention on green wave coordination; the higher the index, the more severe the disruption to the green wave.

[0039] Optionally, taking each manual intervention mode cluster as a unit, the occurrence time of all intervention events within the manual intervention mode cluster is extracted, and the preset common period, phase difference, and coordination phase scheme for that time period are obtained from the original scheme. For each manual intervention event, based on its operation record and the corresponding actual release record, the real-time offset of the manual intervention event to the coordination phase is determined. Combining the information of upstream and downstream intersections and the real-time offset of the manual intervention event to the coordination phase, the green wave coordination damage index for each manual intervention event is determined. Based on the green wave coordination damage index for each manual intervention event, the green wave coordination damage index for the same intersection within a preset time period is determined.

[0040] Step 106: Based on the information on the deviation of human intervention demand, the regularity index of intervention, and the green wave coordination damage index of multiple human intervention events, determine the evaluation results of the traffic signal control effect within the preset historical time period.

[0041] The evaluation results are used to assess the effectiveness of the current traffic signal control.

[0042] Optionally, the manual intervention demand deviation information, intervention regularity indicators, and green wave coordination damage indicators for multiple manual intervention events corresponding to all manual intervention events within a preset historical time period are normalized, transforming these indicators to the same scoring range. Different fusion weights are then assigned to the three types of scores, a weighted comprehensive score is calculated, and the evaluation result of the traffic signal control effect is determined based on the weighted score.

[0043] In this embodiment, based on traffic signal operation data within a preset historical time period, operation records and corresponding actual release records for multiple manual intervention events within the preset historical time period are identified. Resource offsets are calculated based on the operation records and corresponding actual release records of each manual intervention event to obtain manual intervention demand offset information for each event. Based on the manual intervention demand offset information of each event, cluster analysis is performed on the multiple manual intervention events to obtain multiple manual intervention pattern clusters. Intervention regularity indicators are calculated based on the operation records and corresponding actual release records of each manual intervention event within the multiple manual intervention pattern clusters. Green wave coordination cascade damage quantification is performed based on the operation records and corresponding actual release records of each manual intervention event within the multiple manual intervention pattern clusters to obtain green wave coordination damage indicators. Based on the manual intervention demand offset information, intervention regularity indicators, and green wave coordination damage indicators of multiple manual intervention events, the evaluation results of traffic signal control effectiveness within the preset historical time period are determined. This application transforms manual control behavior into key evidence for evaluating signal control effectiveness by leveraging the traffic demand characteristics behind manual intervention events, achieving a shift from passive response to proactive diagnosis. This study constructs a system that starts by identifying human intervention events and their actual release sequences, calculates phase-level resource offsets to reveal structural mismatches between demand and solutions, then extracts recurring intervention patterns through cluster analysis of demand offsets, distinguishes between occasional disturbances and regular defects, and quantifies the cascading damage of such regular interventions to arterial green wave coordination. Finally, it integrates these three types of evidence—demand offsets, regularities, and coordination damage—to form a traceable health assessment result. This improves the accuracy and precision of the assessment and enhances the traffic signal system's self-diagnosis and continuous optimization capabilities.

[0044] Based on the above embodiments, this application also provides a process for determining the offset information of human intervention demand in a traffic signal control effect evaluation method. Figure 2 This is a flowchart illustrating the process of determining the offset information of human intervention demand in a traffic signal control effect evaluation method provided in this application embodiment, as shown below. Figure 2 As shown, in step 102 above, resource offsets are calculated based on the operation records and corresponding actual release records of each manual intervention event to obtain the manual intervention requirement offset information for each event, including: Step 201: Based on the operation records of each human intervention event, construct the micro-behavioral feature vector of the human intervention event.

[0045] The micro-behavioral feature vector includes: intersection number, intervention start time, intervention end time, intervention duration, manual operation type, the original plan corresponding to the intervention occurrence time, the planned phase difference in the original plan, and the actual release phase sequence. The intersection number is the unique identifier of the traffic intersection where the intervention occurred; the intervention start time is the time when the manual intervention event first takes effect; the intervention end time is the time when the system reverts to automatic / coordinated mode; the intervention duration is the difference between the intervention end time and the intervention start time; the manual operation type is the operation type of the manual intervention event; the original plan corresponding to the intervention occurrence time is the control plan that should have been running automatically at the intervention start time; the planned phase difference in the original plan is the preset green light start time difference between this intersection and the upstream reference intersection in coordinated control, usually in seconds or phase angles; the actual release phase sequence is the actual phase sequence executed by the traffic signal during the intervention period and the actual release duration of each phase, recorded in time sequence form.

[0046] Optionally, the operation records of each manual intervention event are identified to obtain the intersection number. Intervention start time Intervention End Time Intervention duration Manual operation type The original plan corresponding to the time of intervention The planned phase difference in the original scheme and the actual release phase sequence This allows for the construction of micro-behavioral feature vectors for events involving human intervention. Where i represents the i-th intersection and k represents the k-th human intervention event.

[0047] Step 202: Based on the actual release records, obtain the actual release phase sequence corresponding to the manual intervention event. The actual release phase sequence includes the first... The actual release phase, the start time of the actual release of the phase, the end time of the actual release of the phase, the actual green light duration of the phase during manual intervention, the planned green light duration of the corresponding phase in the original plan, the green light duration offset caused by manual intervention, and the number of phases actually experienced in the manual intervention event. Each actual released phase is the r-th phase executed sequentially within the time window of a manual intervention event; the actual start time of phase release is the specific time when the green light of that phase turns on; the actual end time of phase release is the specific time when the green light of that phase ends (usually when the yellow light turns on); the actual green light duration of that phase during manual intervention is the difference between the actual end time and the actual start time of release, i.e., the actual number of green light seconds that phase actually receives; the planned green light duration for the corresponding phase in the original plan is the number of green light seconds originally planned to be allocated to that phase in the original plan corresponding to the time of the intervention event; the green light duration offset caused by manual intervention is the difference between the actual released green light duration and the planned green light duration; the number of phases actually experienced in the manual intervention event is the number of different phases actually executed from the start to the end of the intervention.

[0048] Optionally, the actual release record is identified to obtain the r-th actual release phase. and the corresponding phase actual release start time Phase actual release end time The actual green light duration for this phase during manual intervention. The planned green light duration for the corresponding phase in the original plan. Green light duration deviation caused by human intervention The number of phases actually experienced in events requiring human intervention According to the r-th actual release phase and the corresponding phase actual release start time Phase actual release end time The actual green light duration for this phase during manual intervention. The planned green light duration for the corresponding phase in the original plan. Green light duration deviation caused by human intervention The number of phases actually experienced in events requiring human intervention .

[0049] Constructing the actual release phase sequence .in, r represents the number of phases, i represents the i-th intersection, and k represents the k-th human intervention event.

[0050] Step 203: Calculate the resource offset based on the micro-behavioral feature vectors of each human intervention event and the corresponding actual release phase sequence to obtain the human intervention demand offset information for each human intervention event.

[0051] Optionally, resource offsets are calculated based on the micro-behavioral feature vectors of each human intervention event and the corresponding actual release phase sequence to obtain the green light resource offset for each human intervention event. Based on the green light resource offsets of each human intervention event, human intervention demand offset information is constructed for each human intervention event.

[0052] In this embodiment, by recording whether intervention occurs, the start and end times, operation type, duration, and actual phase sequence and green light duration changes of the intervention are captured. By comparing with the original automatic timing scheme, the green light resource offset for each phase is calculated, thereby revealing the manual intervention demand offset information. This transforms manual intervention from simply filtered-out abnormal noise into evidence reflecting the mismatch between the signal control scheme and actual traffic demand, improving the authenticity of the evaluation and the accuracy of the diagnosis, and laying a reliable data foundation for subsequent identification of structural timing defects.

[0053] Based on the above embodiments, the manual intervention demand offset information includes: a phase offset vector, the intention to intervene, and an average pressure index. This application also provides a process for determining the manual intervention demand offset information in another traffic signal control effectiveness evaluation method. Figure 3 A flowchart illustrating the process of determining the offset information of human intervention demand in another traffic signal control effect evaluation method provided in this application embodiment is shown below. Figure 3 As shown, in step 203 above, resource offsets are calculated based on the micro-behavioral feature vectors of each human intervention event and the corresponding actual release phase sequence to obtain the human intervention demand offset information for each human intervention event, including: Step 301: Determine the phase shift intensity of each human intervention event based on the micro-behavioral feature vector and the actual release phase sequence.

[0054] Among them, the phase offset intensity is used to measure the degree of deviation between the actual execution of a single phase under human intervention and the original scheme.

[0055] Optionally, the green light duration offset caused by human intervention in each phase of the actual release phase sequence of each human intervention event. Determine the phase shift intensity Where p is the total number of phases, r is the r-th phase, i is the i-th intersection, and k is the k-th intervention event.

[0056] Step 302: Determine the phase shift vector of each artificial intervention event based on the phase shift intensity of each artificial intervention event.

[0057] Optionally, based on the phase shift intensity of each human intervention event And the original plan corresponding to the time of intervention in the micro-behavioral feature vector. Mid-phase Total planned green light duration within the corresponding time window Determine the phase offset vector of each human intervention event. Where p is the total number of phases, r is the r-th phase, i is the i-th intersection, and k is the k-th intervention event. The original solution Mid-phase The total planned green light duration within the corresponding time window. To prevent extremely small constants with a denominator of zero. When When the value is positive, it indicates that manual intervention has increased the release resources for that phase; when... A negative value indicates that the phase was compressed or skipped due to human intervention.

[0058] Optionally, the phase offset vectors of each human intervention event are used to construct the human intervention demand offset vector. .in, For the intersection The number of phases.

[0059] Step 303: Determine the intention of each human intervention event based on the phase offset vector, human operation type, and sequence of intervention phase changes.

[0060] Optionally, offset vector based on human intervention requirements Manual operation type And the sequence of intervention phase changes to determine the intention of artificial intervention. When a certain inlet straight phase When the value is positive and extended for multiple consecutive cycles, it is determined to be a main direction queuing release type intervention; when a left turn phase at an inlet is inserted and the original plan is not released in the current cycle, it is determined to be a left turn overflow dissipation type intervention; when multiple phases are compressed and a certain coordinated phase is forcibly extended, it is determined to be a coordinated direction protection type intervention.

[0061] Step 304: Determine the average stress index of each human intervention event based on the phase offset vector, human intervention intention, and intervention duration.

[0062] Optionally, offset vector based on human intervention requirements Intention to human intervention and duration of intervention Determine the average stress index for each human intervention event. .in, Indicates the cycle length of the original scheme. The norm of the phase demand offset vector is represented by the first norm. The severity mapping value represents different intervention intentions. , , These are the weighting coefficients. Indicates the first Tianlukou A collection of events involving human intervention.

[0063] Optionally, based on the average stress index of each intervention event. Determine the average pressure index at the intersection. .in, The number of days within the preset time period.

[0064] In this embodiment, by constructing micro-behavioral feature vectors and actual release phase sequences, the spatiotemporal characteristics and phase resource adjustment details of each manual intervention are characterized, and the phase offset intensity and offset vector are calculated, realizing the transformation from raw operation logs to quantifiable traffic demand deviations. Combining intervention type and phase change sequence to identify the intention of manual intervention improves the ability to understand dynamic changes in traffic demand. Finally, the offset vector, intention, and intervention duration are integrated to calculate the average pressure index, avoiding the one-sidedness of traditional methods that rely solely on frequency or duration statistics.

[0065] Based on the above embodiments, this application also provides a process for determining the cluster of manual intervention modes in a traffic signal control effectiveness evaluation method. Figure 4 This is a flowchart illustrating the process of determining a cluster of manual intervention patterns in a traffic signal control effectiveness evaluation method provided in this application embodiment. Figure 4 As shown, in step 103 above, based on the human intervention demand offset information of each human intervention event, cluster analysis is performed on multiple human intervention events to obtain multiple human intervention pattern clusters, including: Step 401: Based on the phase offset vector of each human intervention event and the intention of human intervention, vectorize the data according to the daily time ratio and the original scheme of each human intervention event to obtain the clustering input feature vector of each human intervention event.

[0066] The daily time ratio refers to the number of minutes within a day when the intervention begins. For example, if an intervention occurs at 7:30 AM, and there are 24 hours in a day (24 × 60 = 1440 minutes), then 7:30 AM corresponds to 7 × 60 + 30 = 450 minutes. Therefore, the daily time ratio is 450 minutes.

[0067] Optionally, based on the phase offset vector of each human intervention event and the intention of human intervention Based on the day-time ratio and the original plans for each human intervention event. Vectorization is performed to obtain the clustering input feature vectors for each human intervention event.

[0068] Among them, the clustering input feature vector of each human intervention event Let i be the i-th intersection, and k be the k-th intervention event. This represents the total number of minutes in a day. One-hot encoding representing the intent of human intervention. This indicates the original automatic timing scheme number.

[0069] Step 402: Based on the clustering input feature vectors of each human intervention event, perform clustering analysis on each human intervention event according to the preset classification rules to obtain multiple human intervention pattern clusters.

[0070] The preset classification rule is to distinguish artificial intervention events that occur at the same time but have different phase requirements, and to associate and identify events that occur at different times but have the same phase requirements and intervention intentions.

[0071] Optionally, the clustering input feature vectors of each human intervention event are used. Phase offset vectors of each human intervention event in the data and the intention of human intervention Based on preset classification rules, density clustering, hierarchical clustering, or other unsupervised clustering methods are used to cluster the input feature vectors of each human intervention event. Clustering was performed to obtain multiple clusters of artificial intervention patterns. .

[0072] In this embodiment, by constructing clustered input feature vectors, the temporal regularity, phase preference, and scheme dependence of intervention behaviors can be comprehensively characterized. Cluster analysis of human intervention events can effectively identify clusters of human intervention patterns with similar time distribution, phase requirements, and intervention targets. This transforms scattered operation logs into analyzable regular behavior patterns, thereby improving the intelligence and refinement of traffic signal control effect evaluation.

[0073] Based on the above embodiments, this application also provides a process for determining intervention regularity indicators in a traffic signal control effectiveness evaluation method. Figure 5 This is a flowchart illustrating the process of determining intervention regularity indicators in a traffic signal control effectiveness evaluation method provided in this application embodiment, as shown below. Figure 5 As shown, in step 104 above, based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters, intervention regularity indicators are calculated, including: Step 501: Determine the temporal concentration and scheme correlation of each artificial intervention pattern cluster based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster.

[0074] Among them, time concentration indicates whether human intervention events within the same human intervention pattern cluster occur in close time periods. Scheme correlation indicates whether human intervention events within the same human intervention pattern cluster always occur under the same original automatic timing scheme.

[0075] Optionally, based on each cluster of artificial intervention models The proportion of day time in the micro-behavioral feature vectors of various artificial intervention events. Determine the time concentration .in, For time concentration; Indicates the first The first intersection One intervention cluster; The percentage of day time indicates the time when the intervention occurred; This represents the variance of the timing of intervention events within the same cluster of intervention patterns. This represents the maximum permissible variance or normalized baseline of human intervention events within the same cluster of human intervention patterns.

[0076] Optionally, based on each cluster of artificial intervention models The original scheme in the micro-behavioral feature vector of each artificial intervention event. Determine the relevance of the solutions .in, This represents the original plan prior to the intervention; This indicates a specific timing scheme number; This indicates that this type of intervention belongs to the program. The number of times; This represents the total number of interventions in the cluster.

[0077] Step 502: Determine the phase consistency of each artificial intervention pattern cluster based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster and the actual release phase sequence.

[0078] Optionally, based on each cluster of artificial intervention models Phase shift vectors in the micro-behavioral feature vectors of each human intervention event and human intervention pattern clusters Average phase offset vector Determine phase consistency .in, Indicates the first Phase offset vector of the second intervention; Indicates the first The average phase offset vector of the intervention; This indicates the difference between a single intervention and the average pattern of this type; This indicates phase consistency.

[0079] Step 503: Determine regularity indicators based on the temporal concentration, phase consistency, and scheme correlation of each artificial intervention model cluster.

[0080] Optionally, based on the time concentration of each cluster of artificial intervention modalities Phase consistency and the relevance of the plan Identify regularity indicators Among them, regularity indicators For the first The regularity of human intervention events at each intersection. This represents the proportion of this type of intervention event among all interventions. Indicates the intersection within the statistical period The complete set of all human intervention events.

[0081] In this embodiment, regularity indicators are calculated by analyzing the temporal concentration, phase consistency, and scheme correlation of human intervention events. This helps to identify whether human intervention occurs frequently and whether human intervention events recur under specific time periods, phase directions, and timing schemes. This distinguishes between occasional external disturbances and structural timing scheme defects, thus avoiding misjudgments caused by relying solely on intervention frequency or duration, and making the evaluation results more accurate. Based on the above embodiments, this application also provides a process for determining the green wave coordination damage index in a traffic signal control effect evaluation method. Figure 6 This is a flowchart illustrating the process of determining the green wave coordination damage index in a traffic signal control effect evaluation method provided in this application embodiment. Figure 6 As shown, in step 105 above, based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters, the damage quantification of green wave coordination cascade is performed to obtain the green wave coordination damage index, including: Step 601: Determine the coordinated phase center offset of each artificial intervention event based on the micro-behavioral feature vectors and actual release phase sequences of each artificial intervention event in multiple artificial intervention mode clusters.

[0082] Among them, the coordination phase center offset is used to indicate the amount by which the release center of the coordination phase deviates from the original plan after a human intervention event occurs.

[0083] Optionally, based on the actual release phase sequence of each human intervention event in multiple human intervention mode clusters. and micro-behavioral feature vectors The original scheme in the text determines the coordinated phase center offset of each human intervention event. .in, Indicates the green wave coordination phase of the main line; This indicates the actual release center time during the phase coordination process; This indicates the planned release center time of the coordinated phase in the original automatic scheme; This indicates the offset of the coordination phase center.

[0084] Step 602: Based on the coordinated phase center offset of each manual intervention event and the planned data of upstream and downstream intersections, determine the arrival deviation and effective green wave loss of each manual intervention event.

[0085] Arrival deviation includes upstream arrival offset and downstream arrival offset. Upstream arrival offset is the time deviation of a vehicle arriving at the current intersection relative to the green wave window of the current intersection; downstream arrival offset is the time deviation of a vehicle arriving at the downstream intersection relative to the green wave window of the downstream intersection. Effective green wave loss includes upstream effective green wave loss and downstream effective green wave loss. Upstream effective green wave loss is the effective time lost when a vehicle arrives at the current intersection exceeding the allowable range of the green wave window. Downstream effective green wave loss is the effective time lost when a vehicle arrives at the downstream intersection exceeding the allowable range of the green wave window. Planned data includes the travel time of vehicles arriving from adjacent intersections, the base time of the green wave window in the original intersection plan, and the allowable width of the effective green wave window at the intersection.

[0086] Optionally, based on the coordinated phase center offset of each manual intervention event and the planned data of the upstream intersection, the upstream arrival deviation and upstream effective green wave loss of each manual intervention event are determined.

[0087] Optionally, based on the coordinated phase center offset of each manual intervention event and the planned data of the downstream intersection, the downstream arrival deviation and downstream effective green wave loss of each manual intervention event are determined.

[0088] For example, this application uses a downstream intersection as an example to illustrate the coordination phase center offset of each human intervention event. The reference time for the green wave window in the original plan for the downstream intersection Intervention start time And the planned data for downstream intersections, including travel time from the current intersection to the downstream intersection. Determine the downstream arrival deviation .

[0089] According to downstream arrival deviation Downstream intersection Effective green wave allowable window width Determine the effective green wave loss downstream .in, This means that if a vehicle arrives early or late but is still within the green wave window, it is not considered to have caused actual green wave damage. Losses are only calculated when vehicles exceed the green wave window. For example, if the downstream green wave window half-width is 20 seconds and the convoy arrives 30 seconds late, the effective green wave loss is 10 seconds.

[0090] Step 603: Determine the coordination recovery time of each artificial intervention event based on the micro-behavioral feature vector, the coordination phase center offset, and the arrival deviation.

[0091] The coordination recovery time refers to how long it takes for the trunk line green wave coordination to return to stability after a manual intervention event.

[0092] Optionally, this application uses a downstream intersection as an example for illustration, and the calculation method for the upstream intersection is the same as that for the downstream intersection. This application does not impose any limitations on this aspect.

[0093] Optionally, based on the duration of intervention in the micro-behavioral feature vector of each human intervention event. Coordinate phase center offset and downstream arrival deviation Determine the coordinated recovery time for each human intervention event. .in, This indicates the impact of the duration of the intervention itself on recovery time; This indicates the effect of the coordinated phase offset on the recovery time; This indicates the impact of green wave loss at upstream and downstream intersections on recovery time; Indicates the relationship with the first A set of upstream and downstream intersections that have green wave coordination relationships; , , The weighting coefficient is determined based on the requirements of the current solution, and this application embodiment does not impose any restrictions on it.

[0094] Step 604: Determine the green wave coordination damage index based on the micro-behavioral characteristic vectors of each human intervention event and the coordination recovery time.

[0095] The green wave coordination damage index is the proportion of green wave failure time at an intersection caused by human intervention events to the total coordination control time. A higher green wave coordination damage index indicates a more severe disruption to the overall coordination of the main road caused by a single point of human intervention event.

[0096] Optionally, based on the duration of intervention in the micro-behavioral feature vector of each human intervention event. and coordination of recovery time Determine the green wave coordination damage index .in, For the first The proportion of green wave failure time caused by manual intervention at each intersection to the total coordinated control time. This indicates the total duration of coordinated control at the intersection.

[0097] In this embodiment, the micro-behavioral characteristics of manual intervention events and the actual release sequence are analyzed to quantify the phase center offset. Combined with the planned phase difference and travel time at upstream and downstream intersections, the vehicle arrival deviation and effective green wave loss are accurately calculated, thereby assessing the coordination recovery time and ultimately forming a green wave coordination damage index. This achieves an upgrade from point-to-line evaluation, accurately identifying key nodes and typical patterns that frequently interfere with green wave operation, and improving the stability and intelligent management level of the traffic signal system in dynamic environments.

[0098] Based on the above embodiments, this application also provides a process for determining the evaluation result in a traffic signal control effectiveness evaluation method. Figure 7 This is a flowchart illustrating the process of determining the evaluation result in a traffic signal control effectiveness evaluation method provided in an embodiment of this application, as shown below. Figure 7 As shown, in step 106 above, based on the information on the deviation of human intervention demand for multiple human intervention events, the regularity index of intervention, and the green wave coordination damage index, the evaluation results of the traffic signal control effect within the preset historical time period are determined, including: Step 701: Normalize the artificial intervention demand deviation information, intervention regularity indicators, and green wave coordination damage indicators for multiple artificial intervention events to obtain normalized artificial intervention demand deviation information, normalized intervention regularity indicators, and normalized green wave coordination damage indicators.

[0099] Optionally, avoid including human intervention pressure in the human intervention demand offset information for multiple human intervention events. Intervention regularity indicators and green wave coordination damage indicators Normalization is performed to obtain the normalized artificial intervention pressure. Normalized intervention regularity indicators and the normalized green wave coordinated damage index .in, This indicates a normalization function established based on historical samples, expert thresholds, or statistical quantiles, with the normalization result ranging from 0 to 1.

[0100] Step 702: Based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index, determine the basic pressure risk, structural risk, and coordination damage risk.

[0101] Among them, basic pressure risk is used to reflect the pressure of human intervention; structural risk is used to represent structural risk, requiring both high pressure and strong regularity; and coordinated damage risk is used to represent coordinated damage risk, requiring both high pressure, strong regularity, and a large green wave impact.

[0102] Optionally, based on the normalized human intervention demand offset information Normalized intervention regularity indicators and the normalized green wave coordinated damage index Determine basic pressure risks Structural risks and coordinating damage risks .

[0103] For example, if the pressure of manual intervention at an intersection is high but irregular, it may simply be due to an accident, construction, or a temporary event; if the pressure of manual intervention is high and regular, it is more likely that there is a structural defect in the timing scheme; if the pressure is high, regular, and the green wave impact rate is high, it indicates that the defect not only affects the individual intersection but also disrupts the trunk coordination system.

[0104] Step 703: Determine the assessment results of the traffic signal control effectiveness based on the basic pressure risk, structural risk, and coordination damage risk.

[0105] Optionally, the assessment result of the traffic signal control effect can be determined by weighted summation based on the basic pressure risk, structural risk, and coordination damage risk. Alternatively, the assessment result of the traffic signal control effect can be determined by using a pre-set risk assessment model based on the basic pressure risk, structural risk, and coordination damage risk. The pre-set risk assessment model is a model pre-trained using the basic pressure risk, structural risk, coordination damage risk, and assessment results.

[0106] In this embodiment, normalization eliminates the differences in dimensions and orders of magnitude between different indicators, realizes the unified comparability of multi-source heterogeneous data, enables the evaluation results to reflect the frequency and duration of human intervention, quantifies its cascading impact on trunk green wave coordination, and provides the evaluation results with a clear chain of evidence, which can distinguish between occasional disturbances and structural problems, thereby improving the accuracy of traffic signal control effect evaluation.

[0107] Based on the above embodiments, this application also provides a process for determining indicators in a traffic signal control effectiveness evaluation method. Figure 8 This is a flowchart illustrating the process of determining indicators in a traffic signal control effectiveness evaluation method provided in this application embodiment, as shown below. Figure 8As shown, in step 701 above, based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index, the basic pressure risk, structural risk, and coordination damage risk are determined, including: Step 801: Determine the basic pressure risk based on the normalized human intervention demand offset information.

[0108] Optionally, the normalized human intervention demand offset information can be used. As a basic pressure risk .

[0109] Step 802: Determine structural risks based on the normalized human intervention demand offset information and the normalized intervention regularity indicators.

[0110] Optionally, the pressure for human intervention in the normalized human intervention demand offset information can be included. and normalized intervention regularity indicators The product of these factors serves as structural risk. .

[0111] Step 803: Determine the coordination damage risk based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index.

[0112] Optionally, the pressure for human intervention in the normalized human intervention demand offset information can be included. Normalized intervention regularity indicators and the normalized green wave coordinated damage index The product of these factors serves as a means to coordinate the risk of damage. .

[0113] In this embodiment, the baseline pressure reflects the overall reliance on human intervention, which is then combined with regularity analysis to determine whether it is a structural timing defect. Finally, the impact of green wave damage assessment on trunk line coordination is introduced, achieving step-by-step identification. This avoids misjudging the nature of the problem based solely on the frequency or duration of intervention, and can more accurately distinguish between occasional disturbances and defects in the scheme that need optimization, thus improving the scientific rigor and practicality of the evaluation results.

[0114] Based on the above embodiments, this application also provides a process for determining the evaluation result in another traffic signal control effectiveness evaluation method. Figure 9 A flowchart illustrating the determination of evaluation results in another traffic signal control effectiveness evaluation method provided in this application embodiment is shown below. Figure 9 As shown, in step 703 above, based on the basic pressure risk, structural risk, and coordination damage risk, the assessment results of the traffic signal control effectiveness are determined, including: Step 901: Determine the signal control health index based on basic pressure risk, structural risk, and coordination damage risk.

[0115] The traffic signal control health index is the health index of the intersection. The higher the risk, the lower the traffic signal control health index; the mere existence of intervention pressure will result in a deduction of points; the regularity of the pressure will result in a further deduction of points; when the regular problem also disrupts the green wave, a severe deduction will be made; if structural problems and green wave damage are both obvious, the health index will be further reduced by coupling penalty terms.

[0116] Optionally, based on the underlying pressure risk Structural risks and coordinating damage risks Determine the health index of signal control .in, Indicates the first The health index of signal control at each intersection , , Indicates the weight of each risk item; This represents the coupling penalty term between structural defects and green wave damage.

[0117] Step 902: Determine the evaluation results of the traffic signal control effectiveness based on the signal control health index.

[0118] Optionally, the evaluation result of traffic signal control effectiveness is determined based on the signal control health index and preset health level thresholds. The preset health level thresholds include good, slightly risky, sub-healthy, and poor. (Signal Control Health Index) This indicates that the intersection is under good control. This indicates that the intersection's control status is at a low risk, meaning there is room for improvement. This indicates that the intersection's control status is sub-healthy, which is also a warning state. This indicates that the intersection's traffic control is in poor condition and requires significant optimization.

[0119] For example, when the underlying stress risk Higher and structural risks When the value is low, it indicates that there is pressure for human intervention but a stable pattern has not yet been formed. It is recommended to prioritize external disturbances such as accidents, construction, illegal parking, and temporary changes in traffic organization. Instead of directly adjusting the fixed timing scheme, it is suggested to introduce short-term sensing control, queue detection, event reporting linkage, or duty prompts.

[0120] When structural risks When the value is high and the cluster center time window is stable, it indicates that the original automatic scheme has a structural mismatch in a specific period, and should be based on the first... Average phase shift vector of intervention The phase green ratio for the corresponding time period is adjusted so that the new timing scheme can cover the window of high incidence of manual intervention in advance.

[0121] When structural risks Higher and first Average phase shift vector of intervention When the dominant component is concentrated at a certain entrance or a certain turning phase, it indicates that the traffic demand in that direction has been underestimated for a long time. The green light duration should be increased, the phase sequence adjusted, a protection phase set up, or the lane function configuration optimized for that phase.

[0122] When coordinating damage risks When the value is high, it indicates that single-point manual intervention has disrupted trunk line coordination. The maximum locking time of manual greening should be limited, and upstream and downstream coordination phase difference compensation, restoration transition cycle and regional linkage protection should be triggered simultaneously when manual takeover is carried out to reduce green wave breakage.

[0123] when Continuously below the severe warning threshold and , , When all values ​​are high, it indicates that simple signal parameter optimization may have reached its marginal utility limit. It is recommended to carry out traffic organization optimization, including reversible lanes, tidal flow lanes, entrance channelization, prohibition of partial turns, bus priority reconstruction, or intersection geometry modification.

[0124] In this embodiment, the signal control health index is calculated by comprehensively considering basic pressure risk, structural risk, and coordination damage risk, thus realizing a shift from single-indicator evaluation to a multi-level progressive evaluation. This improves the accuracy and scientific rigor of the assessment, enabling it to truly reflect the adaptability and stability of signal control in actual operation.

[0125] Based on the above embodiments, this application also provides a process for determining human intervention events in a traffic signal control effectiveness evaluation method. Figure 10 This is a flowchart illustrating the process of determining human intervention events in a traffic signal control effectiveness evaluation method provided in this application embodiment, as shown below. Figure 10 As shown, in step 101 above, based on the traffic signal operation data within a preset historical time period, the operation records of multiple manual intervention events within the preset historical time period and the corresponding actual release records are identified, including: Step 1001: Obtain the operation data of traffic signals within a preset historical time period.

[0126] The operational data can identify the control mode (automatic control, sensor control, coordinated control, manual control, etc.), cycle start and end times, phase number, phase start time, phase end time, actual green light duration, original automatic green light duration, coordinated scheme number, coordinated phase, planned phase difference, and manual operation type. Manual operation types include forced phase jump, phase locking, early termination, and phase insertion.

[0127] Optionally, the operation data of traffic signals within a preset historical time period can be determined through the operation log of the traffic signal controller.

[0128] Step 1002: Determine the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information for each artificial intervention event from the operational data.

[0129] The intervention time information includes the cycle start and end times, phase start time, and phase end time. The manual operation type information includes forced phase jump, phase lock, early termination, and movement during phase insertion. The original intervention plan information includes the original automatic plan green light duration. The information for multiple actual intervention phases includes the actual green light duration, coordination plan number, coordination phase, and planned phase difference.

[0130] Step 1003: Based on the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information, determine the operation record for each manual intervention event.

[0131] Optionally, the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information of each manual intervention event can be used as the operation record of each manual intervention event.

[0132] Step 1004: Based on multiple actual intervention phase information, determine the actual release record for each manual intervention event.

[0133] Optionally, the actual intervention phase information of each human intervention event can be used as a sequence set to determine the actual release record of each human intervention event.

[0134] In this embodiment, by systematically extracting key information such as intervention time, operation type, original timing scheme, and actual release phase from historical operational data, the originally scattered operation logs are transformed into standardized operation records and actual release records, achieving refined modeling of manual intervention behavior. This accurately reconstructs the specific execution process of each manual intervention and provides a reliable data foundation for subsequent resource offset analysis, pattern identification, and coordination impact assessment, thereby improving the objectivity and scientific rigor of traffic signal control effectiveness evaluation.

[0135] Based on the same inventive concept, this application also provides a traffic signal control effect evaluation device corresponding to the traffic signal control effect evaluation method. Since the principle of the device in this application is similar to the traffic signal control effect evaluation method described above in this application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0136] Figure 11 This is a schematic diagram of the structure of a traffic signal control effect evaluation device provided in an embodiment of this application, as shown below. Figure 11 As shown, the device includes: The identification module 1101 is used to identify the operation records of multiple manual intervention events and the corresponding actual release records within the preset historical time period based on the operation data of traffic signals within the preset historical time period. The first calculation module 1102 is used to calculate the resource offset based on the operation record and the corresponding actual release record of each manual intervention event, so as to obtain the manual intervention requirement offset information of each manual intervention event. Analysis module 1103 is used to perform cluster analysis on multiple human intervention events based on the human intervention demand offset information of each human intervention event to obtain multiple human intervention pattern clusters; The second calculation module 1104 is used to calculate intervention regularity indicators based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters. The quantification module 1105 is used to quantify the green wave coordination cascade damage based on the operation records and corresponding actual release records of each manual intervention event in multiple manual intervention mode clusters, and to obtain the green wave coordination damage index. The determination module 1106 is used to determine the evaluation results of traffic signal control effect within a preset historical time period based on the information on the deviation of human intervention demand, intervention regularity indicators, and green wave coordination damage indicators of multiple human intervention events. In one possible implementation, the first calculation module 1102 is specifically used to: construct a micro-behavioral feature vector of each human intervention event based on the operation records of each human intervention event; Based on the actual release records, obtain the actual release phase sequence corresponding to the manual intervention event; Resource offsets are calculated based on the micro-behavioral feature vectors of each human intervention event and the corresponding actual release phase sequence to obtain the human intervention demand offset information for each human intervention event.

[0137] In one possible implementation, the artificial intervention demand offset information includes: a phase offset vector, the intention of artificial intervention, and an average pressure index; the first calculation module 1102 is specifically used to: determine the phase offset intensity of each artificial intervention event based on the micro-behavioral feature vector and the actual release phase sequence. The phase shift vector of each artificial intervention event is determined based on the phase shift intensity of each artificial intervention event; Based on the phase offset vector, manual operation type, and sequence of intervention phase changes for each manual intervention event, the intention of the manual intervention for each event is determined. The average stress index for each intervention event is determined based on the phase offset vector, the intention of the intervention, and the duration of the intervention.

[0138] In one possible implementation, the analysis module 1103 is specifically used to: vectorize the phase offset vector of each human intervention event and the intention of human intervention based on the day-time ratio and the original scheme of each human intervention event to obtain the clustering input feature vector of each human intervention event. Based on the clustering input feature vectors of each human intervention event, clustering analysis is performed on each human intervention event according to preset classification rules to obtain multiple human intervention pattern clusters.

[0139] In one possible implementation, the second calculation module 1104 is specifically used to: determine the temporal concentration and scheme correlation of each artificial intervention pattern cluster based on the micro-behavioral feature vector of each artificial intervention event in each artificial intervention pattern cluster. Based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster and the actual release phase sequence, the phase consistency of each artificial intervention pattern cluster is determined. Based on the temporal concentration, phase consistency, and correlation of each artificial intervention model cluster, regularity indicators are determined.

[0140] In one possible implementation, the quantization module 1105 is specifically used to: determine the coordinated phase center offset of each artificial intervention event based on the micro-behavioral feature vectors of each artificial intervention event in multiple artificial intervention mode clusters and the actual release phase sequence. Based on the coordinated phase center offset of each manual intervention event and the planned data of upstream and downstream intersections, the arrival deviation and effective green wave loss of each manual intervention event are determined. Based on the micro-behavioral characteristic vectors, coordination phase center offset, and arrival deviation of each human intervention event, the coordination recovery time of each human intervention event is determined. Based on the micro-behavioral characteristic vectors of each human intervention event and the coordination recovery time, the green wave coordination damage index is determined.

[0141] In one possible implementation, the determining module 1106 is specifically used to: normalize the artificial intervention demand offset information, intervention regularity index and green wave coordination damage index of multiple artificial intervention events respectively, to obtain normalized artificial intervention demand offset information, normalized intervention regularity index and normalized green wave coordination damage index. Based on the normalized artificial intervention demand deviation information, the normalized intervention regularity indicators, and the normalized green wave coordination damage indicators, the basic pressure risk, structural risk, and coordination damage risk are determined. The assessment results of the traffic signal control effectiveness are determined based on the basic pressure risk, structural risk, and coordination damage risk.

[0142] In one possible implementation, the determining module 1106 is specifically used to: determine the basic pressure risk based on the normalized human intervention demand offset information; Structural risks are identified based on the normalized information on the deviation of human intervention demand and the normalized indicators of intervention regularity. Based on the normalized artificial intervention demand deviation information, the normalized intervention regularity index, and the normalized green wave coordination damage index, the coordination damage risk is determined.

[0143] In one possible implementation, the determining module 1106 is specifically used to: determine the signal control health index based on the basic pressure risk, structural risk, and coordination damage risk; The assessment results of the traffic signal control effectiveness are determined based on the signal control health index.

[0144] In one possible implementation, the identification module 1101 is specifically used to: acquire the operation data of traffic signals within a preset historical time period; The intervention time, manual operation type, original intervention plan, and multiple actual intervention phases for each manual intervention event were determined from the operational data. Based on the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information, determine the operation records for each manual intervention event; Based on multiple actual intervention phase information, the actual release records for each manual intervention event are determined.

[0145] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.

[0146] This application also provides a computer device. Figure 12This application provides a schematic diagram of the structure of a computer device, as shown in the embodiment of the present application. Figure 12 As shown, the system includes a processor 1201 and a memory 1202, and optionally, a bus 1203. The memory 1202 stores machine-readable instructions executable by the processor 1201. When the computer device is running, the processor 1201 and the memory 1202 communicate via the bus 1203. When the machine-readable instructions are executed by the processor 1201, the steps of the traffic signal control effect evaluation method described above are performed.

[0147] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the traffic signal control effect evaluation method described above.

[0148] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and devices described above can be referred to the corresponding processes in the method embodiments, and will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some communication interfaces; the indirect coupling or communication connection of devices or modules can be electrical, mechanical, or other forms.

[0149] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0150] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for evaluating the effectiveness of traffic signal control, characterized in that, The method includes: Based on the traffic signal operation data within a preset historical time period, identify the operation records of multiple manual intervention events within the preset historical time period and the corresponding actual release records; Based on the operation records and corresponding actual release records of each manual intervention event, the resource offset is calculated to obtain the manual intervention requirement offset information for each manual intervention event. Based on the human intervention demand offset information of each of the aforementioned human intervention events, cluster analysis is performed on multiple human intervention events to obtain multiple human intervention pattern clusters; Based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters, calculate the intervention regularity index; Based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters, the green wave coordination cascade damage quantification is performed to obtain the green wave coordination damage index; Based on the human intervention demand offset information of multiple human intervention events, the intervention regularity index, and the green wave coordination damage index, the evaluation result of the traffic signal control effect within the preset historical time period is determined.

2. The method according to claim 1, characterized in that, The step involves calculating resource offsets based on the operation records and corresponding actual release records of each manual intervention event to obtain the manual intervention requirement offset information for each event, including: Based on the operation records of each of the aforementioned artificial intervention events, construct the micro-behavioral feature vector of the artificial intervention event; Based on the actual release records, obtain the actual release phase sequence corresponding to the manual intervention event; Based on the micro-behavioral feature vectors of each of the aforementioned human intervention events and the corresponding actual release phase sequences, resource offsets are calculated to obtain the human intervention demand offset information for each of the aforementioned human intervention events.

3. The method according to claim 2, characterized in that, The artificial intervention demand offset information includes: phase offset vector, artificial intervention intention, and average stress index; The resource offset calculation based on the micro-behavioral feature vectors of each of the aforementioned human intervention events and the corresponding actual release phase sequences yields the human intervention demand offset information for each of the aforementioned human intervention events, including: Based on the micro-behavioral feature vector and the actual release phase sequence, determine the phase shift intensity of each of the artificial intervention events; The phase shift vector of each artificial intervention event is determined based on the phase shift intensity of each artificial intervention event; Based on the phase offset vector, manual operation type, and sequence of intervention phase changes of each of the aforementioned manual intervention events, the intention of the manual intervention for each of the aforementioned manual intervention events is determined; The average stress index of each human intervention event is determined based on the phase offset vector of each human intervention event, the intention of the human intervention, and the duration of the intervention.

4. The method according to claim 3, characterized in that, The step involves clustering multiple human intervention events based on their human intervention demand offset information to obtain multiple human intervention pattern clusters, including: Based on the phase offset vector of each of the artificial intervention events and the intention of the artificial intervention, vectorization is performed based on the day-time ratio and the original scheme of each of the artificial intervention events to obtain the clustering input feature vector of each of the artificial intervention events; Based on the clustering input feature vectors of each of the aforementioned human intervention events, clustering analysis is performed on each of the aforementioned human intervention events according to preset classification rules to obtain multiple clusters of the aforementioned human intervention patterns.

5. The method according to claim 2, characterized in that, The step of calculating intervention regularity indicators based on the operation records and corresponding actual release records of each manual intervention event in the multiple manual intervention mode clusters includes: Based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster, the temporal concentration and scheme correlation of each artificial intervention pattern cluster are determined. Based on the micro-behavioral feature vectors of each artificial intervention event in each artificial intervention pattern cluster and the actual release phase sequence, the phase consistency of each artificial intervention pattern cluster is determined. The regularity indicators are determined based on the temporal concentration, phase consistency, and scheme correlation of each artificial intervention pattern cluster.

6. The method according to claim 2, characterized in that, The step involves quantifying the green wave coordination cascade damage based on the operation records and corresponding actual release records of each manual intervention event within the multiple manual intervention mode clusters, to obtain a green wave coordination damage index, including: Based on the micro-behavioral feature vectors and actual release phase sequences of each of the artificial intervention events in the multiple artificial intervention mode clusters, the coordinated phase center offset of each of the artificial intervention events is determined; Based on the coordinated phase center offset of each manual intervention event and the planned data of upstream and downstream intersections, the arrival deviation and effective green wave loss of each manual intervention event are determined. The coordination recovery time for each of the artificial intervention events is determined based on the micro-behavioral feature vectors of each event, the coordination phase center offset, and the arrival deviation. The green wave coordination damage index is determined based on the micro-behavioral feature vectors of each artificial intervention event and the coordination recovery time.

7. The method according to claim 1, characterized in that, The step of determining the evaluation result of traffic signal control effectiveness within the preset historical time period based on the manual intervention demand offset information of multiple manual intervention events, the intervention regularity index, and the green wave coordination damage index includes: The artificial intervention demand offset information, the intervention regularity index, and the green wave coordination damage index of multiple artificial intervention events are normalized to obtain normalized artificial intervention demand offset information, normalized intervention regularity index, and normalized green wave coordination damage index. Based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index, the basic pressure risk, structural risk, and coordination damage risk are determined. The assessment results of the traffic signal control effectiveness are determined based on the aforementioned basic pressure risk, structural risk, and coordination damage risk.

8. The method according to claim 7, characterized in that, The determination of basic pressure risk, structural risk, and coordination damage risk based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index includes: The basic pressure risk is determined based on the normalized human intervention demand offset information; The structural risks are determined based on the normalized artificial intervention demand offset information and the normalized intervention regularity indicators. The coordination damage risk is determined based on the normalized artificial intervention demand offset information, the normalized intervention regularity index, and the normalized green wave coordination damage index.

9. The method according to claim 7, characterized in that, The assessment result of determining the effectiveness of the traffic signal control based on the basic pressure risk, the structural risk, and the coordination damage risk includes: Based on the aforementioned basic pressure risk, structural risk, and coordination damage risk, a signal control health index is determined; The evaluation result of the traffic signal control effect is determined based on the signal control health index.

10. The method according to claim 1, characterized in that, The step of identifying operation records and corresponding actual release records of multiple manual intervention events within a preset historical time period based on traffic signal operation data includes: Obtain the operational data of traffic signals within the preset historical time period; From the operational data, determine the intervention time information, manual operation type information, original intervention plan information, and multiple actual intervention phase information for each manual intervention event; Based on the intervention time information, the manual operation type information, the original intervention plan information, and multiple actual intervention phase information, the operation record for each manual intervention event is determined; Based on the multiple actual intervention phase information, the actual release record of each of the artificial intervention events is determined.

Images