Traffic signal optimization method based on parameter calibration and improved webster timing method

Abstract

The application provides a traffic signal optimization method based on parameter calibration and improved Webster timing method, the method comprising: arranging detection coils in each lane in front of a stop line at an intersection, collecting vehicle passing information and storing locally; based on traffic flow theory analysis, calibrating saturated headway and calculating saturated flow of each lane according to the vehicle passing information, reducing the influence of oversaturated traffic flow on signal timing; for phase overlap and stage secondary release, the concept of phase residual flow ratio is proposed to solve the adaptation problem of Webster timing method under the signal timing requirements of the phase overlap and the stage secondary release; in the signal timing self-optimization link, if the signal timing meets the minimum green constraint relationship, the signal timing ends; if the signal timing does not meet the minimum green constraint relationship, the minimum green adjustment process is entered; when real-time optimization is performed, the updated signal timing is executed in the next cycle and the vehicle passing information is recorded again; when timing optimization is performed, adjustment is performed within a certain fixed time period based on historical data. The problems of unreasonable setting of basic parameters and limited use of traditional timing methods in actual application scenarios are solved, and a fast, accurate and stable traffic signal adaptive optimization system is established.

Description

Technical Field

[0001] This invention relates to the field of urban intelligent transportation technology, specifically to a traffic signal optimization method based on parameter calibration and an improved Webster timing method. Background Technology

[0002] Against the backdrop of the gradual development of new technologies and systems such as artificial intelligence and vehicle-road cooperation in the field of intelligent transportation, intersection traffic signal control remains an indispensable and crucial aspect of any intelligent transportation system and an essential element for improving traffic management. Intersection signal timing optimization methods that can meet the needs of complex scheme designs and traffic flow characteristics have significant application value. However, in practical timing optimization operations and system applications, there is a debate regarding various timing optimization paths.

[0003] Traditional signal timing methods based on traffic flow theory suffer from problems such as unreasonable basic parameter settings and limited application scenarios. Therefore, they have been widely replaced in signal timing optimization systems by methods based on reinforcement learning and optimization techniques. However, signal timing based on traffic flow theory and its deterministic output can provide traffic engineers with immediate optimization feedback, enabling rapid location and adjustment of signal timing parameters. Its near-zero computational requirements also save significant hardware computing costs in the construction of edge optimization systems or large-scale central optimization systems.

[0004] As the most representative timing model based on traffic flow theory, the classic Webster timing method has limitations in phase scheme structure and intersection saturation, making it difficult to adapt to varied and complex application scenarios and phase scheme designs. In actual scheme design, phase overlap (releasing the same turn in two consecutive phases) and secondary release (secondary release based on the discrete traffic flow design of parking aggregation) are very common, which often renders the Webster timing method useless or unable to output correct results in actual optimization systems.

[0005] Therefore, it is of great significance to address the problems of unreasonable basic parameter settings and limited use of the traditional Webster timing method in practical application scenarios, improve its engineering application capabilities and applicability, and provide algorithmic support for establishing a fast, accurate and stable traffic signal timing optimization system. Summary of the Invention

[0006] The technical problem this invention aims to solve is how to address the issues of unreasonable basic parameter settings and limited use in practical applications of the traditional Webster timing method, thereby providing algorithmic support for establishing a fast, accurate, and stable traffic signal timing adaptive optimization system.

[0007] To address the aforementioned technical problems, this invention provides a traffic signal optimization method based on parameter calibration and an improved Webster timing method, the method comprising:

[0008] S1. Install detection coils in each lane before the stop line at the intersection to collect vehicle information and store it locally;

[0009] S2. Based on the vehicle passage information, saturated headway is calibrated and saturated flow of each lane is calculated to reduce the impact of oversaturated traffic flow on signal timing.

[0010] S3. Calculation of Stage Flow Ratio Based on Phase Remaining Flow Ratio: The concept of the phase remaining flow ratio is proposed for phase overlap and stage secondary release, addressing the adaptation problem of the Webster timing method under the signal timing requirements of phase overlap and stage secondary release, wherein:

[0011] Regarding the phase overlap: a stage-independent phase set I is introduced, which is the complement of the intersection of the current stage and the next stage phases in the current stage phase set. The flow ratio is allocated by traversing the stages in the phase scheme. The method for allocating the stage flow ratio is as follows:

[0012] Y = max{y′ i y′ j ,…},i,j∈I

[0013] Among them, y i ′ is the remaining flow ratio of phase i in the current stage;

[0014] After allocating the current phase flow ratio, update the remaining flow ratio of each phase, where the remaining flow ratio of each phase satisfies the formula:

[0015] y′ i =max{0, y′ i -Y k δ ik}

[0016] Where, when phase i is included in stage k, δ ik Equal to 1, otherwise δ ik Equals 0;

[0017] For the second release in the aforementioned stage: the flow ratio of the stage is allocated in reverse, and the average value is calculated with the result of the forward flow ratio allocation;

[0018] After the phase flow ratio allocation is completed, if all phase remaining flow ratios are zero, the signal timing phase begins; if any phase remaining flow ratio is non-zero, the current phase scheme is not applicable to the current traffic conditions.

[0019] S4. In the signal timing autonomous optimization stage, if the signal timing satisfies the minimum green constraint, the signal timing ends; if the signal timing does not satisfy the minimum green constraint, the minimum green adjustment process begins; wherein:

[0020] The minimum green constraint relationship is:

[0021]

[0022] Among them, g s This is the effective green light time for a certain period, l s It is a phase of lost time. It is the sum of the non-green light times of the phase.

[0023] The algorithm for the minimum green adjustment satisfies:

[0024]

[0025] Among them, C max It is the upper limit threshold of the signal period, set C w It is the signal period during the minimum green adjustment process:

[0026] If C w ≥C max Then, after allocating the effective green light time for the aforementioned stage, the minimum green light adjustment will end.

[0027] If C w <C max Then, except for the stage of minimum green adjustment, the remaining stage time is allocated according to the flow ratio.

[0028] Furthermore, the method also includes:

[0029] When the method is applied to road sections with large traffic flow changes, the updated signal timing is executed in the next cycle, and the vehicle passage information is recorded based on the updated signal timing to achieve real-time optimization of traffic signals;

[0030] When the method is applied to road sections with long-term stable traffic, the vehicle passage information is updated within a fixed time period based on historical data to optimize traffic signal timing.

[0031] Furthermore, the vehicle passage information in S1 includes lane lid, vehicle vid, vehicle type type, passage status, and passage time.

[0032] Furthermore, the vehicle passage information is filtered, which includes deleting the vehicle passage information that intermittently affects vehicles, and using the intermittently affecting vehicles as breakpoints, splitting the periodic vehicle passage of each lane into multiple sets of data.

[0033] Furthermore, the method for calibrating the saturation flow rate in S2 is as follows:

[0034]

[0035] Where n is satisfied h vid When the number of samples (n) is less than the minimum sample size, the saturation flow rate of the lane is taken as the design value, h. vid It is the time distance between adjacent vehicle heads, t vid The time it takes for each vehicle to pass through the detection coil, l is the set vehicle length, w is the set width of the detection coil, and h... s This represents the average distance between the front of the vehicle when it is parked.

[0036] Furthermore, given that the Webster timing method requires the sum of the stage flow ratios to be no greater than 0.9 and is only applicable to signal timing in unsaturated traffic, step S4 also involves flow ratio reset. The method for resetting the flow ratio is as follows:

[0037] When ∑Y>0.9, the flow ratio Y of all stages is redistributed according to their respective proportions of the whole, and the redistribution method is as follows:

[0038]

[0039] When ∑Y≤0.9, Y′=Y.

[0040] Furthermore, S4 also involves the total signal loss time L, the initial signal period C0, and the effective green light time g of the stage. s Calculation:

[0041] The formula for calculating the total signal loss time L is:

[0042] L=∑l s =∑max(rt p +l p )

[0043] Among them, rt p It is the sum of the red light times of the phase within the stage, l p It is the phase initiation loss;

[0044] The formula for calculating the initial signal period C0 is:

[0045]

[0046] Where L is the total signal loss time, Y′ s It is the flow ratio mentioned;

[0047] The effective green light time during the aforementioned stage is gs The calculation formula is:

[0048]

[0049] Where C0 is the initial signal period, L is the total signal loss time, and Y′ s It is the flow ratio.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] This invention addresses the problems of unreasonable basic parameter settings and limited use of the traditional Webster timing method in practical application scenarios by calibrating saturation flow rate and improving the Webster timing method to adapt to non-standard phase schemes and oversaturated traffic flow conditions. Furthermore, it enables traffic signal timing optimization or real-time autonomous optimization based on application scenarios. Attached Figure Description

[0052] Figure 1 This is a flowchart of the present invention;

[0053] Figure 2 This is a schematic diagram illustrating the intermittent impact on vehicles during the data filtering step of this invention.

[0054] Figure 3 This is a schematic diagram of the vehicle headway in this invention;

[0055] Figure 4 This is a diagram of the signal control topology in this invention;

[0056] Figure 5 This is a schematic diagram illustrating the collection of vehicle passage information in one embodiment of the present invention;

[0057] Figure 6 This is a comparative schematic diagram of a standard phase scheme and a phase scheme with overlap in one embodiment of the present invention. Detailed Implementation

[0058] To make the technical solutions and effects of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0059] This invention provides a traffic signal optimization method based on parameter calibration and an improved Webster timing method. The method includes traffic data acquisition, saturation flow calibration, flow ratio allocation, and signal timing optimization. Based on this, an autonomous signal timing optimization workflow can be formed at the signal cycle level or within a specified time range. The flow ratio allocation and signal timing optimization methods are applicable to signal timing optimization in any scenario. The following details the traffic data acquisition, saturation flow calibration, flow ratio allocation, and signal timing optimization methods, such as... Figure 1 As shown:

[0060] 1. Vehicle passage information collection and storage

[0061] Loop detectors or virtual loop detectors based on video and radar are deployed in each lane before the stop line at the intersection to collect and store vehicle information locally. Vehicle information includes lane ID, vehicle VID, vehicle type, vehicle status (vehicle entering or leaving the detection loop), and vehicle time (in milliseconds).

[0062] Within the set statistical period [t] s , t e Within [t], extract the time within [t] s , t e The vehicle passage records of all lanes in the section are used as the statistical data for this period.

[0063] For each lane, count the number of vehicle records that entered the detection loop for each lane, convert them to the number of cars based on vehicle type, and use this as the traffic volume q for the statistical period. lia .

[0064] 2. Saturation flow calibration based on vehicle passage characteristic analysis

[0065] Based on traffic flow theory analysis, the saturation headway is calibrated according to vehicle passage information and the saturation flow of each lane is calculated to reduce the impact of oversaturated traffic flow on signal timing.

[0066] (1) Data filtering

[0067] Considering the intermittent flow characteristics of signalized intersections, it is first necessary to eliminate the impact of red light blocking and congestion on continuous flow recognition. Comparing the two, the stopping time caused by critical congestion is less than the red light blocking time. With a preset threshold of vehicle stationary speed v, vehicle length l, and detection coil width w, the intermittent flow time threshold can be set to (l+w) / v. Checking adjacent vehicle passage records, there exists a t... vid，1 -t vid，0 >(l+w) / v or t vid+1，0 -t vid，1 If >(l+w) / v, then let the vidth vehicle be the intermittently affected vehicle.

[0068] Further, based on the determination of intermittent impact on vehicles, the statistical period [t] for each lane is calculated. s , t e The vehicle passage records within the [database name] are cleaned and reorganized, deleting passage records that intermittently affect vehicles, and dividing the remaining passage records into multiple groups of statistical data based on breakpoints. For example... Figure 2 As shown, gray represents vehicles that are intermittently affected.

[0069] (2) Calculate the headway

[0070] like Figure 3 As shown, for each set of statistical data, starting with the record of a vehicle entering the detection loop as "status", the time it takes for each vehicle to pass through the loop and its headway relative to the next vehicle in the workshop are calculated. The headway relative to the next vehicle satisfies the formula:

[0071] t vid =t vid，1 -t vid，0

[0072] h vid =t vid+1，0 -t vid，0

[0073] Among them, h is set vid The time distance between the front of the vehicle, t vid It is the time it takes for each vehicle to pass through the detection coil.

[0074] (3) Saturation flow rate calibration

[0075] Based on Greenshields' speed-density linear relationship model, with parking density as K... m Speed ​​limits at city intersections as V f It can be seen that the expected speed of the continuous flow with the largest flow rate is V. f / 2, expected density is Where h s It is the average distance between the front of the vehicle when it is parked.

[0076] Considering the discreteness of actual traffic flow, when the vehicle speed is high, the vehicle speed and spacing tend to be discretely distributed. Therefore, traffic flow with a speed lower than the expected speed is used as the identification object of continuous flow saturation flow, that is, greater than the expected density.

[0077] Time distance between adjacent train heads h vid Time t between passing the coil and the vehicle in front vid At the desired density, the following relationship exists:

[0078]

[0079] That is, the higher the density, The smaller the value, the better, according to the filtering rules for identifying objects of continuous flow saturation flow, when the following relationship is satisfied:

[0080]

[0081] If the vehicle is determined to be following another vehicle continuously, the headway can be used for saturation flow calibration.

[0082] The method for determining the saturation flow rate for each lane is as follows:

[0083]

[0084] Where n is satisfied h vid The number of samples is such that when the number of samples n is less than the minimum sample size, the saturation flow rate of the lane is taken as the design value.

[0085] Saturation flow calibration identifies continuous traffic segments in basic vehicle data and uses saturation headway that meets the characteristics of continuous operation to calibrate the basic parameters of signal timing based on traffic flow model selection. This effectively identifies the impact of mixed traffic flow and congested traffic flow on intersection efficiency and reduces the error between theoretical model output and actual application requirements.

[0086] 3. Calculation of stage flow ratio based on remaining flow ratio

[0087] (1) Phase flow ratio

[0088] Based on intersection channelization and traffic light settings, the relationship between controllable signal phase and phase-lane correlation is clarified, and the phase-flow ratio is generated.

[0089] In this invention, the phase is the smallest control unit at the intersection, and there is no overlapping traffic flow between phases.

[0090] The phase flow ratio y is equal to the ratio of the phase flow rate to the phase saturation flow rate, i.e.

[0091]

[0092] Where m is the number of phase-associated lanes, and α is the saturation index of each phase assigned according to the control objective.

[0093] (2) Phase flow ratio allocation

[0094] Furthermore, by combining the signal scheme design to complete the phased flow ratio allocation, the phases in this invention are combinations of phases and carriers of release time. The phases are combined in a certain order to form the scheme. The signal control topology is as follows: Figure 4 As shown,

[0095] The traditional Webster timing method is only applicable to standard phase schemes (without secondary release of phases) and has the drawback that manual processes can be transformed based on experience, which is difficult for the system to implement.

[0096] To address irregular signal control scenarios and timing requirements such as phase overlap and phased secondary release, this invention proposes a phase residual flow ratio y′. p The concept is to address the adaptation problem of Webster timing method to the signal timing requirements of phase overlap and secondary release in the aforementioned stages.

[0097] During the phase flow ratio allocation process, the remaining flow of a phase is equal to the phase flow ratio minus the flow ratio already allocated to the phase.

[0098] During initialization, the ratio of residual flow rate for each phase is equal to the phase flow rate ratio, i.e., y′=y.

[0099] For phase overlap: introduce a phase-independent phase set I: the set of phases included in the current phase and not included in the next phase, that is, the phase-independent phase set I is the complement of the phase intersection of the current phase and the next phase in the current phase phase set.

[0100] The traffic ratio is allocated by traversing the stages in the plan. The allocation method is as follows:

[0101] Y = max{y′ i y′ j ,…},i,j∈I

[0102] Among them, y i ′ is the residual flow ratio of phase i.

[0103] After calculating the flow ratio for the current stage, the remaining flow ratio for each phase is updated. The remaining flow ratio for each phase satisfies the formula:

[0104] y′ i =max{0, y′ i -Y k δ ik}

[0105] Where, δ ik It is a 0-1 variable; when phase i is included in stage k, δ ik Equals 1; otherwise δ ik It equals 0.

[0106] For phased secondary release: the phase flow ratio can be reversibly allocated and averaged with the forward allocation result.

[0107] The allocation method based on the remaining flow ratio can effectively address common phase overlap and secondary release scheme designs, reduce the limitations of phase scheme design on the application scope of theoretical models, and improve the applicability of theoretical models.

[0108] After the phase flow ratio allocation is completed, if the phase remaining flow ratio is 0, the signal timing phase is entered; if there is a phase remaining flow ratio that is not zero, the current phase scheme is not applicable to the current traffic conditions.

[0109] 4. Signal timing optimization

[0110] In the signal timing autonomous optimization process, if the signal timing satisfies the minimum green constraint, the signal timing ends; if the signal timing does not satisfy the minimum green constraint, the minimum green adjustment process begins.

[0111] (1) Flow ratio reset

[0112] Webster's timing method requires the sum of stage flow ratios to be no greater than 0.9 and is only applicable to signal timing schemes for unsaturated traffic. This invention proposes an improved calculation method that makes it applicable to signal timing schemes for oversaturated traffic.

[0113] When ∑Y>0.9, the flow ratio Y of all stages is redistributed according to their respective proportions of the overall flow Y / ∑Y, as follows:

[0114]

[0115] When ∑Y≤0.9, Y′=Y.

[0116] The signal timing-flow ratio is reset to address the issue that the original theoretical model is not applicable under saturated and oversaturated traffic flow conditions.

[0117] (2) Calculate the total signal loss time

[0118] The total signal loss time L is equal to the sum of the stage loss times l. s The phase loss time is equal to the sum of the phase red light times within the phase. p Phase start-up loss l p The maximum value, i.e.

[0119] L=∑l s =∑max(rt p +l p )

[0120] rt p The composition is relatively complex and is related to the red light time during phase transition and the late start and early end time configuration of the phase within the phase (from GB / T20999-2017).

[0121] (3) Calculate the initial period

[0122]

[0123] (4) Calculation of effective green light time

[0124]

[0125] (5) Minimum Green Constraint

[0126] Minimum green is a configuration parameter for the phase, and the acquired g s The effective green light time for a phase needs to be converted into a check of the phase's minimum running time. The phase's minimum running time must meet the minimum green light requirements for all phases, i.e., it needs to satisfy the following constraints:

[0127]

[0128] in, It is the sum of the non-green light time of the phase, which is related to the phase transition and the late start and early stop configuration of the phase within the phase (from GB / T20999-2017).

[0129] If a stage does not meet the minimum green constraint, the minimum green adjustment process is initiated; otherwise, the signal timing ends.

[0130] (6) Minimum Green Adjustment

[0131] The minimum green adjustment algorithm follows three principles: the maximum cycle principle, the flow ratio principle, and the remaining allocation principle.

[0132] A. Maximum Period Principle: Signal Period C w There is an upper limit threshold C max During the adjustment process, if the calculated signal period C w Reaching or exceeding C max At that time, C w =C max Then, after allocating the effective green light time for each phase, the minimum green light adjustment ends; but C max It should be greater than the sum of the minimum running times of all stages, i.e.

[0133]

[0134] B. Flow ratio principle: In C w Not exceeding C max In the case of a phase, the green credit ratio is greater than or equal to the flow ratio;

[0135] C. Residual Proportion Principle: In C w Not exceeding C maxIn the case of the minimum green adjustment phase, the remaining phase time is allocated according to the flow ratio for the other phases.

[0136] The signal timing process provides the minimum green check adjustment principle and clear method, avoiding imbalance between the green ratio and flow ratio and excessive signal period during the calibration phase.

[0137] Example

[0138] 1. Vehicle passage information collection and storage

[0139] Loop detectors or video / radar-based virtual loop detectors are deployed in each lane before the stop line at intersections to collect vehicle passing information. The detectors are positioned within 0-30 meters before the stop line to ensure vehicles enter the solid lane line area, reducing the possibility of lane changes. Vehicle passing information includes lane ID, vehicle VID, vehicle type, passing status [vehicle entering / exiting the detection loop], and passing time (in milliseconds).

[0140] The cycle begins with the start of each signal release and ends when a signal release is completed. The vehicle data recorded during this period is used as the raw data for the current cycle.

[0141] Traffic flow is calculated by lane, and the status is determined by the number of small, medium, and large vehicles entering the lane. These numbers are then multiplied by an equivalent conversion factor and summed to obtain the periodic traffic flow q. lia .

[0142] Vehicle type small cars mid-size car large vehicles Equivalent conversion factor 1 2 3

[0143] 2. Saturation flow calibration based on vehicle passage characteristic analysis

[0144] (1) Data filtering

[0145] Assuming the vehicle stationary speed threshold is 1 m / s, the average vehicle length is 5 m, and the loop width is 1 m, then when the passing time of the i-th vehicle and the (i+1)-th vehicle satisfies t i，1 -t i，0 >6 or t i+1，0 -t i，1 When the relationship is greater than 6, the i-th vehicle is an intermittently affecting vehicle. Delete the records of intermittently affecting vehicles and use these intermittently affecting vehicles as breakpoints to split the periodic vehicle passages in each lane into multiple sets of data, thus completing data cleaning.

[0146] (2) Calculate the headway

[0147] For each set of statistical data, starting with the record containing the entry value "status", calculate the time t for each vehicle to pass through the loop. vid And the headway h between it and the next car in the workshop vidThe method is as follows:

[0148] t vid =t vid，1 -t vid，0

[0149] h vid =t vid+1，0 -t vid，0

[0150] (3) Saturation flow rate calibration

[0151] Based on the assumptions of an average vehicle length of 6m and a 2m interval between adjacent vehicles when parked on the road, the lane saturation flow rate is calibrated as follows:

[0152]

[0153] Where n is satisfied The number of hvid. When the number of n is less than the minimum sample size, the lane saturation flow rate is taken as the design value.

[0154] 3. Calculation of stage flow ratio based on remaining flow ratio

[0155] (1) Phase flow ratio

[0156] Based on intersection channelization and traffic light settings, the controllable signal phases and phase-lane correlations are clearly defined to generate phase-flow ratios. For example... Figure 5 The detection layout diagram shown illustrates the relationship between controllable phase and lane / detector:

[0157]

[0158]

[0159] The phase flow ratio y is equal to the ratio of the phase flow rate to the phase saturation flow rate, i.e.

[0160]

[0161] Where m is the number of phase-associated lanes, and α is the saturation index of each phase assigned according to the control objective.

[0162] (2) Phase flow ratio allocation

[0163] Furthermore, the flow ratio allocation is completed in conjunction with the signal scheme design.

[0164] like Figure 6The diagram shows a comparison between a standard phase scheme and a phase scheme with overlap. In the overlap phase scheme, some phases (phases 5 and 6) are released twice. The traditional method of using the maximum value of the phase-related phase flow ratio as the phase flow ratio cannot solve this type of scenario or requires other preset conditions. This invention implements an allocation method for the secondary release scenario based on the introduced residual flow ratio.

[0165] During the phase flow ratio allocation process, the remaining flow of a phase is equal to the phase flow ratio minus the flow ratio already allocated to the phase. Initially, the remaining flow ratio of each phase is equal to the phase flow ratio, i.e., y′ = y.

[0166] For phase overlap: Introduce a phase-independent phase set I, which is the set of phases included in the current phase but not included in the next phase; that is, the complement of the intersection of the phases of the current phase and the next phase in the current phase's phase set. Flow ratio allocation is performed by traversing the phases in the scheme, using the following method:

[0167] Y = max{y′ i y′ j ,…},i,j∈I

[0168] Among them, y i ′ represents the remaining flow ratio for phase i. After calculating the flow ratio for the current phase, the remaining flow ratio for each phase is updated.

[0169] y′ i =max{0, y′ i -Y k δ ik}

[0170] Where, δ ik It is a 0-1 variable; when phase i is included in stage k, δ ik Equals 1; otherwise δ ik It equals 0.

[0171] If, after the phase flow ratio allocation is completed, there is a phase with a remaining flow ratio that is not zero, then the current phase scheme is not applicable to the current traffic conditions; otherwise, the signal timing phase can proceed.

[0172] 4. Signal timing optimization

[0173] (1) Flow ratio reset

[0174] When ∑Y>0.9, the flow ratio Y of all stages is redistributed according to their respective proportions of the overall flow Y / ∑Y, as follows:

[0175]

[0176] When ∑Y≤0.9, Y′=Y.

[0177] (2) Calculate the total signal loss time

[0178] The total signal loss time L is equal to the stage loss time l s The sum of the phase loss time equals the sum of the phase red light times within the phase. p Phase start-up loss l p The maximum value, i.e.

[0179] L=∑l s =∑max(rt p +l p )

[0180] rt p The composition is relatively complex and is related to the red light time during phase transition and the late start and early end time configuration of the phase within the phase (from GB / T20999-2017).

[0181] (3) Calculate the initial period

[0182]

[0183] (4) Calculation of effective green light time

[0184]

[0185] (5) Minimum green check and adjustment

[0186] The following outlines the specific adjustment process for integrating the first round of minimum green checks:

[0187] A. Set an adjustment requirement identifier Adj[s] for each stage, and set a periodic adjustment identifier CAdj. Both identifiers are 0-1 variables and are initialized to 0.

[0188] B. Perform minimum green checks sequentially on the stages where Adj[s] = 0, and if the condition is met... Continue to the next stage of inspection; if not satisfied. The comparison ends when Adj[s] = 1 and CAdj ​​= 1.

[0189] C. If CAdj ​​= 0, proceed to process H; otherwise, if CAdj ​​= 0, continue.

[0190] D. For the phase where MGAdj[s] = 1, the effective green light time for that phase. Since the sum of the effective green light times in each phase equals the difference between the cycle time and the total lost time, the following relationship holds:

[0191]

[0192] E. Determine Cw+1 Does it exceed the period threshold C? max If the time limit is exceeded, proceed to process G; otherwise, calculate the current period C. w+1 The effective green light time for each stage is as follows:

[0193] (a) Calculate the effective green light time for all stages where Adj[s] = 0 based on the flow ratio:

[0194] g s =C w+1 Y′ s

[0195] (b) The effective green light time for the Adj[s]=1 stage is the minimum effective green light time.

[0196] F. Check the effective green light time (equal to the minimum effective green light time) for all stages where Adj[s] = 1, and determine g. s Is it less than C? w+1 Y s If the judgment result is true, set Adj[s] = 0 and CAdj ​​= 1. Return to process B.

[0197] G. Set the effective green time for all stages where Adj[s] = 1 equal to the minimum effective green time, i.e. For all stages where Adj[s] = 0, the effective green light time is allocated according to the remaining allocation principle:

[0198]

[0199] Return to process B.

[0200] H. Output the effective green light time for each stage, and according to C = ∑g s +L calculation cycle.

[0201] This invention achieves traffic signal timing optimization or real-time autonomous optimization by calibrating saturated flow rate and improving the Webster timing method to adapt to non-standard phase schemes and oversaturated traffic flow conditions.

[0202] The following are application scenarios for traffic signal timing optimization or real-time autonomous optimization based on this invention:

[0203] When applied to long-term stable traffic sections, such as urban arterial roads and highways, this invention can be used to optimize traffic signal timing. Calculations and adjustments are made based on historical data and traffic forecasts to improve traffic efficiency and reduce congestion. Traffic signal timing optimization methods typically involve adjustments within fixed time periods, such as morning and evening rush hours.

[0204] When applied to road sections with significant traffic flow fluctuations and frequent traffic condition variations, such as intersections and junctions, this invention can enable real-time autonomous signal optimization. By acquiring real-time vehicle information from traffic monitoring equipment, vehicle detectors, cameras, etc., the invention dynamically adjusts the signal timing scheme based on this information.

[0205] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A traffic signal optimization method based on parameter calibration and improved Webster timing method, characterized in that, The method comprises: S1, laying detection coils in each lane in front of the intersection stop line for collecting vehicle passing information and storing locally; S2, calibrating saturated vehicle headway based on the vehicle passing information and calculating saturated flow of each lane to reduce the influence of oversaturated traffic flow on signal timing; S3, calculating phase flow ratio based on phase residual flow ratio: for phase overlap and phase secondary release, the concept of phase residual flow ratio is proposed to solve the adaptation problem of Webster timing method under the signal timing requirements of phase overlap and phase secondary release, wherein: For phase overlap: introduce stage-independent phase set I, which is the complement of the intersection of the current stage and the next stage in the current stage phase set, traverse the stages in the phase scheme to allocate flow ratio, and the allocation method of the stage flow ratio is: Y = max {y', y",...}, i, j e I i j ,...}, i, j e I​ where y i is the remaining flow ratio for the current phase of phase i; After the current stage flow ratio is allocated, update each phase residual flow ratio, and each phase residual flow ratio satisfies the formula: y′ i = max{0, y′ i -Y k δ ik} where δ ik equals 1 when phase i is included in stage k, otherwise δ ik equals 0; For phase secondary release: allocate the stage flow ratio in reverse and take the average with the forward flow ratio allocation result; After the stage flow ratio allocation is completed, if all the phase residual flow ratios are zero, proceed to the signal timing link; if there is a non-zero phase residual flow ratio, the current phase scheme is not applicable to the current traffic condition; S4, in the signal timing link, if the signal timing satisfies the minimum green constraint relationship, the signal timing ends; if the signal timing does not satisfy the minimum green constraint relationship, proceed to the minimum green adjustment process; wherein: The minimum green constraint relationship is: where g s is the phase effective green time, l s is the phase lost time, is the sum of the non-green times of the phases, The algorithm of the minimum green adjustment satisfies: where C max is a signal period upper threshold value, set C w is a signal period, in the minimum green adjustment process: If C w ≥ C max then end the minimum green adjustment after the phase effective green time has been distributed. If C w <C max Then, except for the phase of the minimum green adjustment, the remaining phases are allocated the remaining phase time in the ratio of the flows.

2. The traffic signal optimization method based on parameter calibration and improvement of Webster timing method according to claim 1, characterized in that, The method further comprises: When the method is applied to a road section with large traffic flow changes, the updated signal timing is executed in the next cycle, and the vehicle passing information is recorded based on the updated signal timing, realizing real-time optimization of traffic signals; When the method is applied to a road section with long-term stable traffic, update the vehicle passing information within a fixed time period based on historical data to optimize traffic signal timing.

3. The traffic signal optimization method based on parameter calibration and improvement of Webster timing method according to claim 1, characterized in that, The vehicle passing information in S1 includes lane lid, vehicle vid, vehicle type type, vehicle passing status status, and vehicle passing time time.

4. The traffic signal optimization method based on parameter calibration and improvement of Webster timing method according to claim 3, characterized in that, Data filtering is performed on the vehicle passing information, which includes deleting the vehicle passing information of intermittent influencing vehicles and splitting the vehicle passing of each lane cycle into multiple groups of data with the intermittent influencing vehicles as breakpoints.

5. The method for optimizing traffic signal based on parameter calibration and improving Webster timing method according to claim 1, characterized in that, The method for calibrating saturated flow in S2 is: where n is the number satisfying h vid is the number of vehicles, when the number of n is less than the minimum sample size, the saturation flow of the lane takes the design value, h vid is the adjacent headway, t vid is the time for each vehicle to pass through the detection coil, l is the set body length, w is the set width of the detection coil, h s is the average headway when parking.

6. The method for optimizing traffic signal based on parameter calibration and improving Webster timing method according to claim 1, wherein, For the Webster timing method, the sum of the stage flow ratio is required to be not greater than 0.9, and it is only applicable to non-saturated traffic signal timing, and the flow ratio reset is also involved in S4, and the method of flow ratio reset is: When ∑Y>0.9, re-allocate the flow ratio Y of all stages according to their proportion in the whole, and the allocation method is as follows: When ∑Y≤0.9, Y'=Y.

7. The traffic signal optimization method based on parameter calibration and improvement of Webster timing method according to claim 6, characterized in that, The S4 also involves the total signal loss time L, the initial signal period C0 and the phase effective green light time g s Calculation: The calculation formula of the total signal loss time L is: L = ∑i s = ∑max(rti p + l p ) where rt p is the sum of the phase internal phase red light times, l p is the phase start-up loss; The calculation formula of the initial signal cycle C0 is: Where L is the total signal loss time, Y′ s It is the flow ratio mentioned; The phase effective green time g s The calculation formula is: where Co is the initial signal period, L is the total signal loss time, Y' is the signal loss time, and Y is the signal loss time of the signal to be transmitted. s is the flow ratio.

Images