WO2023065057A1 - 交通信号智能控制架构方法 - Google Patents

交通信号智能控制架构方法 Download PDF

Info

Publication number
WO2023065057A1
WO2023065057A1 PCT/CN2021/000212 CN2021000212W WO2023065057A1 WO 2023065057 A1 WO2023065057 A1 WO 2023065057A1 CN 2021000212 W CN2021000212 W CN 2021000212W WO 2023065057 A1 WO2023065057 A1 WO 2023065057A1
Authority
WO
WIPO (PCT)
Prior art keywords
traffic
signal
flow direction
flow
dimensional
Prior art date
Application number
PCT/CN2021/000212
Other languages
English (en)
French (fr)
Inventor
孟卫平
Original Assignee
孟卫平
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 孟卫平 filed Critical 孟卫平
Priority to PCT/CN2021/000212 priority Critical patent/WO2023065057A1/zh
Publication of WO2023065057A1 publication Critical patent/WO2023065057A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/065—Traffic control systems for road vehicles by counting the vehicles in a section of the road or in a parking area, i.e. comparing incoming count with outgoing count

Definitions

  • the invention relates to the field of traffic control. Specifically, a traffic signal artificial intelligence control architecture approach.
  • the purpose of the invention is to solve the problem of optimizing the structure of the response of the signal mode to the traffic information.
  • the present invention proposes a solution to achieve the above objectives, and proposes a control framework using traffic information prediction-signal mode including artificial intelligence methods, that is, AI-SPS framework. details as follows:
  • a traffic signal intelligent control framework method characterized in that it comprises:
  • S1 obtain road network parameters; create a number of road network traffic signal modes, referred to as signal modes, to form a signal mode library, each signal mode provides the control service corresponding to the road network traffic characteristics of the mode, and optimizes the composition matching with the road network traffic characteristics Signal; obtain or configure the current signal mode and its parameters of the road network, referred to as the current signal and its parameters;
  • the road network parameters include the length of each road section in the road network and its traffic time; the traffic time includes the driving time or/or the starting time with the congested convoy; The length is divided by the set driving speed, or includes the braking time minus the set driving speed; the time for the congested fleet to start refers to the time it takes for the congested fleet to leave the original place from the first car in the team to the time when the tail car in the team leaves the original place, which is equal to the starting time of the congested fleet.
  • the congestion coefficient is equal to the ratio of the fleet length to the road section fleet, when it is equal to 1, it means serious congestion, or including the start coefficient of the fleet is calculated according to the experimental value range of 0.10 to 0.26, or 0.18, the unit: Seconds/meters; Note: In this application, “or/or with the expression,” and “or with the expression,” only refer to the “or/or with” and “or with” immediately after the “expression”, does not involve the subsequent comma The listed “expressions”, such as “expression 1, or/or with expression 2, expression 3", “expression 1” and “expression 3” are in a coexistent relationship, and “expression 2" is in an “or” relationship with other expressions , or, "or” exists in a relationship;
  • Said signal mode includes any signal that sets the green light propagation direction between phases at crossings, which is called green wave; this flow direction is called green wave flow direction as a characteristic parameter of signal mode green wave, and the mode is recorded as green wave-flow direction;
  • the flow direction of the green wave is determined by the phase time difference of the ratio signal configuration between the intersections, that is, the order of the phase difference.
  • the phase time difference between two adjacent intersections is the relative phase difference , which is equal to the traffic time set for the road section between the adjacent intersections;
  • the signal mode with no phase difference between the ratio signals running between the intersections is a standing wave, which does not propagate green light changes, that is, the ratio mode, the green wave degradation mode, providing equal flow in each flow direction
  • the control service function of traffic characteristics refers to the configuration of each phase duration of the intersection according to the set period and the set ratio;
  • Balance means that the green light signal has no flow direction, and the phase difference of each intersection is synchronized
  • the controlled traffic flow direction includes both the same part and the opposite part.
  • the road sections with the same flow direction use the driving time and the road sections with the opposite flow direction use the congested convoy start time;
  • the guidance refers to the guidance of the traffic in the two opposite directions.
  • the green wave flow direction or combined with the green wave function is a green wave optimization signal, and its modes are respectively recorded as green wave-flow direction-guidance, green wave-flow direction-unblocking, green wave-flow direction-mixing, standing wave-unidirectional -Equilibrium;
  • the signal mode parameter also includes a period, and the distribution duration of each phase is the phase timing;
  • the phase includes the direction phase representing the control direction of the intersection, or the shunt phase with the direction phase to control the left/right steering;
  • the traffic characteristics of the road network referred to as the traffic mode
  • the parameters include flow direction, or/or with the status, or/or with the flow, or/or with each phase of the intersection; Congestion, or/or mixed with it; congestion refers to the condition of the fleet when the length of the road segment minus the length of the fleet reaches or is less than the set value, or the congestion coefficient reaches or exceeds the set value, and the set value is referred to as the road segment congestion set value; mixed Refers to some road sections congested and some road sections not congested; said traffic mode parameter flow direction or the traffic mode combined with the traffic mode parameter status, recorded as traffic mode-flow direction-movement, or with traffic mode-flow direction-congestion, or with traffic mode-flow direction - Mixed, or with traffic patterns - homogeneous - movement;
  • the matching of the signal mode and the traffic characteristics of the road network refers to the matching with the corresponding parameters of the traffic mode, including the green wave flow direction optimization signal matching the green wave flow direction and the traffic mode flow direction, or matching and combining with the green wave function and traffic mode conditions
  • Green wave optimization signal including green wave-flow direction-guidance matching traffic mode-flow direction-movement, standing wave-uniform direction-equilibrium matching traffic mode-unidirectional-movement, or matching green wave-flow direction-blockage matching traffic mode- Flow direction-congestion, or traffic mode-flow direction-mixing with green wave-flow direction-mixing; or/or cycle optimization signal matching with signal mode period and traffic mode flow; or/or phase matching with signal mode and traffic mode
  • the command signal mode refers to the signal mode of the mandatory execution command that does not need to be matched with the traffic information; 2) or determine the optimization signal from the preliminary judgment according to the priority rule; 3) make a corresponding mode transition period according to the determined optimization signal and the current signal;
  • the traffic signal intelligent control framework method of the present invention is characterized in that comprising,
  • the signal patterns in the signal pattern library include superimposed signals
  • the road network parameters include road network topology: topological parallelogram, or include virtual intersections and virtual road sections;
  • the green wave parameter flow direction also includes two intersecting flow directions, and the green wave is called a two-dimensional green wave, or recorded as a green wave-double flow direction or a two-dimensional green wave-flow direction, and its dimension reduction form includes a one-dimensional green wave- Flow direction, and 0-dimensional standing wave-unidirectional;
  • the green wave optimization signal includes two-dimensional green wave-flow direction-guiding, one-dimensional green wave-flow direction-guiding, 0-dimensional standing wave-unidirectional-balance, two-dimensional green wave Wave-flow direction-unblocking, one-dimensional green wave-flow-direction-unblocking, two-dimensional green wave-flow-mixing, one-dimensional green wave-flow-mixing, or with two-dimensional green wave-flow-direction, one-dimensional green Wave-flow-direction:
  • the flow direction of the traffic pattern parameters also includes two crossing main directions, and this traffic mode is called two-dimensional traffic, or recorded as traffic-two-flow direction or two-dimensional traffic-flow direction, and its dimensionality reduction form includes one main direction as one-dimensional Traffic-flow direction, and 0-dimensional traffic-unidirectional;
  • the traffic mode includes two-dimensional traffic-flow-movement, one-dimensional traffic-flow-movement, 0-dimensional traffic-unidirectional-movement, two-dimensional traffic-flow-congestion , one-dimensional traffic-flow-congestion, two-dimensional traffic-flow-mix, one-dimensional traffic-flow-mix, or with two-dimensional traffic-flow-convection, one-dimensional traffic-flow-convection;
  • the green wave optimization signal also includes an optimization signal matching the two-dimensional green wave with the two-dimensional traffic, that is, the two-dimensional green wave optimization signal, which may optionally include two-dimensional green wave-flow direction-guidance and two-dimensional traffic- Flow direction-movement matching, one-dimensional green wave-flow direction-guiding and one-dimensional traffic-flow direction-movement matching, 0-dimensional standing wave-unidirectional-equilibrium and 0-dimensional traffic-unidirectional-movement matching, or two-dimensional green wave- Flow direction-flow-congestion matching with two-dimensional traffic-flow-congestion, or matching with one-dimensional green wave-flow-flow-blocking and one-dimensional traffic-flow-congestion, or matching with two-dimensional green wave-flow-direction-mixing and two-dimensional traffic- Flow direction-mixing and matching, or matching with one-dimensional green wave-flow-direction-mixing and one-dimensional traffic-flow-direction-mixing, or matching with two-dimensional green wave-flow-direction-mixing, or matching with two-
  • the traffic signal intelligent control framework method of the present invention is characterized in that comprising,
  • the superimposed state signal also includes, under the set traffic characteristic condition, let the set vehicle fleet of one phase of the intersection obtain several continuous intersection green light signals, referred to as solitary waves, which are superimposed in the two-dimensional green wave;
  • the signal pattern parameter also includes the path temporary timing table of the several consecutive intersections, which is called the solitary wave temporary timing table;
  • the set traffic characteristic conditions include instructions to allow the set motorcades at the designated crossing to flow to the phase to forcibly occupy the green light time of other phases at the crossing ahead, to pass through, referred to as forced solitary waves; Occupying the remaining time of the green lights of other phase convoys at the intersection ahead, referred to as Yushi, passing, referred to as Yushi solitary wave; both are collectively referred to as solitary wave traffic;
  • the optimization signal of the signal mode matching traffic mode also includes a solitary wave-flow direction-solitary wave temporary timing table and a set fleet-flow direction-path intersection table, and the solitary wave temporary timing table includes a forced solitary wave temporary timing table or/ Or a temporary timing table with the remaining time solitary wave;
  • the predicted traffic information also includes, the real-time measured and predicted set traffic characteristic conditions include instruction solitary waves: including mandatory solitary waves, or/or solitary waves with residual time; configuration of temporary timing tables for mandatory solitary waves Calculation: Compulsory designation of the flow direction phase of the intersection, including the direction phase or diversion phase, setting the green light time of the fleet occupying the phase of other phase fleets at the intersection ahead, and calculating and establishing the temporary timing table of the mandatory solitary wave; the temporary timing table of the solitary wave in the remaining time Calculation of the configuration: designate intersection flow direction phase including direction phase or shunt phase to set the fleet to occupy the relevant phase of the intersection ahead and the remaining green light time to pass after the fleet provides this phase, and calculate and establish the temporary timing table of the remaining time solitary wave according to this; calculate all
  • the method for the remaining time is referred to as the A-A algorithm, including the use of the flow direction traffic phase at each set intersection, including the special information of the directional phase diversion phase tail
  • the decision to optimize the signal 1) predicting the matching of the traffic mode and the signal mode also includes receiving the command solitary wave, determining the phase of the intersection to set the fleet, and superimposing the solitary wave on the current signal mode for preliminary judgment of the optimized signal, including using
  • the solitary wave prediction algorithm establishes and obtains the temporary solitary wave timing table of several consecutive intersection green light signals;
  • the 2) or the priority rule also includes the mandatory solitary wave preliminary judgment prior to the remaining time solitary wave preliminary judgment, and the solitary wave preliminary judgment Judgment takes precedence over other preliminary judgments;
  • the execution optimization signal further includes: executing a temporary solitary wave timing table.
  • the traffic signal intelligent control framework method of the present invention is characterized in that comprising,
  • the superimposed state signal also includes, under the condition of setting traffic information characteristics, let the vehicle with the closest phase to the intersection, that is, the first vehicle in the phase, obtain the green light signal from the red light, which is called differential green wave, and is superimposed in the two-dimensional green wave ;
  • Said driving refers to a vehicle running normally at a prescribed speed;
  • the signal mode parameters also include the minimum safe switching time between the green light and the red light between the phases, which is called the differential green wave time, or the phase variable time ⁇ t Th0 , and the two phases involved;
  • the minimum safe switching time of the phase is The time means that the first vehicle in the phase can safely stop at the front stop line with normal braking when its control phase signal is switched from green to red, and at the same time allow the other phase first vehicles to pass safely after the control phase signal is switched from red to green
  • the minimum time at the front intersection; the distance between the first vehicle in the phase and the front control phase calculated according to the minimum safe switching time of the phase is the minimum safety distance of the phase;
  • the set traffic information characteristic conditions include that the first vehicle q0 of the phase when each set intersection flows to the traffic phase red light is at or below the minimum safety distance of the phase set for the phase, and the first vehicle q0 of the phase of the ratio signal green light of the intersection is at its other There are no pedestrians outside the minimum safe distance of the phase or with the corresponding pedestrian phase, that is, there is no car or no pedestrian in the green light phase of the ratio signal, which is collectively called the differential traffic in which the first vehicle is in a differentiable position in the red light traffic phase, also called differential green wave condition; or with the continuation of the differential green wave condition under the following setting sequence rules, when the green light phase of "occupied with the ratio signal green light time" continues to detect a vehicle, or when a vehicle is detected in other shunt phases with the same flow direction, Or other flow direction when a vehicle is detected;
  • the optimized signal for matching the signal pattern with the traffic pattern also includes differential green wave time and its two phases and differential green wave condition and its two phase matching;
  • the predicted traffic information further includes the real-time measured differential green wave condition of the set traffic characteristic condition
  • said 1) matching of predicted traffic mode and signal mode in said decision optimization signal also includes, according to the measured differential green wave condition or receiving instructions to differentiate the green wave signal mode, in order to satisfy the differential green wave signal mode under the current signal mode
  • the first vehicle at the red light phase under the wave condition is matched with the differential time optimized signal superimposed on the current signal mode for preliminary judgment, that is, the first vehicle at the red light phase that satisfies the above "differential green wave condition" under the current signal mode obtains an occupancy "ratio signal".
  • One of the phase variables of the green light phase without cars or "with no pedestrians" phase takes time or is called the differential time; the 2) or priority rule also includes differential green wave preliminary judgments taking precedence over other preliminary judgments;
  • executing the optimization signal also includes: immediately switching or keeping the green light to the phase for obtaining the differential time;
  • the traffic signal intelligent control framework method of the present invention is characterized in that comprising,
  • the signal mode library includes the operation of adding modes or mode parameters, setting a new signal mode to provide the service of the set traffic mode or serving the set traffic mode, and matching the signal mode parameters with the traffic mode it serves , and its initial judgment of optimized signal, its priority, and its implementation operation are respectively configured in the signal pattern library in S1, in the predicted traffic information in S2, in the priority rule for determining the preliminary judgment of optimized signal in S3, and in S4 to execute the operation of optimized signal middle;
  • its S2 prediction traffic information is characterized by including,
  • the measured or predicted road network traffic information and its characteristics include that by setting a number of attention points equipped with traffic sensing devices in the road network, the set positions in the road section or/or intersections are used to obtain all Traffic flow characteristic data or/or vehicle fleet characteristic data of the set flow direction of the above-mentioned attention points, so as to analyze and express the characteristics of the road network traffic information;
  • its S2 prediction traffic information is characterized by including,
  • the traffic flow characteristic data of the set flow direction of the focus point includes corresponding set values of traffic flow or its variation set values, or/or the set value of the difference between the flow of the flow direction relative to the set flow direction, reaching the set value
  • the flow direction of the above-mentioned flow setting value or the flow difference setting value is taken as the main direction of the focus point; or the fleet characteristic data related to the focus point setting flow direction includes a number of corresponding fleet length setting values or its variation setting values ; or/or the set value of the difference between the set flow to the length of the fleet and the length of the road section where it is located, the situation that reaches the set value of the difference is the set flow to the congested road section of the focus point;
  • its S2 acquisition of predicted traffic information is characterized by including,
  • the point of interest is configured with a calculation and processing unit for traffic information prediction at this point, referred to as a point of interest unit, and each point of interest unit uses the traffic information of the point of interest to set the flow direction corresponding to the setting prediction method to make predictions; several of the points of interest
  • the focus unit forms the forecasting layer;
  • the forecasting method refers to any method for predicting future data based on existing data, including the repetitive forecasting method, empirical method, mean method, and maximum value method that directly use real-time detection data as predicting future data.
  • Statistical optimization methods, artificial intelligence methods including neural network methods and expert system methods.
  • its S2 acquisition of predicted traffic information is characterized by including,
  • intersection of the attention point is used to set the flow direction of the traffic information prediction calculation processing unit, referred to as the intersection unit, each intersection unit is also responsible for predicting the traffic information characteristics of the flow direction phase diversion phase, or involves including the flow direction upstream adjacent road section The characteristics of the traffic information of the vehicle source or upstream intersection unit and its setting value.
  • its S2 prediction traffic information is characterized by including
  • the number, distribution, and types of road sections or intersections of the attention points are determined by the distribution of traffic flows predicted by the setting prediction method.
  • its S3 decision to optimize the signal is characterized by including,
  • the preliminary judgment of the green wave optimization signal includes green wave flow direction optimization and green wave function optimization, and the steps include 1) determining the flow direction of the traffic mode: calculating the main direction flow of the attention point of each flow direction to reach the flow set value or/or the flow rate
  • the difference setting value is the sum of the traffic flow in the main direction of the attention point.
  • the flow direction with a large sum value is set as the main traffic direction of the road network, that is, the flow direction of the traffic mode.
  • the flow direction is recorded as two-dimensional traffic-flow direction
  • the main traffic direction of a road network is the single flow direction of traffic mode, which is recorded as one-dimensional traffic-flow direction
  • the traffic mode flow direction without the main traffic direction of the road network is recorded as zero-dimensional traffic-average flow direction.
  • the traffic congestion road section number, sum calculates the number of road sections where the length of the caravan of the same flow direction reaches the set value of the fleet length or/or the difference between the length of the road section and the length of the road section reaches or is less than the set value of the road section congestion, referred to as the traffic congestion road section number, sum, and the flow direction with the largest sum value is set as the main direction of road network traffic congestion, that is, the main direction of traffic mode congestion.
  • the flow direction setting value is the flow direction congestion setting value
  • the two flow directions are the main directions of traffic congestion in the two road networks, which are recorded as traffic-dual flow-congestion or two-dimensional traffic-flow-congestion, when only one flow direction has If there are more congested road sections in the flow direction, and reach or exceed the set value of the flow direction congestion, while the other flow direction does not have the number of congested road sections, then the flow direction is the main direction of traffic congestion in the road network, which is recorded as one-dimensional traffic-flow-congestion, When there is no flow direction with the number of congested road sections reaching or exceeding the set value of the flow direction congestion, the flow direction is the part of the congestion flow direction, including small area congestion flow or several road section congestion flow directions; the combination characteristics with the traffic mode flow direction include two-dimensional traffic- Flow-movement, or/or with zero-dimensional traffic-unidirectional-movement, or/or with one-dimensional traffic-flow-movement, or/or with two-dimensional traffic-flow-jamming, or/
  • the data of its predicted target includes designated traffic characteristic data, adjust and optimize the signal parameters of the traffic with the described prediction method, according to the existing designated traffic Feature data and green wave parameter data group to optimize the signal parameters of the traffic feature data that meet the prediction target, and the preliminary judgment of the optimized signal is called the direct optimization signal prediction method;
  • its S3 decision to optimize the signal is characterized by including,
  • Preliminary judgment of cycle optimization signal, or preliminary judgment of phase timing optimization also includes forecast optimization using any intersection signal timing algorithm, Webster timing method, conflict point method, estimation method, critical lane method, and cycle phase
  • the prediction method related to timing optimization signal is called cycle phase timing optimization signal prediction method
  • its S3 decision to optimize the signal is characterized by including,
  • the prediction method also includes a direct optimization signal prediction method and a cycle phase timing optimization signal prediction method, through the optimization signal learning method with the goal of minimizing the vehicle waiting time or/or minimizing the waiting length of the fleet, the road is found According to the new change relationship between the traffic information characteristic data obtained by the traffic information focus point of the network and the corresponding optimized signal parameter data, the new changed relationship is created as a new optimized signal and added to the signal pattern library;
  • its S3 decision to optimize the signal is characterized by including
  • the priority rule determines the priority order of the preliminary judgment of each optimization signal according to the data sorting including calculation prediction/estimation of the reduction of driving waiting time/reduction of vehicle fleet/increase in flow rate that will be caused by the preliminary judgment of each optimization signal;
  • its S31 decision to optimize the signal is characterized by including
  • the priority rules are sorted according to the size of the estimated data, and the former is prioritized over the latter: including the command signal mode, or a preliminary judgment with two-dimensional green wave-flow direction-blockage removal, or two-dimensional green wave-flow direction -Mixed preliminary judgment, or with one-dimensional green wave-flow direction-blocking preliminary judgment, or with one-dimensional green wave-flow direction-mixed preliminary judgment, or with two-dimensional green wave-flow direction-reference preliminary judgment, or with one-dimensional Green wave-flow direction-direction preliminary judgment, two-dimensional green wave-flow direction-guidance preliminary judgment, one-dimensional green wave-flow direction-guidance preliminary judgment, standing wave-uniform direction-equilibrium preliminary judgment;
  • the traffic signal intelligent control framework method of the present invention is characterized in that the forced solitary wave temporary timing includes,
  • F201 calculates the time required to pass through the local intersection, referred to as the forced solitary wave time; subtract the forced solitary wave time from the local phase green light time, and get a number less than 0, then use other phases accordingly Amount of time borrowed to use;
  • the remaining green light time is obtained, if it is less than 0, the absolute value is the time to predict the corresponding amount of borrowing from other phases; the set time is the time it takes to assume that the phase convoy passes;
  • Fig. 1 is a flow chart of a traffic signal control framework method
  • Fig. 2 is a road section and traffic data-flow diagram of 2 set focus points
  • Figure 3 is a structural diagram of the "prediction and decision-making" structure of the road network traffic signal control framework for two road sections with set points of interest;
  • Fig. 4 is the crossing and traffic characteristics-flow or fleet length diagram of community setting attention point
  • Figure 5 is a structural diagram of the "prediction and decision-making" structure of the road network traffic signal control architecture at the intersection of the community's set focus point;
  • Figure 2 1--The starting point (0, 0) of network intersection node code identification is the lower left intersection of the road network, and the mark ⁇ (0, 0), (6, 4) ⁇ represents the origin is (0, 0), vertical and horizontal
  • the maximum value (6, 4) of coordinates (i, j) is 6, 4 respectively.
  • Each coordinate (i, j) is the intersection of two roads, and there are 7 roads and 5 roads intersecting with it.
  • the azimuth is consistent with the traffic flow, and the unit is vehicle-number of vehicles; 3--intersection distance-driving time/congested fleet start time is recorded as #-#/#: unit : m-s/s;
  • Note 1 The intersection (0, 4) in the upper left corner of the figure is a virtual intersection, represented by a small dotted line, and the connected road section (0, 4) is a virtual road section, represented by a dotted line;
  • Note 2 The triangle mark at the intersection (0, 0) in the lower left corner of the road network in the figure indicates that the intersection is the base point intersection of the mode phase difference of the current superposition signal, and the two hollow solid arrows point to the current direction of the green wave flow of the signal, east and north;
  • Note 3 Figure The dotted triangle mark at the intersection (6, 4) in the upper right corner of the road network indicates that the intersection will be configured as the intersection of the mode phase difference base point of the optimized superposition signal, and the two hollow dotted arrows point to the green wave flow
  • Figure 3 1-Signal pattern library; 2-Set the attention point road section unit and the traffic information data receiving and prediction module, 2 attention point road section units form the prediction layer, 3-Hollow arrows represent layers and layers, modules and modules Data connection, the arrow points to the data connection between the prediction layer and the analysis layer, 4-the single-flow direction optimization signal preliminary judgment unit in the analysis layer or the function optimization signal, the arrow points to this is the north flow direction optimization signal preliminary judgment unit, 5-analysis layer The initial judgment unit of the middle and double flow direction optimization signal or the function optimization signal. The arrow points to this is the initial judgment unit of the east flow direction and the north flow direction optimization signal.
  • Figure 4 set attention point intersection in one of the 2-12 set attention point intersection communities (3, 1 ), indicated by the arrow, represented by an octagon, the attention point is equipped with a traffic flow phase traffic information flow collection device at the intersection, or a diversion phase traffic information flow collection device, or a fleet head and tail information collection device, surrounding the intersection
  • the "East”, “West”, “South” and “North” orientation labels and the 8 numbers separated by double angle brackets "#-#/#/#”#-#/#/#” denote The azimuth traffic #-hezhi#/, left#/, right# three phases are measured and wait to pass the crossing.
  • the measured traffic flow on the north side of the intersection waiting for southbound traffic is 1, the fleet going straight is 1, the left turn is 0, the right turn is 0, the predicted traffic is 8, and the queue length is 7/0/0.
  • This orientation is opposite to the traffic flow direction, and the unit is car
  • Note 1 the westward black arrows on the horizontal road section (2, 4) and the horizontal road section (2, 5) represent the westbound congested convoy and its captain;
  • Note 2 The dotted inverted triangle in the intersection (3, 2) represents that the intersection is the base point of the mode phase difference of the green wave that flows eastward;
  • Figure 5 2- Set the traffic information data receiving and prediction module of the intersection unit of the attention point, 12 intersection unit prediction modules form the prediction layer, and the arrow points to the intersection unit prediction module of the intersection (5, 3) where the attention point is set; 8 -Preliminary judgment of solitary wave optimization signal in the analysis layer, 9-solitary wave management module in the decision-making layer; 10-preliminary judgment of differential green wave optimization signal in the analysis layer, 11-solitary wave conflict resolution module in the decision-making layer;
  • Embodiment 1 referring to Fig. 1 and Fig. 2,
  • E1-S11 obtain the road network parameters, including the length of each road section in the road network and its traffic time, as shown in the mark 3 in Figure 2;
  • the traffic time includes the time for starting the congested fleet of the road section; wherein the driving time refers to the vehicle to set The time taken to complete the road section at the speed of the vehicle is equal to the length of the road section divided by the set driving speed, or including the braking time minus the set driving speed;
  • the topology of the road network is composed of approximate quasi-parallelograms including virtual intersections and virtual road sections;
  • Several superimposed signal modes of road network traffic shown in Figure 2 form a signal mode library, each of which provides the control service of the traffic mode corresponding to the mode, and forms an optimized signal that matches the traffic mode; obtain or configure the current operation
  • the signal mode and its parameters are called the current signal and its parameters, including the signal period of 68 seconds, and the phase timing of each intersection is: the phase timing ratio of the north-south flow direction and the east-west flow direction is
  • the current signal is marked in Figure 2, the two-dimensional green wave-east and north-guided, the base point intersection of the mode phase difference is at the intersection (0,0) in the lower left corner of the road network, and the two cross flows are east and north, and the mode cycle of each intersection
  • Table 1 The remainder and periodic complement are shown in Table 1:
  • Said signal mode includes any signal that sets the green light propagation direction between phases at crossings, which is called green wave; this flow direction is called green wave flow direction as a characteristic parameter of signal mode green wave, and the mode is recorded as green wave-flow direction;
  • the green wave flow direction is calculated according to the phase time of the most upstream intersection, which is called the mode phase difference base point or the green wave source point, and configures the phase difference of the relevant phases of other intersections.
  • the phase difference is called the mode absolute phase difference;
  • the phase time difference is the relative phase difference, which is equal to the traffic time set for the road section between the adjacent intersections;
  • the signal mode with no phase difference between the ratio signals running between the intersections is a standing wave, which does not propagate the change of the green light, that is, the ratio mode, green wave Degradation mode, providing the control service function of equal traffic characteristics in each flow direction;
  • the ratio signal refers to the configuration of each phase duration according to the set period and set ratio at the intersection;
  • the green wave characteristic parameters also include green wave functions, including guidance or equalization or/ Or it can be opposite guidance with decongestion or/or mixed or/or with convection guidance, wherein guidance means that when the green wave flow direction is the same as the controlled traffic flow direction, the relative phase difference is calculated by using the set driving time for the traffic time of each road section; balanced means The green light signal has no flow direction, and the phase difference of each intersection is synchronized; clearing congestion means that the flow direction of the green wave is opposite to the
  • the road section of the same flow direction uses the driving time and the road section of the opposite flow part uses the congested convoy start time;
  • the guidance refers to the guidance of the traffic in the two opposite directions, which is a kind of guidance.
  • the green wave signals are respectively recorded as green wave-flow direction-guiding, green wave-flow direction-unblocking, green wave-flow direction-mixing, standing wave-uniform direction-equilibrium; the parameters of this signal mode also include period, each phase distribution
  • the duration is the phase timing; the phase includes the directional phase representing the control direction of the intersection, or the shunt phase that controls the left/right steering in the directional phase; this embodiment does not include the blocking function because there is no fleet length sensing system configured;
  • the green wave signal includes two-dimensional green wave-flow direction-guiding, one-dimensional green wave-flow direction-guiding, 0-dimensional standing wave-unidirectional- Balanced, or with two-dimensional green wave-flow direction-direction, one-dimensional green wave-flow direction-direction; two-dimensional green wave-flow direction in the present embodiment, wherein flow direction comprises east and north or north and east, east and south or South and east, west and north or north and west, west and south or south and west, wherein the flow direction includes east, south, west, north, and zero-dimensional standing wave-unidirectional; the two-dimensional green wave flow direction or green Wave function combination, included in the present embodiment under the signal mode column of the signal mode library in Table 2, as shown in mode numbers 4 to 8, two-dimensional green wave-flow direction-guiding, or with two-dimensional green wave-flow direction-guiding , or with one-dimensional green wave-flow direction-direction, one-dimensional green wave-flow direction-guidance,
  • the position of a two-dimensional green wave base point intersection also determines the two Cross flow direction parameter; the mode phase difference base point crossing position of one-dimensional green wave is set at the crossing on the channel of the same cross non-green wave flow direction of the most upstream of each channel of said green wave, that is, one-dimensional green wave base point passage crossing; zero-dimensional station
  • the intersection of the base point of the wave mode phase difference can be set to any intersection;
  • the traffic mode parameters include the flow direction, or/or with the condition, or/or with the flow rate, or/or with the traffic flow of each phase of the intersection; wherein, the flow direction of the traffic mode includes two main directions intersecting, and the traffic mode is called two-dimensional Traffic, or denoted as traffic-double flow or two-dimensional traffic-flow, its dimension reduction form includes a main direction as one-dimensional traffic-flow, and 0-dimensional traffic-unidirectional, two-dimensional traffic-flow includes in the present embodiment East and North or North and East, East and South or South and East, West and North or North and West, West and South or South and West, whose degenerate form includes a main direction characterized by one-dimensional traffic-flow direction, including East , south, west, north, and 0-dimension-unidirectional; traffic conditions in this embodiment do not include congestion, mixing, and traffic mode conditions include motion and non-congestion; congestion refers to the length of the road section minus the length of the fleet reaching or less than the set value, or Congestion coefficient reaches or exceeds setting
  • the matching of the signal mode and the traffic mode refers to the matching of each corresponding parameter, including the green wave flow direction optimization signal matching the flow direction of the signal mode and the flow direction of the traffic mode; the green wave function matching the green wave function and the traffic mode condition
  • the green wave optimization signal of the optimized signal combination is included in the present embodiment under the signal mode column and the matching traffic mode column of the signal mode library in Table 2, as shown in the mode serial numbers 4 to 8, the two-dimensional green wave-flow direction-guidance and two-dimensional One-dimensional traffic-flow-movement matching, one-dimensional green wave-flow-guiding and one-dimensional traffic-flow-movement matching, zero-dimensional standing wave-uniform-equilibrium and zero-dimensional traffic-unidirectional-movement matching, or two-dimensional Green wave-flow direction-direction matching with two-dimensional traffic-flow direction-convection, or matching with one-dimensional green wave-flow direction-direction and one-dimensional traffic-flow direction-convection; or/or matching with the signal pattern cycle and traffic pattern A period optimization signal for
  • E1-S201-S205 select the focus of the road network: forecast according to the traffic distribution and past experience data Figure 2
  • the road network is prone to peak traffic for a long time in the four directions of vertical and horizontal traffic, and the two road sections are the horizontal section (3, 3) and the vertical section
  • the position in the road section (4, 3) is determined to set the two road sections as the midpoint of the road section of interest in the road network, and the two-way traffic flow sensor is configured at the set position, and the lane coil or vehicle positioning data is used to obtain the east, west, south, north 4 Traffic flow in each flow direction, to express the road network traffic information and its characteristics
  • S203 the traffic information prediction calculation processing unit configured for the position of the road section of the attention point, referred to as the road section unit, uses a repetitive prediction method to predict each time of the attention point traffic flow in each direction of traffic
  • S202 wherein the traffic flow direction characteristic determination refers to that the absolute value of the difference between the traffic flow of the set flow direction of the road section unit and the traffic flow
  • the absolute value of the difference between the south and north flow of the point of interest D2 in Figure 2 is calculated by
  • 6 in the previous cycle, Greater than the set difference 2, the south flow direction is the main direction of the attention point, the unit is the number of vehicles, the forecast remains the same as before, continue to set the south flow direction as the main direction of the attention point, and no other flow direction is the main direction of the attention point , send the main direction and flow information of the predicted focus point D2 to the “single flow direction-south-optimized signal preliminary judgment unit”, “dual flow direction-south and west-optimized signal preliminary judgment unit” and “double flow direction-south and west-optimized signal preliminary judgment unit” of green wave optimization in the analysis layer respectively Flow direction-south and east-optimized signal preliminary judgment unit” as shown in Figure 3, the "single flow direction optimized signal preliminary judgment unit” marked 3-4 and the "dual flow direction
  • the forecast remains the same as before, and continue to set the west flow direction as the attention point.
  • Point to the main direction, and there is no other flow direction as the main direction of the attention point, and the predicted main direction and flow information of the attention point D2 are respectively sent to the analysis layer in the green wave optimization "single flow direction-west-optimized signal preliminary judgment” and “dual flow direction- South and West - Preliminary Judgment of Optimizing Signal” "Dual Flow - West and North - Preliminary Judgment of Optimizing Signal”;
  • E1-S31 decide to optimize the signal: 1), according to the traffic mode obtained by receiving the command signal mode or analyzing the predicted traffic information, match the corresponding superimposed signal mode from the signal mode library to form the optimized signal
  • the preliminary judgment also includes: 1.1), the preliminary judgment steps of the green wave optimization signal include: 1.1.1), determining the flow direction of the traffic mode: calculating the flow of the main direction of the attention point in each flow direction to reach the set value of the flow rate or/or set the difference with the flow rate
  • the value is the sum of the traffic in the main direction of the attention point.
  • the flow direction with a large sum value is set as the main traffic direction of the road network, that is, the flow direction of the traffic mode.
  • the main traffic direction of a road network is a single traffic mode flow direction, which is recorded as a one-dimensional traffic-flow direction, and the traffic mode flow direction without a road network traffic main direction is recorded as a zero-dimensional traffic-unidirectional direction;
  • the sum of the traffic whose flow difference reaches the set value 2 in the south direction is the predicted value 8 of the flow rate of the vertical road section (4, 3) at the attention point
  • the sum of the flow of all the flows whose flow difference reaches the set value 2 in the west direction is the predicted value 9 of the traffic flow on the horizontal road section (3, 3) of the attention point
  • the two main traffic directions of the road network, south and west are used as the dual flow directions of the predicted traffic mode, that is, two-dimensional traffic-south and west; it can also be based on other traffic information
  • Data prediction other road network traffic flow characteristics include two-dimensional traffic-other flow direction, zero-dimensional traffic-unidirectional, one-dimensional traffic
  • E1-S41 execute optimization signal: mode transition period control: first run to complete the mode transition period, and then run the new cycle signal;
  • Embodiment 2 referring to Fig. 4, on the basis of the parameter, structure and method of embodiment 1, update and increase include
  • E2-S11 obtain the road network parameters, including the length of each road section in the road network and its traffic time, as shown in the mark 3 in Figure 4; the traffic time also includes the time when the congested motorcade of the road section starts; wherein the congested motorcade starts when it is used Refers to the time taken by the congested fleet from the first car to leave the original place to the last car to leave the original place, which is equal to the fleet start factor * congestion coefficient * road section length, where the congestion coefficient range is a number less than or equal to 1, and when it is equal to 1, it means serious congestion , or include the convoy starting coefficient calculated according to the experimental value range of 0.10 to 0.26, or take the middle as 0.18, unit: second/meter;
  • the signal mode functions also include blockage removal, mixing, and partial blockage removal by mixing; the mixing means that one flow is directed to the green wave and the other is flowed to the green wave to perform blockage removal; the mixed partial blockage refers to that one flow is directed to the green wave to perform guidance And the other part of the flow to the green wave to unblock or solitary wave is specially used for the following two-dimensional green wave-flow direction-mixed part of the unblocking, or refers to a part of the flow to the green wave to perform guidance and the other part of the road to perform unblocking or The solitary wave is specially used for the following one-dimensional green wave-flow direction-mixing or partial dredging; its green wave flow direction or the signal pattern combined with the green wave function, which is included in the signal pattern column of the signal pattern library in Table 6 in this embodiment Next, as shown in model numbers 4 to 17, including two-dimensional green wave-flow direction-removal of blockage, two-dimensional green wave-flow direction-mixing, two-dimensional green wave-flow direction
  • Described traffic pattern situation also comprises congested, mixes, and mixed part is congested; Described mixing refers to the situation that one flow is congested and the other flows to non-congested i.e. movement; The mixed part congests and refers to that one flows to traffic situation and another flow to traffic situation is partial Road section movement part road section is congested; Described traffic flow direction or combination with traffic situation, be included in the present embodiment under the signal pattern column of signal pattern storehouse in table 6, as shown in pattern serial number 4 to 17, comprise two-dimensional traffic-flow direction-congestion , two-dimensional traffic-flow-mixed, two-dimensional traffic-flow-mixed with partial congestion; degenerate includes one-dimensional traffic-flow-jammed, one-dimensional traffic-flow-mixed, one-dimensional traffic-flow-mixed with partial congestion; Parameters and their values as listed in the traffic mode in Table 6:
  • the function optimization signal that the green wave function of the signal pattern matches with the traffic pattern situation also includes the green wave function-thaw congestion matching traffic pattern situation-congestion, the green wave function-mixing and matching traffic pattern situation-mixing, the green wave function-mixing part dredging Congestion matching traffic mode situation-mixed partial congestion;
  • the optimization signal of green wave mode and traffic mode matching also includes the combination optimization of two-dimensional green wave flow direction and function combination matching two-dimensional traffic flow direction and situation combination, included in table 6 in the present embodiment
  • Under the signal mode column and matching traffic mode column in the signal mode library as shown in the mode number 4 to 17, it includes two-dimensional green wave-flow direction-decongestion and two-dimensional traffic-flow direction-congestion matching, two-dimensional green wave-flow direction- Mixed and two-dimensional traffic-flow-mixing and matching, two-dimensional green wave-flow-mixed partial congestion and two-dimensional traffic-flow-mixed with partial congestion; one-dimensional green wave-flow-blocking and one-dimensional traffic-flow- Congestion matching, one
  • the current signal two-dimensional green wave-east and north-guided pattern absolute phase difference cycle remainder and cycle complement of the present embodiment are as shown in Table 1 in embodiment 1;
  • S201 select road network focus points: according to traffic flow and fleet distribution and past experience data, in the road network shown in Figure 4, 12 intersection communities ⁇ (2, 1), (5, 3) ⁇ often have peak traffic and peak fleet captains , it is determined to set the 12 intersection communities as intersection communities of concern, and each intersection in the area is equipped with traffic flow in the direction of arrival at the set position to indicate the arrival flow sensor and the vehicle fleet sensor, and obtain the traffic flow in the four directions of east, west, south, north, and south of the intersection
  • the traffic flow and the information of the first vehicle and the last vehicle of the fleet that is, the captain information, to express the characteristics of the road network traffic information
  • S203 the traffic information prediction and calculation processing unit of the intersection of the concerned point, referred to as the intersection unit, using the wavelet function as the excitation function
  • the three-layer BP neuron network that is, the wavelet neuron network prediction method, and the traffic data of the previous 25 signal cycles, predict the traffic flow and fleet length of each traffic flow direction at the intersection of the concerned point; the traffic flow
  • one attention point can be at most There are two main directions of attention points;
  • the traffic condition characteristic determination rule is that the difference between the fleet captain of the intersection unit setting the flow direction and the length of the road section reaches or is less than the set value 2, and the flow direction of the larger flow is a concern of the attention point
  • one point of interest can have at most two main direction states of the point of interest;
  • the sensors and their set traffic data, and the set prediction method to predict the predicted traffic characteristic data of the 12 points of interest in Figure 4, as shown in Table 7, the east flow direction and west direction of each point of interest The flow rate and its difference in the flow direction, the fleet length and the difference between the east flow direction and the west flow direction of each point of interest, and the flow direction where they reach or exceed the set value of the main direction;
  • the main direction and flow status information of these predicted points of interest are respectively sent to "Single-flow direction-East/West/South/North-optimized signal preliminary judgment unit" and "Double-flow direction-East and South/South and West/West and North/North and East-optimized signal preliminary judgment unit" of green wave optimization in the analysis layer ";
  • F in the table represents flow characteristics
  • Q represents vehicle fleet characteristics
  • E2-S31 in deciding to optimize the signal 1) Obtain the traffic pattern according to the traffic information analyzed and predicted, match the signal pattern, and form the preliminary judgment of the optimized signal including: 1.1), the preliminary judgment steps of the green wave optimized signal include: 1.1.1), determine Main traffic flow direction of the road network: calculate the points of interest in the same main direction. The main direction flow reaches the set value of the flow rate or/or the total value of the flow difference with the set value of the flow rate. The main direction with a larger value is set as the main direction of the community traffic That is, the flow direction of the traffic mode, including two-dimensional traffic-flow direction, one-dimensional traffic-flow direction, zero-dimensional traffic-homogeneous direction, calculated in Table 8:
  • this embodiment sets the set value to the flow direction 90% of the total number of road sections, then the two flow directions are set as the main direction of the two-dimensional traffic congestion, when only one flow direction has more flow congestion road sections, and all reach or exceed the flow congestion set value, then the flow direction
  • the flow direction is set as the main direction of the one-dimensional traffic jam.
  • the flow direction is set as the traffic partial congestion flow direction. When there is no flow direction.
  • the number of traffic congestion sections is set as the two-dimensional traffic-flow-movement, and the traffic feature combination includes two-dimensional traffic-flow-movement, or/or with zero-dimensional traffic-unidirectional-movement, or/or with one-dimensional Traffic-flow-movement, or/or with two-dimensional traffic-flow-jam, or/or with one-dimensional traffic-flow-jam, or/or with two-dimensional traffic-flow-mixed, or/or with one-dimensional traffic- Flow direction-mixing, or/or mixing with two-dimensional traffic-flow direction-movement with partial congestion, or/or mixing with one-dimensional traffic-flow direction-movement with partial congestion; according to the calculated and determined setting values in this embodiment, it is obtained as Table 7 analyzes and calculates the traffic mode status part, and the traffic mode flow direction is two-dimensional traffic-south and west-movement mixed with partial congestion, and two road sections are congested: horizontal road section (2,3) and (2,4);
  • the priority rule also includes the preliminary judgment of two-dimensional green wave-flow direction-blockage removal is prior to the preliminary judgment of two-dimensional green wave-flow direction-mixing, and the preliminary judgment of two-dimensional green wave-flow direction-mixing is prior to the one-dimensional green wave - Preliminary judgment of flow direction and plugging removal, the preliminary judgment of one-dimensional green wave-flow direction-guiding plugging removal is prior to the preliminary judgment of two-dimensional green wave-flow direction-guided mixed part of plugging removal, and the preliminary judgment of two-dimensional green wave-flow direction-guided mixed part of plugging removal Priority to the two-dimensional green wave-flow direction-leading preliminary judgment; the optimized signal of the decision: two-dimensional green wave-south and west-guided mixed partial blockage/solitary wave;
  • Two-dimensional green wave-south and west-guidance mixes part of the congestion/solitary wave, and the current signal, two-dimensional green wave-north and east-guidance, making the corresponding mode transition period of the two, 3.0) each intersection maintains the previous current signal The period is 68 seconds and the phase timing; 3.1) Calculate the transition duration of the two signal modes at each intersection: 3.1.1) Optimize the two-dimensional green wave of the signal-south and west-guided mixed partial unblocking/solitary wave,? ? ? Refer to the fast mode method for other modes.
  • the base point of the mode phase difference is selected at the most upstream intersection of the two-dimensional traffic flow, that is, the intersection (6, 4) marked by the dotted triangle in the upper right corner of Figure 2, and the green wave flow to the west is selected as the mainstream of the mode. direction, the green wave flow that passes through the model phase difference base point intersection (6, 4) to the south becomes the model sub-current direction, and the north-south channel becomes the channel phase difference base point of the main direction channel of all models,? ? ?
  • Flow direction #/#/# in the table represents the phase timing of the intersection represented by the table coordinates, such as "coordinate (0,1) east-west 22/12/” indicates the east flow direction and west flow direction of the intersection (0,1)
  • the straight-going phase of the intersection is 22 seconds, the left-turning phase is 12 seconds, and there is no right-turning phase;
  • phase difference is the remainder of the mode phase difference period at the intersection;
  • the above temporary solitary wave timing table for the westward congested convoy at intersection (3, 2) can be applied to relieve the intersection (4, 2) flows westward to the congested convoy, so the above-mentioned solitary wave temporary timing table executes 2 cycles, and the congested convoy at intersection (4, 2) is untied;
  • E2-S42 execute optimization signal: mode transition period control: first run to complete the mode transition period, and then run the new cycle signal; execute the solitary wave temporary timing table.
  • Embodiment 3 S101, increase the new signal mode green wave parameter-phase difference fluctuation; as the road network in embodiment 2, signal and attention point and traffic data configuration and setting; Increase in its S12, S22, S32 respectively as follows,
  • trq d/v0-(1/v0+a)*Q
  • ⁇ trq -(1/v0+a)* ⁇ Q
  • Described traffic mode establishes new parameter crossing and sets flow direction fleet length Q and variation ⁇ Q thereof, and their setting value;
  • Said signal pattern and traffic pattern matching increase green wave-intersection flow direction-phase difference fluctuation and traffic-intersection flow direction-phase fleet change
  • E3-S101-S2 predicting traffic information, adding a new processing function to the intersection unit of the concerned point: finding that the length of the fleet acquired by the sensor system of the designated flow direction reaches the set value and the characteristic data of the intersection also reaches the set value
  • the specified flow direction coordinates referred to as column crossing row flow to fleet fluctuation, or row crossing column flow to fleet fluctuation, sent to the analysis layer; in this embodiment, the flow direction of column crossing row flow to fleet fluctuation refers to row horizontal, east or west , a direction, the flow direction of the row crossing column to the convoy fluctuation refers to the column longitudinal direction, south or north, a direction;
  • E3-S101-S3 decide to optimize the signal, 1) the analysis layer adds a new optimization signal preliminary judgment unit: 4 respectively analyze the column fluctuation preliminary judgment unit at the intersection of the sensing system of the 4-column assembly fleet in this embodiment, and 3 analyze respectively
  • the row fluctuation preliminary judgment unit of the intersection of the 3-row assembly team sensor system the column fluctuation preliminary judgment unit analyzes and specifies whether the number of intersections of the row crossings of all intersections in the same column flows to the fleet fluctuation reaches the set value, To determine the traffic mode-column intersection row flow-vehicle change, match the green wave-column intersection row flow-phase difference fluctuation preliminary judgment; the row fluctuation preliminary judgment unit analyzes the column intersection row flow direction fleet fluctuation of all intersections specified in the same line Whether the number of crossings reaches the set value to determine the traffic mode-column crossing flow-vehicle change, match the green wave-column crossing flow-phase difference fluctuation preliminary judgment; 2) the priority rule increases the fluctuation preliminary judgment only prior
  • E3-S101-S4 execute optimization signal: add phase difference fluctuation control: first run to complete the current phase difference cycle, and then run the phase difference fluctuation, which is the new phase difference cycle signal;

Abstract

一种交通信号智能控制架构方法，包括步骤：1)创建信号模式库，其一个信号模式提供一种交通模式特征的高效服务，2)预测交通信息特征，3)分析交通模式，4)从信号模式库中匹配与交通模式相应信号模式的优化信号，5)根据优先规则决定优化信号，6)制作从现行信号转换成优化信号的模式过渡期，7)执行优化信号。本方法的优点1)最突出的是其匹配系列智能预测交通模式于其专有的叠加信号优化模式库的控制架构，每个信号模式为与之匹配的交通模式提供极为高效交通服务，准确地构建在了人-机能力边界上，可极大地减少行车等待，远超30％，不能通过现有人工智能的算法优化出来；2)实现了交通流量流向双广谱负载模式化高效控制与高效响应；3)是接收动态学习创建新信号模式入库的活的架构。

Description

交通信号智能控制架构方法 技术领域
本发明涉及交通控制领域。具体地说，是交通信号人工智能控制架构方法。
背景技术
近年提出了多种方法创建高效的交叉两个方向传播的两维绿波信号模式，以及基于两维绿波多交叉传播方向绿波信号模式，即叠加绿波信号(Superpositioned Signals，SPS)；这些高效的叠加绿波信号对于设定的网络交通流具有最优控制效果；将使用包括人工智能(Artificial Intelligence，AI)在内的预测方法预测到的交通信息与这些叠加绿波信号模式及时匹配调用，可以极大地提高交通效率。
发明内容
本发明的目的是为解决信号模式对交通信息的响应优化架构问题。
本发明提出了实现上述目的解决方案，提出了使用包括人工智能方法的交通信息预测-信号模式的控制架构，即AI-SPS架构。具体如下：
一种交通信号智能控制架构方法，其特征在于包括：
S1，获取路网参数；创建若干路网交通信号模式，简称信号模式，组成信号模式库，每种信号模式提供该模式相应的路网交通特征的控制服务，与路网交通特征匹配的构成优化信号；获取或配置路网现在运行的信号模式及其参数，简称现行信号及其参数；
所述路网参数包括路网中各个路段长度及其交通用时；该交通用时包括行车用时或/或与拥堵车队启动用时；该行车用时指车辆以设定车速行驶完成路段所用时间，等于该路段长度除以设定行车速度，或包括减去设定行车速度的刹车时间；该拥堵车队启动用时指拥堵车队从队首车驶离原地到队尾车驶离原地所用时间，等于车队启动系数*拥堵系数*路段长度，该拥堵系数等于车队长度与路段车队之比，等于1时表示严重拥堵，或包括该车队启动系数按实验值范围0.10至0.26计算，或取中为0.18，单位：秒/米；注：本申请中“或/或与表述，”和“或与表述，”仅仅指紧随该“或/或与”和“或与”之后的“表述”，不涉及逗号后续列出的“表述”，例如“表述1，或/或与表述2，表述3”中“表述1”和“表述3”是并存关系，“表述2”是与其它表述是“或”存关系，或，“或与”存关系；
所述信号模式包括任何设定路口间相位间绿灯传播流向的信号，称为绿波；该流向作为信号模式绿波的特征参数称为绿波流向，模式记为绿波-流向；该设定绿波流向由各路口间运行的比率信号配置的相位时间差即相位差的大小排序所决定，从较小相位差路口流向较大相位差路口；两个相邻路口间的相位时间差是相对相位差，等于该相邻路口间路段设定的交通用时；各路口间运行的比率信号间没有相位差的信号模式是驻波，不传播绿灯变化，即比率模式，绿波退化模式，提供各流向均等交通特征的控制服务功能；所述比率信号指路口根据设定周期和设定比率配置各相位时长；该绿波特征参数或还包括功能称为绿波功能，该功能包括引导或均衡，或/或与疏堵，或/或与混合，或/或与对流引导即对引，其中，引导指当绿波流向与所控交通流向相同，各路段交通用时采用设定行车用时计算相对相位差；均衡指绿灯信号无流向，各路口0相位差同步；疏堵指绿波流向与所控交通流向相反，各路段交通用时采用设定拥堵车队启动用时计算相对相位差；混合指当绿波流向与所控交通流向既包括相同部 分也包括相反部分，相同流向部分的路段采用行车用时而相反流向部分的路段采用拥堵车队启动用时；对引指对相对的两个方向的交通实行引导，是引导的一种；绿波流向或与绿波功能组合为绿波优化信号，其模式分别记作绿波-流向-引导，绿波-流向-疏堵，绿波-流向-混合，驻波-均向-均衡；该信号模式参数还含周期，各相位分配时长即相位配时；该相位包括代表交叉路口控制方向的方向相位，或与方向相位中控制左/右转向的分流相位；
所述路网交通特征，简称交通模式，参数包括流向，或/或与状况，或/或与流量，或/或与路口各相位分流量；其中，状况包括运动即非拥堵，或/或与拥堵，或/或与混合；拥堵指路段长度减去车队长度达到或小于设定值，或拥堵系数达到或超过设定值，的车队状况，该设定值简称为路段拥堵设定值；混合指部分路段拥堵部分路段非拥堵；所述交通模式参数流向或与交通模式参数状况组合的交通模式，记作交通模式-流向-运动，或与交通模式-流向-拥堵，或与交通模式-流向-混合，或与交通模式-均向-运动；
所述信号模式与路网交通特征的匹配指与交通模式各对应参数的匹配，包括绿波流向与交通模式流向一致匹配的绿波流向优化信号，或与绿波功能与交通模式状况对应匹配组合的绿波优化信号，包括绿波-流向-引导匹配交通模式-流向-运动，驻波-均向-均衡匹配交通模式-均向-运动，或与绿波-流向-疏堵匹配交通模式-流向-拥堵，或与绿波-流向-混合匹配交通模式-流向-混合；或/或与信号模式周期与交通模式流量对应匹配的周期优化信号；或/或与信号模式相位配时与交通模式相位分流流量对应匹配的相位配时优化信号；其中，“流向一致匹配”指流向相同，“对应匹配”指根据该匹配涉及该两个模式参数组数值之间设定的系列数值组中交通模式参数数值所对应信号模式参数数值，可选择的数值组包括交通模式参数状况与信号模式参数功能数值组，交通模式参数流量与信号模式参数周期时长数值组，交通模式参数相位分流流量与信号模式参数相位配时数值组，交通模式参数车队队长相关值与信号模式设定对应参数数值组；
S2，预测交通信息：根据获取路网的交通信息，包括实测或/或与过去的若干设定时段，m＞＝1个周期c，的各流向车流量特征或/或与车队特征，预测未来设定若干时段，n＞＝1个周期c，的路网交通信息；当不设定所述预测未来若干周期时，下周期为所预测时段；
S3，决定优化信号：1)根据接到指令信号模式或分析所述预测交通信息得到的交通模式，从所述信号模式库中找到与交通模式匹配的信号模式，形成所述优化信号的初步判断，包括1.1)绿波优化信号初步判断，1.2)或与周期优化信号初步判断或与相位配时优化信号初步判断，所述指令信号模式指无需与交通信息匹配的强制执行所指令的信号模式；2)或与根据优先规则从所述初步判断中决定出优化信号；3)根据决定的优化信号和现行信号制作相应的模式过渡期；
S4，执行优化信号：模式过渡期控制：先运行完成模式过渡期，后运行新周期信号；
根据本发明所述交通信号智能控制架构方法：其特征在于包括，
S11，所述信号模式库中的信号模式包括叠加信号；
所述路网参数包括路网拓扑形式：拓扑平行四边形，或包括虚拟路口、虚拟路段；
所述叠加信号指设定时段(含若干个n＞＝1信号周期c)内运行2个及以上交叉流向的绿波信号，也就是任何基于交叉的两个流向的绿波，即两维绿波，的信号，及其任何降维形式；当所述两维绿波中的一个流向绿波相位差被配置为0时，就降维成一维绿波；
所述绿波参数流向还包括交叉的两个流向，该绿波称为两维绿波，或记为绿波-双流向或两维绿波-流向，其降维形式包括一维绿波-流向，及0维驻波-均向；所述绿波优化信号，包括两维绿波-流向-引导，一维绿波-流向-引导，0维驻波-均向-均衡，两维绿波-流向-疏堵，一维绿波-流向-疏堵，两维绿波-流向-混合，一维绿波-流向-混合，或与两维绿波-流向-对引，一维绿波-流向-对引：
所述交通模式参数流向还包括交叉的两个主流向，该交通模式称为两维交通，或记为交通-双流向或两维交通-流向，其降维形式包括一个主流向的为一维交通-流向，及0维交通-均向；所述交通模式，包括两维交通-流向-运动，一维交通-流向-运动，0维交通-均向-运动，两维交通-流向-拥堵，一维交通-流向-拥堵，两维交通-流向-混合，一维交通-流向-混合，或与两维交通-流向-对流，一维交通-流向-对流；
所述绿波优化信号还包括所述两维绿波与所述两维交通匹配的优化信号，即两维绿波优化信号，可选择的包括两维绿波-流向-引导与两维交通-流向-运动匹配，一维绿波-流向-引导与一维交通-流向-运动匹配，0维驻波-均向-均衡与0维交通-均向-运动匹配，或与两维绿波-流向-疏堵与两维交通-流向-拥堵匹配，或与一维绿波-流向-疏堵与一维交通-流向-拥堵匹配，或与两维绿波-流向-混合与两维交通-流向-混合匹配，或与一维绿波-流向-混合与一维交通-流向-混合匹配，或与两维绿波-流向-对引与两维交通-流向-对流匹配，或与一维绿波-流向-对引与一维交通-流向-对流匹配；
S21，所述预测交通信息中的交叉的两个及以上个流向交通信息，或与其状况；
根据本发明所述交通信号智能控制架构方法：其特征在于包括，
S12，所述叠加态信号还包括，设定交通特征条件下让路口一个相位的设定车队获得若干个连续路口绿灯信号，简称孤波，叠加在两维绿波中；
所述信号模式参数还包括所述若干个连续路口的路径临时配时表，称为孤波临时配时表；
所述设定交通特征条件包括指令，让指定路口流向相位的设定车队强制占用前方路口其它相位车队的绿灯时间，通过，简称强制孤波；或/或与让指定路口流向相位的设定车队占用前方路口其它相位车队的绿灯剩余时间，简称余时，通过，简称余时孤波；两者统称孤波交通；
所述信号模式匹配交通模式的优化信号还包括孤波-流向-孤波临时配时表与设定车队-流向-路径路口表，孤波临时配时表包括强制孤波临时配时表或/或与余时孤波临时配时表；
S22，所述预测交通信息还包括，实时实测与预测的所述设定交通特征条件包括指令孤波：包括强制孤波，或/或与余时孤波；强制孤波临时配时表的配置计算：强制指定路口流向相位含方向相位或分流相位的设定车队占用前方路口其它相位车队的相位绿灯时间，依此计算建立所述强制孤波临时配时表；余时孤波临时配时表的配置计算：指定路口流向相位含方向相位或分流相位设定车队占用前方路口相关相位车队提供本相位后剩余的绿灯时间通过，依此计算建立所述余时孤波临时配时表；计算所述余时的方法简称A-A算法，包括使用各设定路口流向交通相位含方向相位分流相位队尾q特殊信息，长车队；所述余时孤波请见中国专利201710897777.6；
S32，所述决定优化信号1)预测交通模式与信号模式的匹配还包括，接到指令孤波，确定路口相位设定车队，将孤波叠加于现行信号模式上的优化信号初步判断，包括用孤波预测算法建立获得相关若干个连续路口绿灯信号的孤波临时配时表；所述2)或与所述优先规则还包括强制孤波初步判断优先于余时孤波初步判断，孤波初步判断优先于其它初步判断；
S42，所述执行优化信号还包括：执行孤波临时配时表。
根据本发明所述交通信号智能控制架构方法：其特征在于包括，
S13，所述叠加态信号还包括，设定交通信息特征条件下让距离路口一个相位最近的行车，即相位首车，由红灯获得绿灯信号，称微分绿波，叠加在两维绿波中；所述行车指以规定速度正常行驶的车辆；
所述信号模式参数还包括相位间的绿灯与红灯间的最小安全切换通过时间，称为微分绿波时间，或相变量子用时Δt Th0，及所涉及两相位；所述相位最小安全切换通过时间指能够让相位首车在其控制相位信号由绿灯切换成红灯时，以正常刹车安全停于前方停车线，同时让其它相位首车在其控制相位信号由红灯切换成绿灯后安全通过前方路口的最小时间；以该相位最小安全切换通过时间所对应计算出的相位首车与前方控制相位的距离就是相位最小安全距离；
所述设定交通信息特征条件包括各设定路口流向交通相位红灯时的相位首车q0处于或小于该相位设定的相位最小安全距离，而该路口比率信号绿灯的相位首车q0处于其相位的最小安全距离外或与相应的行人相位也无行人，即比率信号绿灯相位无车或与无行人，合称为红灯交通相位首车在可微分位置的微分交通，也称微分绿波条件；或与在下列设定顺序规则下延续微分绿波条件，当“正在占用比率信号绿灯时间的”绿灯相位继续检测到有车时，或与该同流向其它分流相位检测到有车时，或其它流向检测到有车时；
所述信号模式与交通模式匹配的优化信号还包括微分绿波时间及其两相位与微分绿波条件及其两相位匹配；
S23，所述预测交通信息还包括，实时实测的所述设定交通特征条件的微分绿波条件；
S33，所述决定优化信号所述1)预测交通模式与信号模式的匹配还包括，根据实测到微分绿波条件或接到指令微分绿波信号模式，为在现行信号模式下满足所述微分绿波条件的红灯相位首车，匹配叠加于现行信号模式上微分时间的优化信号初步判断，即让在现行信号模式下满足上述“微分绿波条件”的红灯相位首车获得占用“比率信号绿灯相位无车或与无行人”相位的一个所述相变量子用时或称所述微分时间；所述2)或与优先规则还包括微分绿波初步判断优先于其它初步判断；
S43，执行优化信号还包括：立即将绿灯切换给或保持给获得微分时间的相位；
根据本发明所述交通信号智能控制架构方法：其特征在于包括，
S101，所述信号模式库包括增加模式或模式参数的操作，设定一个新信号模式提供设定交通模式的服务或服务于设定的交通模式，将这一信号模式参数与其服务的交通模式匹配，及其优化信号初步判断及其优先级、实现操作，分别配置到S1中信号模式库中，S2中预测交通信息中，S3中决定优化信号的初步判定优先规则中，S4执行优化信号的操作中；
根据本发明所述交通信号智能控制架构方法：其S2预测交通信息特征在于包括，
S201，所述实测或预测路网交通信息及其特征包括指通过对路网中设定若干装配交通传感装置的关注点，路段中的设定位置或/或与交叉路口，用来获取所述关注点的设定流向的交通流量特征数据或/或与车队特征数据，以分析表达所述路网交通信息特征；
根据本发明所述交通信号智能控制架构方法：其S2预测交通信息特征在于包括，
S202，所述关注点设定流向的交通流量特征数据包括相应的若干交通流量设定值或与其变化设定值，或/或与该设定流向相对流向的流量之差设定值，达到所述流量设定值或流量之差设定值的流向作为该关注点主流向；或与所述该关注点设定流向的车队特征数据包括相应的若干车队长度设定值或与其变化设定值；或/或与该设定流向车队长度与其所在路段长度之差设定值，达到该之差设定值的状况为该关注点的设定流向拥堵路段；
根据本发明所述交通信号智能控制架构方法：其S2获取预测交通信息特征在于包括，
S203，所述关注点配置用于该点交通信息预测的计算处理单元，简称关注点单元，每个关注点单元用设定预测方法对应的关注点设定流向的交通信息做预测；若干个所述关注点单元组成预测层；所述预测方法指任何根据已有数据预测未来数据的方法，包括直接使用实时检测数据作为预测未来数据的重复式预测法，经验方法，均值法，最大值法，统计优化方法，含神经元网络方法和专家系统方法的人工智能方法。
根据本发明所述交通信号智能控制架构方法：其S2获取预测交通信息特征在于包括，
S204，所述关注点路口用于自身设定流向的的交通信息预测计算处理单元，简称路口单元，每个路口单元或与还负责预测流向相位分流相位交通信息特征，或涉及包括流向上游临近路段的车源或与流向上游路口单元交通信息特征及其设定值。
根据本发明所述交通信号智能控制架构方法：其S2预测交通信息特征在于包括
S205，所述关注点的数量、分布、路段或路口的类型是通过所述设定预测方法预测的流量流向分布决定。
根据本发明所述交通信号智能控制架构方法：其S3决定优化信号特征在于包括，
S301，所述绿波优化信号初步判断包括绿波流向优化和绿波功能优化，步骤包括1)确定交通模式流向：计算各同流向的关注点主流向流量达到流量设定值或/或与流量差设定值即关注点主流向的流量之和，该之和值大的流向设定为路网交通主流向，即交通模式流向，包括交叉的两个路网交通主流向的为交通模式双流向，记为两维交通-流向，一个路网交通主流向的为交通模式单流向，记为一维交通-流向，没有路网交通主流向的为交通模式流向，记为零维交通-均向；2)或与确定交通模式状况：计算各同流向的关注点车队长度达到车队长度设定值或/或与其路段长度之差达到或小于路段拥堵设定值的路段数，简称流向拥堵路段数，之和，该之和值大的流向设定为路网交通拥堵主流向，即交通模式拥堵主流向，当交叉的两个流向都拥有较多该流向拥堵路段数，且都达到或超过该流向设定值即流向拥堵设定值，则该两个流向是两个路网交通拥堵主流向，记为交通-双流向-拥堵或两维交通-流向-拥堵，当仅该一个流向拥有较多流向拥堵路段数，且达到或超过该流向拥堵设定值，而另一个流向没有流向拥堵路段数，则该流向是一个路网交通拥堵主流向，记为一维交通-流向-拥堵，当没有流向拥有流向拥堵路段数达到或超过该流向拥堵设定值，则该流向为该部分拥堵流向，包括小区域拥堵流动或几个路段拥堵流向；与交通模式流向组合特征包括两维交通-流向-运动， 或/或与零维交通-均向-运动，或/或与一维交通-流向-运动，或/或与两维交通-流向-拥堵，或/或与一维交通-流向-拥堵，或/或与两维交通-流向-混合，或/或与一维交通-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵，或/或与一维交通-流向-运动混合含部分拥堵；3)决定匹配的绿波优化参数及其分布：两维交通-流向-运动匹配两维绿波-流向-引导，或/或与零维交通-均向-运动匹配零维驻波-均向-均引，或/或与一维交通-流向-运动匹配一维绿波-流向-引导，或/或与两维交通-流向-拥堵匹配两维绿波-流向-疏堵，或/或与一维交通-流向-拥堵匹配一维绿波-流向-疏堵，或/或与两维交通-流向-混合匹配两维绿波-流向-混合，或/或与一维交通-流向-混合匹配一维绿波-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵匹配两维绿波-流向-引导混合部分疏堵，或/或与一维交通-流向-运动混合含部分拥堵匹配一维绿波-流向-引导混合部分疏堵；或与绿波-流向-引导混合部分疏堵包括局部疏堵、孤波疏堵，所述局部疏堵指将局部拥堵作为小区路网配置为相应的路网疏堵信号模式，也包括所述两维绿波-流向-疏堵和一维绿波-流向-疏堵，所述孤波疏堵指使用所述孤波技术将拥堵路段的车队疏解开，也包括余时孤波，指令孤波；
或/或与直接用所述预测方法进行绿波优化信号预测初步判断：其所预测目标的数据包括指定交通特征数据，用所述预测方法调整优化控制该交通的信号参数，根据已有指定交通特征数据和绿波参数数据组来优化出满足预测目标的交通特征数据的信号参数的优化信号初步判断，称为直接优化信号预测方法；
根据本发明所述交通信号智能控制架构方法：其S3决定优化信号特征在于包括，
S302，周期优化信号初步判断，或与相位配时优化初步判断，还包括使用任何路口信号配时算法进行预测优化，韦伯斯特配时法，冲突点法，估算法，临界车道法，和周期相位配时优化信号相关的预测方法，称为周期相位配时优化信号预测方法；
根据本发明所述交通信号智能控制架构方法：其S3决定优化信号特征在于包括，
S303，所述预测方法，还包括直接优化信号预测方法和周期相位配时优化信号预测方法，通过以车辆等待时间最小化或/或与车队等待长度最小化为目标的优化信号学习方法，发现路网的交通信息关注点获取的交通信息特征数据与相对应的优化信号参数数据之间的新变化关系，将该新变化关系创建为新优化信号，加进信号模式库；
根据本发明所述交通信号智能控制架构方法：其S3决定优化信号特征在于包括
S304，所述优先规则根据包括计算预测/估计各优化信号初步判断将引起的行车等待时间减少/车队车队减少/流量增加的数据排序决定各优化信号初步判断的优先排序；
根据本发明所述交通信号智能控制架构方法：其S31决定优化信号特征在于包括
S305，所述优先规则根据所述估计数据的大小排序，排在前者优先于后者：包括指令信号模式，或与两维绿波-流向-疏堵初步判断，或与两维绿波-流向-混合初步判断，或与一维绿波-流向-疏堵初步判断，或与一维绿波-流向-混合初步判断，或与两维绿波-流向-对引初步判断，或与一维绿波-流向-对引初步判断，两维绿波-流向-引导初步判断，一维绿波-流向-引导初步判断，驻波-均向-均衡初步判断；
根据本发明所述交通信号智能控制架构方法：其特征在于强制孤波临时配时包括，
F201，根据预测获取的孤波车队长度，计算通过本地路口所需的时间，简称强制孤波用时；用本地相位绿灯时间减去这个强制孤波用时，得到小于0的数，则将其它相位相应量的时间借过来使用；
根据强制孤波车队长度和预测达到前方路口的时间，预测前方路口相关相位的信号灯色，及前方路口同关相位绿灯时间中减去一个设定时间后的剩余绿灯时间，再减去这个强制孤波用时，得到剩余绿灯时间，小于0，则该绝对值就是预测借用其它相位相应量的时间；其中设定时间是假设该相位车队通过用时；
如上预测计算设定若干个连续路口；制定所述若干路口的临时配时表；
本发明优点如下：1)匹配系列预测交通模式于其专门的叠加信号优化信号库的控制架构，构建在人-机能力边界上，即准确地使用人工智能在其所无能及人能力所擅长和其所擅长和人能力所难为之处，可极大地减少行车等待，远超30％，不能通过现有人工智能的算法优化出来，无法创造信号模式及其控制架构就像人工智能无法发现创造定律F＝MA，E＝MC 2，麦克斯韦方程组一样；2)该结构具备通过智能学习扩展优化信号模式；3)将广谱交通流量负载控制和广谱交通流向负载控制统一在一个信号模式控制架构下，实现了交通流量流向双广谱负载高效控制与高效响应；4)恰当重复运用分析控制交通模式分布与布置设定关注点可产生众多组合优化控制方案；5)恰当选用控制路网设定关注点及其设定预测交通信息特征可节省设备成本与运行成本，提高运行效率；6)当选用路网所有路口作为设定关注点及车队作为设定交通信息特征时，可以获得最高控制效率和控制精度。
附图说明
图1是交通信号控制架构方法流程图；
图2是2个设定关注点路段与交通数据-流量图；
图3是2个设定关注点路段的路网交通信号控制架构“预测决策”结构图；
图4是小区设定关注点路口与交通特征-流量或车队长度图；
图5是小区设定关注点路口的路网交通信号控制架构“预测决策”结构图；
附图中的编号索引：
图2：1--网络路口节点编码标识起始点(0，0)是路网的左下角路口，记号{(0，0)，(6，4)}代表原点是(0，0)，纵横坐标(i，j)最大值(6，4)即各为6，4，每个坐标(i，j)处是两条道路的交叉路口，有7条道路与5条与之交叉的道路，交叉出35个交叉路口；2--2个设定关注点路段，用顺段落流向的菱形和字母D#代表，分别有D1，D2，分布于2个不同路段，横路段(3，3)，纵路段(4，3)，该关注点装配有双向交通信息流量采集装置及关注点单元，每个路段关注点位置围绕着一组“东”与“西”或“南”与“北”方位标注及其各引出的并用双尖括号“《”隔开的2个数字#-，表示该方位“实测流量#-《预测流量#-”，如“南8-《8-”表示该路段南流向的实测流量是8，预测流量8，该方位位向与交通流向一致，单位是辆-车数；3--路口间距-行车用时/拥堵车队启动用时被记作#-#/#：单位：米-秒/秒；注1：图左上角路口(0，4)是虚拟路口，用虚线小方块代表，其相连的行路段(0，4)是虚拟路段，用虚线代表；注2：图路网左下角路口(0，0)的三角形标注表示该路口为现行叠加态信号的模式相位差基点路口，两个空心实线箭头指向现行信号绿波流向，东和北；注3：图路网右上角路口(6，4)的虚线三角 形标注表示该路口将被配置为优化叠加态信号的模式相位差基点路口，两个空心虚线箭头指向优化信号绿波流向，西和南；以下图中通用本图标注，另外声明定义的除外；
图3：1-信号模式库；2-设定关注点路段单元与交通信息数据接收及预测模块，2个关注点路段单元组成预测层，3-空心箭头表示层与层间，模块与模块间数据连接，箭头所指是预测层与分析层数据连接，4-分析层中单流向优化信号初判单元或与功能优化信号，箭头所指这个是北流向优化信号初判单元，5-分析层中双流向优化信号初判单元或与功能优化信号，箭头所指这个是东流向与北流向优化信号初判单元，6-分析层中周期及绿信比优化信号初判，7-决策层中信号优化决策；以下关注点预决策结构图中通用本图标注，另外声明定义的除外；图4：2-12个设定关注点交叉路口小区中之一的设定关注点路口(3，1)，箭头所指，用八角形代表，该关注点装配有路口交通流向相位交通信息流量采集装置，或分流相位交通信息流量采集装置，或车队队头队尾信息采集装置，围绕着该交叉路口的“东”“西”“南”“北”方位标注及其各引出的用双尖括号“#-#/#/#《#-#/#/#”隔开的的8个数字#表示该方位流量#-和直#/、左#/、右#三相位实测等待通过路口车队长度和预测的相关值，如“北3-0/0/0《8-7/0/0”表示该路口北侧等待南行的实测流量是1，车队直行相位1，左转0，右转0，预测流量8，车队长7/0/0，该方位位向与交通流向相反，单位是车数，车数队长q相应秒数用标准车长6.25米换算成米队长再用队扰时间tqx＝(1/v0+a)*q换算成秒，用v0绿波秒速12.5米秒，α＝0.18及tqx＝0.26计算得到，如20米对应于5秒3辆车；注1：横路段(2，4)和横路段(2，5)上的西向黑箭头代表西行拥堵车队及其队长；注2：路口(3，2)中的虚线倒三角代表该路口是东流向疏堵绿波的模式相位差基点；
图5：2-设定关注点路口单元交通信息数据接收及预测模块，12个路口单元预测模块组成预测层，箭头所指是其中设定关注点路口(5，3)路口单元预测模块；8-分析层中孤波优化信号初步判断，9-决策层中孤波管理模块；10-分析层中微分绿波优化信号初步判断，11-决策层中孤波冲突解决模块；
具体实施方式
结合附图详细描述本发明3个实施例：
实施例1，参见图1和图2，
E1-S11，获取路网参数，包括路网中各个路段长度及其交通用时，如图2中标注3所示；所述交通用时包括路段的拥堵车队启动用时；其中行车用时指车辆以设定车速行驶完成路段所用时间，等于该路段长度除以设定行车速度，或包括减去设定行车速度的刹车时间；路网拓扑形式为包含有近似准平行四边形包括虚拟路口、虚拟路段；创建用于图2所示路网交通的若干叠加信号模式，组成信号模式库，其每种信号模式提供该模式相应的交通模式的控制服务，与交通模式匹配的构成一个优化信号；获取或配置现在运行信号模式及其参数，称为现行信号及其参数，包括信号周期68秒，各路口相位配时是：南北流向与东西流向相位配时比为1∶1，各为34秒，各流向有分流直行左行相位配时比2∶1，各为22秒，12秒；
现行信号如图2标注，两维绿波-东与北-引导，其模式相位差基点路口在路网左下角路口(0，0)，其2个交叉流向为东与北，各路口模式周期余数与周期补数如表1：
表1现行信号两维绿波-东与北-引导的路口模式周期余数与周期补数，周期＝68秒
4 42＞42/26 52＞52/16 60＞60/18 72＞4/64 82＞14/54 90＞22/46 102＞34/34
3 30＞30/38 40＞40/28 48＞48/20 60＞60/8 70＞2/66 78＞10/58 90＞22/46
2 20＞20/48 30＞30/38 38＞38/30 50＞50/18 60＞60/8 68＞0/68 80＞12/56
1 12＞12/56 22＞22/46 30＞30/38 42＞42/26 52＞52/16 60＞60/8 72＞4/64
0 0＞0/68 10＞10/58 18＞18/50 30＞30/38 40＞40/28 48＞48/20 60＞60/8
i/j 0 1 2 3 4 5 6
各路口模式绝对相位差#及其周期余数＞#周期补数/#，如表1中‘#＞#/#’；
所述信号模式包括任何设定路口间相位间绿灯传播流向的信号，称为绿波；该流向作为信号模式绿波的特征参数称为绿波流向，模式记为绿波-流向；该设定绿波流向根据其最上游路口，称为模式相位差基点或绿波源点，的相位时间计算配置其它各路口相关相位的相位差，该相位差称作模式绝对相位差；两个相邻路口间的相位时间差是相对相位差，等于该相邻路口间路段设定的交通用时；各路口间运行的比率信号间没有相位差的信号模式是驻波，不传播绿灯变化，即比率模式，绿波退化模式，提供各流向均等交通特征的控制服务功能；所述比率信号指路口根据设定周期和设定比率配置各相位时长；该绿波特征参数还包括绿波功能，包括引导或均衡或/或与疏堵或/或与混合或/或与对流引导即对引，其中，引导指当绿波流向与所控交通流向相同，各路段交通用时采用设定行车用时计算相对相位差；均衡指绿灯信号无流向，各路口0相位差同步；疏堵指绿波流向与所控交通流向相反，各路段交通用时采用设定拥堵车队启动用时计算相对相位差；混合指当绿波流向与所控交通流向既包括相同部分也包括相反部分，相同流向部分的路段采用行车用时而相反流向部分的路段采用拥堵车队启动用时；对引指对相对的两个方向的交通实行引导，是引导的一种；其绿波信号分别记作绿波-流向-引导，绿波-流向-疏堵，绿波-流向-混合，驻波-均向-均衡；该信号模式的参数还包括周期，各相位分配时长即相位配时；该相位包括代表交叉路口控制方向的方向相位，或与方向相位中控制左/右转向的分流相位；本实施例不包括疏堵功能因为没有配置车队长度传感系统；
所述叠加信号指设定时段(含若干个n＞＝1信号周期c)内运行2个及以上交叉流向的绿波信号，也就是任何基于交叉的两个流向的绿波，即两维绿波，的信号，及其任何降维形式；当所述两维绿波中的一个流向绿波相位差被配置为0时，就降维成一维绿波，同理，一维绿波就降维成零维驻波；所述绿波流向还包括交叉的两个流向，该绿波称为两维绿波，或记为绿波-双流向或两维绿波-流向，其降维形式包括一维绿波-流向，及0维驻波-均向；所述绿波信号，包括两维绿波-流向-引导，一维绿波-流向-引导，0维驻波-均向-均衡，或与两维绿波-流向-对引，一维绿波-流向-对引；本实施例中两维绿波-流向，其中流向包括东与北或北与东、东与南或南与东、西与北或北与西、西与南或南与西，其中流向包括东、南、西、北，及零维驻波-均向；所述两维绿波流向或与绿波功能组合，本实施例中包括在表2中信号模式库信号模式栏目下，如模式序号4至8所示，两维绿波-流向-引导，或与两维绿波-流向-对引，或与一维绿波-流向-对引，一维绿波-流向-引导，及零维驻波-均向-均引：该序号也作为优先级别号，较小的号优先于较大的号，序号2，3不在此列；
表2实施例1信号模式库中信号模式与交通模式匹配表
Figure PCTCN2021000212-appb-000001
注[1]：对引需要满足设定路网结构要求及其设定值；
设定所述2个交叉流向最上游的交汇点路口为所述模式相位差基点路口，即两维绿波基点路口，反之亦然，一个两维绿波基点路口位置也决定了所述2个交叉流向参数；一维绿波的模式相位差基点路口位置设定在所述绿波各通道最上游的同一交叉非绿波流向的通道上路口，即一维绿波基点通道路口；零维驻波模式相位差基点路口可设定任意一个路口；
所述交通模式参数包括流向，或/或与状况，或/或与流量，或/或与路口各相位分流量；其中，交通模式流向包括交叉的两个主流向，该交通模式称为两维交通，或记为交通-双流向或两维交通-流向，其降维形式包括一个主流向的为一维交通-流向，及0维交通-均向，本实施例中两维交通-流向包括东与北或北与东、东与南或南与东、西与北或北与西、西与南或南与西，其退化形式包括一个主流向的特征为一维交通-流向，包括东、南、西、北，及0维-均向；本实施例交通状况不包括拥堵，混合，交通模式状况包括运动即非拥堵；拥堵指路段长度减去车队长度达到或小于设定值，或拥堵系数达到或超过设定值，的车队状况，该设定值简称为路段拥堵设定值；本实施例交通模式流向或与交通模式状况组合的交通模式，包括在表2中信号模式库匹配交通模式栏目下，模式序号4至8所示，两维交通-流向-运动，一维交通-流向-运动，零维交通-均向-运动，或与两维交通-流向-对流，或与一维交通-流向-对流；
所述信号模式与交通模式匹配指各对应参数的匹配，包括所述信号模式流向与交通模式流向匹配的绿波流向优化信号；与所述绿波功能与所述交通模式状况匹配的绿波功能优化信号组合的绿波优化信号，本实施例中包括在表2中信号模式库信号模式栏目与匹配交通模式栏目下，如模式序号4至8所示，两维绿波-流向-引导与两维交通-流向-运动匹配，一维绿波-流向-引导与一维交通-流向-运动匹配，零维驻波-均向-均衡与零维交通-均向-运动匹配，或与两维绿波-流向-对引与两维交通-流向-对流匹配，或与一维绿波-流向-对引与一维交通-流向-对流匹配；或/或与所述信号模式周期与交通模式流量匹配的周期优化信号；或/或与所述信号模式相位配时与交通相位分流特征匹配的相位配时优化信号；
E1-S21，预测交通信息：
E1-S201-S205，选择路网关注点：根据流量分布和以往经验数据预测图2路网纵横往来四个方向上长期容易出现峰值流量和在两个路段是横路段(3，3)和纵路段(4，3)中位置，确定 设置该两个路段为路网中的关注点路段中点位置，配置双向交通流量传感器在所设定位置，用车道线圈或车辆定位数据，获取东西南北4个流向的交通流量，以表达所述路网交通信息及其特征；S203，所述关注点路段位置配置的交通信息预测计算处理单元，简称路段单元，使用重复式预测方法预测所述关注点每个交通流向交通流量；S202，其中交通流向特征判定指所述路段单元设定流向的交通流量与相对流动的交通流量之差绝对值达到或超过设定值2，较大流量的流向为该关注点的一个关注点主流向；本实施例一个关注点最多可以有一个关注点主流向；并根据所选择的设定关注点路段峰值流量或/或与预测的峰值流量，用韦伯斯特路口信号配时法，计算重配信号模式周期相位配时，周期68秒，方向相位比34秒∶34秒，分流相位直行左转22秒∶12秒；
根据所配置的关注点，交通传感器及其设定交通数据，和设定预测方法预测得到：图2中关注点D2的南与北流量差绝对值由前周期的|8-2|＝6，大于设定差值2，南流向为该关注点主流向，单位是辆-车数，预测保持与之前相同，继续设为南流向为该关注点主流向，没有其它流向为该关注点主流向，将预测的关注点D2主流向和流量信息分别送给分析层中绿波优化的“单流向-南-优化信号初判单元”“双流向-南与西-优化信号初判单元”“双流向-南与东-优化信号初判单元”如图3中标注3-4的“单流向优化信号初判单元”和3-5所示的“双流向优化信号初判单元”；关注点D1的东与西流量差绝对值由前周期的|1-9|＝8，大于设定差值2，西流向为该关注点主流向，预测保持与之前相同，继续设为西流向为该关注点主流向，没有其它流向为该关注点主流向，将预测的关注点D2主流向和流量信息分别送给分析层中绿波优化的“单流向-西-优化信号初步判断”“双流向-南与西-优化信号初步判断”“双流向-西与北-优化信号初步判断”；
E1-S31，决定优化信号：1)，根据接到指令信号模式或分析所述预测交通信息得到的交通模式，从所述信号模式库中匹配到所述相应叠加态信号模式，形成优化信号的初步判断还包括：1.1)，绿波优化信号初步判断步骤包括：1.1.1)，确定交通模式流向：计算各同流向的关注点主流向流量达到流量设定值或/或与流量差设定值即关注点主流向的流量之和，该之和值大的流向设定为路网交通主流向，即交通模式流向，包括交叉的两个路网交通主流向的为交通模式双流向，记为两维交通-流向，一个路网交通主流向的为交通模式单流向，记为一维交通-流向，没有路网交通主流向的为交通模式流向，记为零维交通-均向；本实施例中，根据计算所有南流向流量差达到设定值2的流量之和是关注点纵路段(4，3)流量预测值8，和所有西流向流量差达到设定值2的流量之和是关注点横路段(3，3)流量预测值9，得到两个路网交通主流向，南与西，作为预测交通模式双流向，即两维交通-南与西；还可以根据其它交通信息数据预测其它路网交通流向特征包括两维交通-其它流向，零维交通-均向，一维交通-其它流向；1.1.2)，本实施例中没有使用车队长度传感装置及其数据，不关注状况拥堵，默认状况运动，所以，本实施例预测路网交通流向与状况组合特征为两维交通-南与西-运动；还可以预测其它路网交通流向与状况组合特征包括两维交通-其它设定流向-运动，一维交通-其它流向-运动，零维交通-均向-运动；1.1.3)，决定匹配的绿波优化信号参数：本实施例中，两维交通-南与西-运动匹配两维绿波-南与西-引导；本实施例的关注点配置还可以决定匹配的绿波优化信号包括两维交通-其它设定流向-运动匹配两维绿波-对应其它流向-引导，一维交通- 流向-运动匹配一维绿波-对应流向-引导，零维交通-均向-运动匹配零维驻波-均向-均衡；或与两维交通-流向-对流匹配两维绿波-对应流向-对引；或与一维交通-流向-对流匹配一维绿波-对应流向-对引；1.2)周期优化信号初步判断与相位配时优化信号初步判断；交通模式流量或/或与车队长特征参数为若干设定值或与其变化设定值，预测匹配对应流量的叠加信号模式周期为该若干设定值对应的若干周期优化；交通模式相位流量或/或与特征车队长长度为若干设定值或与其变化设定值，预测匹配对应相位流量的叠加态信号模式相位配时为该若干设定值对应的若干优化相位配时，其中预测匹配或使用所述周期相位配时优化信号预测方法；
2)或与根据如下优先规则次序从所述初步判断中决定出优化信号，排序在前者优先于在后者：指令信号模式，两维绿波-流向-引导初步判断，一维绿波-流向-引导步判断，或与两维绿波-流向-对引初步判断，或与一维绿波-流向-对引步判断，或与或/或与周期优化信号初步判断与相位配时优化信号初步判断，驻波优化信号初步判断，所述“或与”优化信号初步判断仅仅指该或与的初步判断，不涉及后续列出的初步判断；本实施例中只会预测出一个交通模式，产生一个优化信号；决定的优化信号：两维绿波-南与西-引导；
3)根据决定的优化信号和现行信号制作相应的模式过渡期：决定的优化信号：
两维绿波-南与西-引导，和现行信号，两维绿波-北与东-引导，制作两者相应的模式过渡期，3.0)各路口保持之前现行信号周期68秒为优化周期68秒；3.1)计算各路口的所述两个信号模式的过渡时长：3.1.1)优化信号模式周期余数计算：两维绿波-南与西-引导，其模式相位差基点选在两维交通流向最上游交汇路口即路口(6，4)如图2中右上角虚线三角形标注的路口，将绿波流向西选做模式主流向，穿过模式相位差基点路口(6，4)的绿波流向南就成为模式副流向，该南北通道就成为所有模式主流向通道的通道相位差基点，每个路口的优化信号模式绝对相位差及其周期余数如下表3：表格中#＞#代表信号模式绝对相位差‘#’及其周期余数‘＞#’，以下符号同义；
表3优化信号两维绿波-南与西-引导的路口模式绝对相位差及其周期余数，周期＝68秒
4 60＞60 50＞50 42＞42 30＞30 20＞20 12＞12 0＞0
3 72＞4 62＞62 54＞54 42＞42 32＞32 24＞24 12＞12
2 82＞14 72＞4 64＞64 52＞52 42＞42 34＞34 22＞22
1 90＞22 80＞12 72＞4 60＞60 50＞50 42＞42 30＞30
0 102＞34 92＞24 84＞16 72＞4 62＞62 54＞54 42＞42
i/j 0 1 2 3 4 5 6
3.1.2)现行信号模式周期补数计算：两维绿波-北与东-引导，其模式相位差基点路口如图2路网左下角路口(0，0)，每个路口的现行信号模式绝对相位差周期余数的周期补数如上表1中#＞#/#的‘/#’值，现行周期＝68秒；
3.1.3)计算所述模式过渡时长：每个路口所述模式过渡时长等于每个路口的优化信号模式绝对相位差的周期余数加每个路口的现行信号模式绝对相位差周期余数的周期补数，取该和的周期余数得到所述每个路口模式过渡时长如下表4：表格中/#+＞#代表之前现行信号模式绝对相位差周期余数的周期补数‘/#’加优化信号模式绝对相位差的周期余数‘＞#’，周期＝68秒，
表4各路口模式过渡时长计算，周期＝68秒；
Figure PCTCN2021000212-appb-000002
Figure PCTCN2021000212-appb-000003
3.2)制作各路口所述模式过渡时长相应的模式过渡期：将模式过渡时长分成两个流向相位时长之和，或与将每个流向再分成分流相位时长之和，如下表5(未标出分流相位时长)：
表5各路口模式过渡期相位配时，周期＝68秒
4 ＞18＝9+9 ＞66＝33+33 ＞60＝30+30 ＞26＝13+13 ＞6＝延红灯 ＞58＝29+29 ＞34＝17+17
3 ＞42＝21+21 ＞22＝11+11 ＞6＝延红灯 ＞50＝25+25 ＞30＝15+15 ＞14＝延红灯 ＞58＝29+29
2 ＞62＝31+31 ＞42＝21+21 ＞26＝13+13 ＞2＝延红灯 ＞50＝25+25 ＞34＝17+17 ＞10＝延红灯
1 ＞10＝延红灯 ＞58＝29+29 ＞42＝21+21 ＞18＝9+9 ＞66＝33+33 ＞50＝25+25 ＞26＝13+13
0 ＞34＝17+17 ＞14＝延红灯 ＞66＝33+33 ＞42＝21+21 ＞22＝11+11 ＞6＝延红灯 ＞50＝25+25
i/j 0 1 2 3 4 5 6
表格中＞#＝#+#代表，模式过渡时长‘＞#’是由相位1时长#加相位2时长#组成；
E1-S41，执行优化信号：模式过渡期控制：先运行完成模式过渡期，后运行新周期信号；
实施例2，参见图4，在实施例1的参数、结构与方法基础上更新与增加包括
E2-S11，获取路网参数，包括路网中各个路段长度及其交通用时，如图4中标注3所示；所述交通用时还包括路段的拥堵车队启动用时；其中所述拥堵车队启动用时指拥堵车队从队首车驶离原地到队尾车驶离原地所用时间，等于车队启动系数*拥堵系数*路段长度，其中拥堵系数范围是小于等于1的数，等于1时表示严重拥堵，或包括所述车队启动系数按实验值范围0.10至0.26计算，或取中为0.18，单位：秒/米；
所述信号模式功能还包括疏堵，混合，混合部分疏堵；所述混合指一个流向绿波执行引导而另一个流向绿波执行疏堵；所述混合部分疏堵指一个流向绿波执行引导而另一个流向绿波执行部分路段的疏堵或孤波，特用于下面两维绿波-流向-混合部分疏堵，或指一个流向绿波部分路段执行引导而另部分路段执行疏堵或孤波，特用于下面一维绿波-流向-混合或与部分疏堵；其绿波流向或与绿波功能组合的信号模式，本实施例中包括在表6中信号模式库信号模式栏目下，如模式序号4至17所示，包括两维绿波-流向-疏堵，两维绿波-流向-混合，两维绿波-流向-混合部分疏堵，一维绿波-流向-疏堵，一维绿波-流向-混合；
所述交通模式状况还包括拥堵，混合，混合部分拥堵；所述混合指一流向拥堵另一流向非拥堵即运动的状况；所述混合部分拥堵指一流向交通状况运动另一流向交通状况是部分路段运动部分路段拥堵；所述交通流向或与交通状况组合，本实施例中包括在表6中信号模式库信号模式栏目下，如模式序号4至17所示，包括两维交通-流向-拥堵，两维交通-流向-混合，两维交通-流向-混合含部分拥堵；退化的包括一维交通-流向-拥堵，一维交通-流向-混合，一维交通-流向-混合含部分拥堵；如表6中交通模式列出的参数及其值：
所述信号模式绿波功能与交通模式状况匹配的功能优化信号还包括绿波功能-疏堵匹配交通模式状况-拥堵，绿波功能-混合匹配交通模式状况-混合，绿波功能-混合部分疏堵匹配交通模式状况-混合部分拥堵；绿波模式与交通模式匹配的优化信号还包括两维绿波流向及 功能组合匹配两维交通流向及状况组合的组合优化，本实施例中包括在表6中信号模式库信号模式栏目与匹配交通模式栏目下，如模式序号4至17所示，包括两维绿波-流向-疏堵与两维交通-流向-拥堵匹配，两维绿波-流向-混合与两维交通-流向-混合匹配，两维绿波-流向-混合部分疏堵与两维交通-流向-混合含部分拥堵；一维绿波-流向-疏堵与一维交通-流向-拥堵匹配，一维绿波-流向-混合与一维交通-流向-混合匹配；一维绿波-流向-混合部分疏堵与一维交通-流向-混合含部分拥堵匹配；
表6实施例2信号模式库中信号模式与交通模式匹配表
Figure PCTCN2021000212-appb-000004
注[2]：圆括号中绿波参数涨落将在“信号模式库更新”实施例3使用与说明；
本实施例的现行信号两维绿波-东与北-引导的模式绝对相位差周期余数与周期补数如实施例1中表1所示；
E2-S21，预测交通信息：
S201，选择路网关注点：根据流量与车队分布和以往经验数据，如图4的路网中12个路口小区{(2，1)，(5，3)}经常出现峰值流量与峰值车队队长，确定设置该12个路口小区为关注点路口小区，该区内每个路口在设定位置配置有来车方向交通流量表示达到流量传感器和车队传感器，获取路口东西南北4个方向的来车流向的交通流量与车队首车与尾车即队长信息，以表达所述路网交通信息特征；S203，所述关注点路口交通信息预测计算处理单元，简称为路口单元，使用以小波函数作为激励函数的三层BP神经元网络，即小波神经元网络预测方法，和此前25个信号周期交通数据，预测所述关注点路口每个交通流向的交通流量和车 队长度；其中交通流向特征判定规则是所述路口单元设定流向的交通流量与相对流动的交通流量之差值达到或超过设定值2，较大流量的流向为该关注点的一个关注点主流向，本实施例一个关注点最多可以有两个关注点主流向；交通状况特征判定规则是所述路口单元设定流向的车队队长与其路段长度之差值达到或小于设定值2，较大流量的流向为该关注点的一个关注点主流向状况，本实施例一个关注点最多可以有两个关注点主流向状况；
根据所配置的关注点，传感器及其设定交通数据，和设定预测方法预测得到：图4中12个关注点的预测交通特征数据，如表7所示，各关注点的东流向与西流向的流量及其差，各关注点的东流向与西流向的车队长度及其差，及它们达到或超过主流向设定值的流向；这些预测的关注点主流向和流量状况信息分别送给分析层中绿波优化的“单流向-东/西/南/北-优化信号初判单元”“双流向-东与南/南与西/西与北/北与东-优化信号初判单元”；
表7小区12个关注点路口预测交通信息特征数据
Figure PCTCN2021000212-appb-000005
注：表中F代表流量特征，Q代表车队特征，：#-#＝#@主流向//#-#＝#@主流向代表交通信息西流向(即路口东方位)数据#减去东流向(即路口西方位)数据#得到之差#，和@主流向/状况，双后斜线//引出同理的北流向数据#减去南流向数据#得到之差#，和@主流向/状况，如，路口(4，2)数据，“F：8-1＝7@西//1-7＝-6@南”和“Q：16-1＝15@西/拥堵//0-0＝0@”表示预测该路口东与西流量之差7得到主流向为西，北与南流量之差-6得到主流向为南，东与西车队长度之差15得到主流向为西，北与南车队长度之差0得到主流向为无；
E2-S31，决定优化信号中1)根据分析预测的交通信息获得交通模式，匹配信号模式，形成优化信号的初步判断包括：1.1)，绿波优化信号初步判断步骤包括：1.1.1)，确定路网主交通流向：计算各同主流向的关注点该主流向流量达到流量设定值或/或与流量差设定值的流量总值，其中值大的主流向设定为小区交通主流向即交通模式流向，包括两维交通-流向，一维交通-流向，零维交通-均向，计算得到表8：
表8 12个关注点小区计算分析交通模式信息
Figure PCTCN2021000212-appb-000006
Figure PCTCN2021000212-appb-000007
表中“(i，j)@流向#”表示横纵坐标i，j的路口(i，j)主流向流量#，如(2，1)@东5代表路口(2，1)主流向东的流量5，“+”代表相加；东流向流量之和是19，南流向流量之和是58，西流向流量之和是64，北流向流量之和是24，较大的两个总流量是南流向58和西流向64，得到交通模式流向是两维交通-南与西；
1.1.2)确定路网主交通状况及部分拥堵分布：计算各同流向的关注点车队长度达到车队长度设定值或/或与其路段长度之差达到或小于路段拥堵设定值的路段数，简称流向拥堵路段数，当交叉的两个流向拥有较多该流向拥堵路段数，且都达到或超过该流向设定值即流向拥堵设定值，本实施例设定该设定值为该流向路段总数的90％，则该两个流向被设定为该两维交通拥堵主流向，当仅该一个流向拥有较多流向拥堵路段数，且都达到或超过该流向拥堵设定值，则该流向被设定为该一维交通拥堵主流向，当没有流向拥有流向拥堵路段数都达到或超过该流向拥堵设定值，则该流向被设定为该交通部分拥堵流向，当没有任何流向拥有流向拥堵路段数，则被设定为该两维交通-流向-运动，交通特征组合包括两维交通-流向-运动，或/或与零维交通-均向-运动，或/或与一维交通-流向-运动，或/或与两维交通-流向-拥堵，或/或与一维交通-流向-拥堵，或/或与两维交通-流向-混合，或/或与一维交通-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵，或/或与一维交通-流向-运动混合含部分拥堵；本实施例根据计算和确定的设定值，得到如表7分析计算交通模式状况部分，所述交通模式流向是两维交通-南与西-运动混合含部分拥堵，两个路段拥堵：横路段(2，3)和(2，4)；
1.1.3)决定匹配的绿波优化参数及其分布：两维交通-流向-运动匹配两维绿波-流向-引导，或/或与零维交通-均向-运动匹配零维驻波-均向-均引，或/或与一维交通-流向-运动匹配一维绿波-流向-引导，或/或与两维交通-流向-拥堵匹配两维绿波-流向-疏堵，或/或与一维交通-流向-拥堵匹配一维绿波-流向-疏堵，或/或与两维交通-流向-混合匹配两维绿波-流向-混合，或/或与一维交通-流向-混合匹配一维绿波-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵匹配两维绿波-流向-引导混合部分疏堵，或/或与一维交通-流向-运动混合含部分拥堵匹配一维绿波-流向-引导混合部分疏堵；或与绿波-流向-引导混合部分疏堵包括部分疏堵、孤波疏堵，所述部分疏堵指将拥堵的小区路口部分配置为相应的小区疏堵信号模式，也包括所述两维绿波-流向-疏堵和一维绿波-流向-疏堵，所述孤波疏堵指使用所述孤波技术将拥堵路段的车队疏解开，也包括余时孤波，指令孤波；本实施例，对于两维交通-南与西-运动混合含部分拥堵，两个路段拥堵：横路段(3，2)和(4，2)，匹配两维绿波-南与西-引导混合部分疏堵/孤波疏堵；
2)或与所述优先规则还包括两维绿波-流向-疏堵初步判断优先于两维绿波-流向-混合初步判断，两维绿波-流向-混合初步判断优先于一维绿波-流向-疏堵初步判断，一维绿波-流向-疏堵初步判断优先于两维绿波-流向-引导混合部分疏堵初步判断，两维绿波-流向-引导混合部分疏堵初步判断优先于两维绿波-流向-对引初步判断；决定的优化信号：两维绿波-南与西-引导混合部分疏堵/孤波；
3)根据决定的优化信号和现行信号制作相应的模式过渡期：决定的优化信号；
两维绿波-南与西-引导混合部分疏堵/孤波，和现行信号，两维绿波-北与东-引导，制作两者相应的模式过渡期，3.0)各路口保持之前现行信号周期68秒及相位配时；3.1)计算各路口的所述两个信号模式的过渡时长：3.1.1)优化信号的两维绿波-南与西-引导混合部分疏堵/孤波，？？？其它模式参考快速模式方法，其模式相位差基点选在两维交通流向最上游交汇路口即路口(6，4)如图2中右上角虚线三角形标注的路口，将绿波流向西选做模式主流向，穿过模式相位差基点路口(6，4)的绿波流向南就成为模式副流向，该南北通道就成为所有模式主流向通道的通道相位差基点，？？？；(1)先按实施例1中对全路网各路口进行两维绿波-南与西-引导模式相位差配时，过程如表2至表4，结果如表5；(2)对路口(3，2)西拥堵车队16，路口(4，2)西拥堵车队12进行孤波相位配时--孤波临时配时表；两个拥堵路段所在通道为第2横通道，其在指令模式两维绿波-南与西-引导下的各路口相位配时如表9，
表9第2横通道的各路口原相位配时表
Figure PCTCN2021000212-appb-000008
注：表中“流向#/#/#”代表表格坐标代表的路口的相位配时，如“坐标(0，1)东西22/12/”表示路口(0，1)的东流向和西流向的直行相位22秒，左转相位12，无右转相位；“相位差”是该路口的模式相位差周期余数；
---计算拥堵路口(3，2)的拥堵车队通过路口所需时间(秒)t x，等于拥堵车队长度(标准小客车数)q乘以路口通过速率(秒数/标准小客车)v x，v x的经验值2至2.6，取v x＝2，t x＝16*2＝32秒；让路口(3，2)西流向原相位时间20秒借用其它相位时间10，得到临时相位配时32秒，如表9对应路口的(3，2)表格中所示；
---计算该拥堵车队通过下游路径各路口需要占用的时间与临时配时，由两段组成计算，
1.拥堵路口(3，2)到西流向最下游装配有车队传感器的路口(2，2)，根据余时孤波算法配时：计算下游路口西流向车队和拥堵路口西流向车队通过路口用时之和，路口(2，2)西流向车队8+路口(2，2)西流向车队16＝24，该通过路口用时之和tq＝24*2＝48秒；
计算拥堵车队达到路口(2，2)的时间损失trq，当trq＞0时，该trq是时间损失，将trq计入所述通过路口用时，当trq＜＝0时，表明无时间损失：trq＝d/v0-(1/v0+a)*q，其中，d是相邻路口间路段长度-米，v0是该路段规定限制时速下的设计绿波时速-米/秒，q是路口(2，2)西流向车辆排队长度-米，a是车队启动系数，其估定范围0.10至0.26，取中0.18，单位：秒/米，该取值可以动态调整，a*q＝tq是车队启动用时＝0.18*8*7＝10.08＝10秒，(d-q)/v0＝(150-8*7)/12.5＝7.5，trq＝7.5-10.08＜0；注：算式中数字7是标准小客车长与车间距；
让路口(2，2)西流向原相位时间20秒借用其它相位时间28，得到临时相位配时48秒，如表9对应路口的(2，2)表格中所示；
2.再向流向没有装配车队传感器的路口(1，2)直至最下游口(0，2)；配时也相应分为余时孤波和指令孤波，根据指令孤波算法配时：强制该下游无传感器各路口使用相邻有车队传感器路口的临时相位配时，如如表9对应路口的(1，2)和(0，2)表格中所示，得到指令孤波部分 的临时相位配时表；与上游的余时孤波部分的合成得到此拥堵车队路径临时相位配时表，如表10所示，
表10第2横通道的各路口孤波临时配时表
Figure PCTCN2021000212-appb-000009
由于路口(4，2)西流向拥堵车队略小于之前的路口(3，2)西流向拥堵车队，上述疏解路口(3，2)西流向拥堵车队的孤波临时配时表可以适用于疏解路口(4，2)西流向拥堵车队，因此上述孤波临时配时表执行2个周期，将路口(4，2)的拥堵车队疏解开；
E2-S42，执行优化信号：模式过渡期控制：先运行完成模式过渡期，后运行新周期信号；执行孤波临时配时表。
实施例3：S101，增加新信号模式绿波参数-相位差涨落；如实施例2中的路网，信号与关注点及交通数据配置与设定；分别在其S12，S22，S32中增加如下，
E3-S101-S1，在所述信号模式库绿波参数中创建一个新参数，名为相位差涨落，指路口设定流向相位差变化；其值是路队相位差及其变化trq，trq＝d/v0-(1/v0+a)*Q，Δtrq＝-(1/v0+a)*ΔQ，(《交通信号泛绿波控制方法》中国发明，201710224791.x和《交通信号泛弦控制方法及其系统》中国发明，201710897777.6)；如表6中第15行；
所述交通模式创建新参数路口设定流向车队长度Q及其变化量ΔQ，及它们的设定值；
所述信号模式与交通模式匹配增加绿波-路口流向-相位差涨落与交通-路口流向-相位车队变化；
E3-S101-S2，预测交通信息，所述关注点路口单元增加新处理功能：发现指定流向车队传感系统获取的车队长度达到设定值和其变化也达到设定值的特征数据及该路口所在指定流向坐标，简称列路口行流向车队涨落，或行路口列流向车队涨落，送分析层；本实施例中，所述列路口行流向车队涨落的流向指行横向，东或西，一个方向，所述行路口列流向车队涨落的流向指列纵向，南或北，一个方向；
E3-S101-S3，决定优化信号，1)分析层增加新优化信号初步判断单元：4个分别分析本实施例中4列装配车队传感系统路口的列涨落初判单元，3个分别分析本实施例中3行装配车队传感系统路口的行涨落初判单元；所述列涨落初判单元分析指定同列所有路口的行路口列流向车队涨落的路口数是否达到设定值，以确定交通模式-列路口行流向-车队变化，匹配绿波-列路口行流向-相位差涨落初步判断；所述行涨落初判单元分析指定同行所有路口的列路口行流向车队涨落的的行路口数是否达到设定值，以确定交通模式-列路口流向-车队变化，匹配绿波-列路口流向-相位差涨落初步判断；2)优先规则增加涨落初步判断仅优先于零维驻波；3)制作模式过渡期，增加重新计算配置现行信号路口模式绝对相位差及其周期余数，装配车队传感系统的路口流向使用相应路队相位差trq作为该路口相对相位差；
E3-S101-S4，执行优化信号：增加相位差涨落控制：先运行完成现行相位差周期，后运行相位差涨落即新相位差周期信号；
Figure PCTCN2021000212-appb-000010

Claims (16)

  1. 一种交通信号智能控制架构方法，其特征在于包括步骤：
    S1，获取路网参数；创建若干路网交通信号模式，简称信号模式，组成信号模式库，每种信号模式提供该模式相应的路网交通特征的控制服务，与路网交通特征匹配的构成优化信号；获取或配置路网现在运行的信号模式及其参数，简称现行信号及其参数；
    所述路网参数包括路网中各个路段长度及其交通用时；该交通用时包括行车用时或/或与拥堵车队启动用时；该行车用时指车辆以设定车速行驶完成路段所用时间，等于该路段长度除以设定行车速度，或包括减去设定行车速度的刹车时间；该拥堵车队启动用时指拥堵车队从队首车驶离原地到队尾车驶离原地所用时间，等于车队启动系数*拥堵系数*路段长度，该拥堵系数等于车队长度与路段车队之比，等于1时表示严重拥堵，或包括该车队启动系数按实验值范围0.10至0.26计算，或取中为0.18，单位：秒/米；注：本文中“或/或与表述，”和“或与表述，”仅仅指紧随该“或/或与”和“或与”之后的“表述”，不涉及逗号后续列出的“表述”，例如“表述1，或/或与表述2，表述3”中“表述1”和“表述3”是并存关系，“表述2”是与其它表述是“或”存关系，或，“或与”存关系；
    所述信号模式包括任何设定路口间相位间绿灯传播流向的信号，称为绿波；该流向作为信号模式绿波的特征参数称为绿波流向，模式记为绿波-流向；该设定绿波流向由各路口间运行的比率信号配置的相位时间差即相位差的大小排序所决定，从较小相位差路口流向较大相位差路口；两个相邻路口间的相位时间差是相对相位差，等于该相邻路口间路段设定的交通用时；各路口间运行的比率信号间没有相位差的信号模式是驻波，不传播绿灯变化，即比率模式，绿波退化模式，提供各流向均等交通特征的控制服务功能；所述比率信号指路口根据设定周期和设定比率配置各相位时长；该绿波特征参数或还包括功能称为绿波功能，该功能包括引导或均衡，或/或与疏堵，或/或与混合，或/或与对流引导即对引，其中，引导指当绿波流向与所控交通流向相同，各路段交通用时采用设定行车用时计算相对相位差；均衡指绿灯信号无流向，各路口0相位差同步；疏堵指绿波流向与所控交通流向相反，各路段交通用时采用设定拥堵车队启动用时计算相对相位差；混合指当绿波流向与所控交通流向既包括相同部分也包括相反部分，相同流向部分的路段采用行车用时而相反流向部分的路段采用拥堵车队启动用时；对引指对相对的两个方向的交通实行引导，是引导的一种；绿波流向或与绿波功能组合为绿波优化信号，其模式分别记作绿波-流向-引导，绿波-流向-疏堵，绿波-流向-混合，驻波-均向-均衡；该信号模式参数还含周期，各相位分配时长即相位配时；该相位包括代表交叉路口控制方向的方向相位，或与方向相位中控制左/右转向的分流相位；
    所述路网交通特征，简称交通模式，参数包括流向，或/或与状况，或/或与流量，或/或与路口各相位分流量；其中，状况包括运动即非拥堵，或/或与拥堵，或/或与混合；拥堵指路段长度减去车队长度达到或小于设定值，或拥堵系数达到或超过设定值，的车队状况，该设定值简称为路段拥堵设定值；混合指部分路段拥堵部分路段非拥堵；所述交通模式流向或与交通模式状况组合的交通模式，记作交通模式-流向-运动，或与交通模式-流向-拥堵，或与交通模式-流向-混合，或与交通模式-均向-运动；
    所述信号模式与路网交通特征的匹配指与交通模式各对应参数的匹配，包括绿波流向与交通模式流向一致匹配的绿波流向优化信号，或与绿波功能与交通模式状况对应匹配组合的绿波优化信号，包括绿波-流向-引导匹配交通模式-流向-运动，驻波-均向-均衡匹配交通模式-均向-运动，或与绿波-流向-疏堵匹配交通模式-流向-拥堵，或与绿波-流向-混合匹配交通模式-流向-混合；或/或与信号模式周期与交通模式流量对应匹配的周期优化信号；或/或与信号模式相位配时与交通模式相位分流流量对应匹配的相位配时优化信号；其中，“流向一致匹配”指流向相同，“对应匹配”指根据该匹配涉及该两个模式参数组数值之间设定的系列数值组中交通模式参数数值所对应信号模式参数数值，可选择的数值组包括交通模式参数状况与信号模式参数功能数值组，交通模式参数流量与信号模式参数周期时长数值组，交通模式参数相位分流流量与信号模式参数相位配时数值组，交通模式参数车队队长相关值与信号模式设定对应参数数值组；
    S2，预测交通信息：根据获取路网的交通信息，包括实测或/或与过去的若干设定时段，m＞＝1个周期c，的各流向车流量特征或/或与车队特征，预测未来设定若干时段，n＞＝1个周期c，的路网交通信息；当不设定所述预测未来若干周期时，下周期为所预测时段；
    S3，决定优化信号：1)根据接到指令信号模式或分析所述预测交通信息得到的交通模式，从所述信号模式库中找到与交通模式匹配的信号模式，形成所述优化信号的初步判断，包括1.1)绿波优化信号初步判断，1.2)或与周期优化信号初步判断或与相位配时优化信号初步判断，所述指令信号模式指无需与交通信息匹配的强制执行所指令的信号模式；2)或与根据优先规则从所述初步判断中决定出优化信号；3)根据决定的优化信号和现行信号制作相应的模式过渡期；
    S4，执行优化信号：模式过渡期控制：先运行完成模式过渡期，后运行新周期信号。
  2. 根据权利要求1所述交通信号智能控制架构方法，其特征在于，
    S11，所述信号模式库中的信号模式包括叠加信号；
    所述路网参数包括路网拓扑形式：拓扑平行四边形，或包括虚拟路口、虚拟路段；
    所述叠加信号指设定时段(含若干个n＞＝1信号周期c)内运行2个及以上交叉流向的绿波信号，也就是任何基于交叉的两个流向的绿波，即两维绿波，的信号，及其任何降维形式；当所述两维绿波中的一个流向绿波相位差被配置为0时，就降维成一维绿波；
    所述绿波参数流向还包括交叉的两个流向，该绿波称为两维绿波，或记为绿波-双流向或两维绿波-流向，其降维形式包括一维绿波-流向，及0维驻波-均向；所述绿波优化信号，包括两维绿波-流向-引导，一维绿波-流向-引导，0维驻波-均向-均衡，两维绿波-流向-疏堵，一维绿波-流向-疏堵，两维绿波-流向-混合，一维绿波-流向-混合，或与两维绿波-流向-对引，一维绿波-流向-对引：
    所述交通模式参数流向还包括交叉的两个主流向，该交通模式称为两维交通，或记为交通-双流向或两维交通-流向，其降维形式包括一个主流向的为一维交通-流向，及0维交通-均向；所述交通模式，包括两维交通-流向-运动，一维交通-流向-运动，0维交通-均向-运 动，两维交通-流向-拥堵，一维交通-流向-拥堵，两维交通-流向-混合，一维交通-流向-混合，或与两维交通-流向-对流，一维交通-流向-对流；
    所述绿波优化信号还包括所述两维绿波与所述两维交通匹配的优化信号，即两维绿波优化信号，可选择的包括两维绿波-流向-引导与两维交通-流向-运动匹配，一维绿波-流向-引导与一维交通-流向-运动匹配，0维驻波-均向-均衡与0维交通-均向-运动匹配，或与两维绿波-流向-疏堵与两维交通-流向-拥堵匹配，或与一维绿波-流向-疏堵与一维交通-流向-拥堵匹配，或与两维绿波-流向-混合与两维交通-流向-混合匹配，或与一维绿波-流向-混合与一维交通-流向-混合匹配，或与两维绿波-流向-对引与两维交通-流向-对流匹配，或与一维绿波-流向-对引与一维交通-流向-对流匹配；
    S21，所述预测交通信息中的交叉的两个及以上个流向交通信息，或与其状况。
  3. 根据权利要求2所述交通信号智能控制架构方法，其特征在于，
    S12，所述叠加态信号还包括，设定交通特征条件下让路口一个相位的设定车队获得若干个连续路口绿灯信号，简称孤波，叠加在两维绿波中；
    所述信号模式参数还包括所述若干个连续路口的路径临时配时表，称为孤波临时配时表；
    所述设定交通特征条件包括指令，让指定路口流向相位的设定车队强制占用前方路口其它相位车队的绿灯时间，通过，简称强制孤波；或/或与让指定路口流向相位的设定车队占用前方路口其它相位车队的绿灯剩余时间，简称余时，通过，简称余时孤波；两者统称孤波交通；
    所述信号模式匹配交通模式的优化信号还包括孤波-流向-孤波临时配时表与设定车队-流向-路径路口表，孤波临时配时表包括强制孤波临时配时表或/或与余时孤波临时配时表；
    S22，所述预测交通信息还包括，实时实测与预测的所述设定交通特征条件包括指令孤波：包括强制孤波，或/或与余时孤波；强制孤波临时配时表的配置计算：强制指定路口流向相位含方向相位或分流相位的设定车队占用前方路口其它相位车队的相位绿灯时间，依此计算建立所述强制孤波临时配时表；余时孤波临时配时表的配置计算：指定路口流向相位含方向相位或分流相位设定车队占用前方路口相关相位车队提供本相位后剩余的绿灯时间通过，依此计算建立所述余时孤波临时配时表；计算所述余时的方法简称A-A算法，包括使用各设定路口流向交通相位含方向相位分流相位队尾q特殊信息，长车队；
    S32，所述决定优化信号1)预测交通模式与信号模式的匹配还包括，接到指令孤波，确定路口相位设定车队，将孤波叠加于现行信号模式上的优化信号初步判断，包括用孤波预测算法建立获得相关若干个连续路口绿灯信号的孤波临时配时表；所述2)或与所述优先规则还包括强制孤波初步判断优先于余时孤波初步判断，孤波初步判断优先于其它初步判断；
    S42，所述执行优化信号还包括：执行孤波临时配时表。
  4. 根据权利要求2所述交通信号智能控制架构方法，其特征在于，
    S13，所述叠加态信号还包括，设定交通信息特征条件下让距离路口一个相位最近的行车，即相位首车，由红灯获得绿灯信号，称微分绿波，叠加在两维绿波中；所述行车指以规定速度正常行驶的车辆；
    所述信号模式参数还包括相位间的绿灯与红灯间的最小安全切换通过时间，称为微分绿波时间，或相变量子用时Δt Th0，及所涉及两相位；所述相位最小安全切换通过时间指能够让相位首车在其控制相位信号由绿灯切换成红灯时，以正常刹车安全停于前方停车线，同时让其它相位首车在其控制相位信号由红灯切换成绿灯后安全通过前方路口的最小时间；以该相位最小安全切换通过时间所对应计算出的相位首车与前方控制相位的距离就是相位最小安全距离；
    所述设定交通信息特征条件包括各设定路口流向交通相位红灯时的相位首车q0处于或小于该相位设定的最小安全距离，而该路口比率信号绿灯的相位首车q0处于其相位的最小安全距离外或与相应的行人相位也无行人，即比率信号绿灯相位无车或与无行人，合称为红灯交通相位首车在可微分位置的微分交通，也称微分绿波条件；或与在下列设定顺序规则下延续微分绿波条件，当“正在占用比率信号绿灯时间的”绿灯相位继续检测到有车时，或与该同流向其它分流相位检测到有车时，或其它流向检测到有车时；
    所述信号模式与交通模式匹配的优化信号还包括微分绿波时间及其两相位与微分绿波条件及其两相位匹配；
    S23，所述预测交通信息还包括，实时实测的所述设定交通特征条件的微分绿波条件；
    S33，所述决定优化信号所述1)预测交通模式与信号模式的匹配还包括，根据实测到微分绿波条件或接到指令微分绿波信号模式，为在现行信号模式下满足所述微分绿波条件的红灯相位首车，匹配叠加于现行信号模式上微分时间的优化信号初步判断，即让在现行信号模式下满足上述“微分绿波条件”的红灯相位首车获得占用“比率信号绿灯相位无车或与无行人”相位的一个所述相变量子用时或称所述微分时间；所述2)或与优先规则还包括微分绿波初步判断优先于其它初步判断；
    S43，执行优化信号还包括：立即将绿灯切换给或保持给获得微分时间的相位。
  5. 根据权利要求1所述交通信号智能控制架构方法，其特征在于包括
    S101，所述信号模式库包括增加模式或模式参数的操作，设定一个新信号模式提供设定交通模式的服务或服务于设定的交通模式，将这一信号模式参数与其服务的交通模式匹配，及其优化信号初步判断及其优先级、实现操作，分别配置到S1中信号模式库中，S2中预测交通信息中，S3中决定优化信号的初步判定优先规则中，S4执行优化信号的操作中。
  6. 根据权利要求1所述交通信号智能控制架构方法，其S2预测交通信息特征特征在于，
    S201，所述实测或预测路网交通信息及其特征包括指通过对路网中设定若干装配交通传感装置的关注点，路段中的设定位置或/或与交叉路口，用来获取所述关注点的设定流向的交通流量特征数据或/或与车队特征数据，以分析表达所述路网交通信息特征。
  7. 根据权利要求6所述交通信号智能控制架构方法，其S2预测交通信息特征特征在于，
    S202，所述关注点设定流向的交通流量特征数据包括相应的若干交通流量设定值或与其变化设定值，或/或与该设定流向相对流向的流量之差设定值，达到所述流量设定值或流量之差设定值的流向作为该关注点主流向；或与所述该关注点设定流向的车队特征数据包括相应 的若干车队长度设定值或与其变化设定值；或/或与该设定流向车队长度与其所在路段长度之差设定值，达到该之差设定值的状况为该关注点的设定流向拥堵路段。
  8. 根据权利要求6所述交通信号智能控制架构方法，其S2获取预测交通信息特征特征在于，
    S203，所述关注点配置用于该点交通信息预测的计算处理单元，简称关注点单元，每个关注点单元用设定预测方法对应的关注点设定流向的交通信息做预测；若干个所述关注点单元组成预测层；所述预测方法指任何根据已有数据预测未来数据的方法，包括直接使用实时检测数据作为预测未来数据的重复式预测法，经验方法，均值法，最大值法，统计优化方法，含神经元网络方法和专家系统方法的人工智能方法。
  9. 根据权利要求8所述交通信号智能控制架构方法，其S2获取预测交通信息特征特征在于，
    S204，所述关注点路口用于自身设定流向的的交通信息预测计算处理单元，简称路口单元，每个路口单元或与还负责预测流向相位分流相位交通信息特征，或涉及包括流向上游临近路段的车源或与流向上游路口单元交通信息特征及其设定值。
  10. 根据权利要求6所述交通信号智能控制架构方法，其S2获取预测交通信息特征特征在于
    S205，所述关注点的数量、分布、路段或路口的类型是通过所述设定预测方法预测的流量流向分布决定。
  11. 根据权利要求1所述交通信号智能控制架构方法，其S3决定优化信号特征在于包括，
    S301，所述绿波优化信号初步判断包括绿波流向优化和绿波功能优化，步骤包括1)确定交通模式流向：计算各同流向的关注点主流向流量达到流量设定值或/或与流量差设定值即关注点主流向的流量之和，该之和值大的流向设定为路网交通主流向，即交通模式流向，包括交叉的两个路网交通主流向的为交通模式双流向，记为两维交通-流向，一个路网交通主流向的为交通模式单流向，记为一维交通-流向，没有路网交通主流向的为交通模式流向，记为零维交通-均向；2)或与确定交通模式状况：计算各同流向的关注点车队长度达到车队长度设定值或/或与其路段长度之差达到或小于路段拥堵设定值的路段数，简称流向拥堵路段数，之和，该之和值大的流向设定为路网交通拥堵主流向，即交通模式拥堵主流向，当交叉的两个流向都拥有较多该流向拥堵路段数，且都达到或超过该流向设定值即流向拥堵设定值，则该两个流向是两个路网交通拥堵主流向，记为交通-双流向-拥堵或两维交通-流向-拥堵，当仅该一个流向拥有较多流向拥堵路段数，且达到或超过该流向拥堵设定值，而另一个流向没有流向拥堵路段数，则该流向是一个路网交通拥堵主流向，记为一维交通-流向-拥堵，当没有流向拥有流向拥堵路段数达到或超过该流向拥堵设定值，则该流向为该部分拥堵流向，包括小区域拥堵流动或几个路段拥堵流向；与交通模式流向组合特征包括两维交通-流向-运动，或/或与零维交通-均向-运动，或/或与一维交通-流向-运动，或/或与两维交通-流向-拥堵，或/或与一维交通-流向-拥堵，或/或与两维交通-流向-混合，或/或与一维交通-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵，或/或与一维交通-流向-运动混合含部分拥堵；3)决定匹配的绿波优化参数及其分布：两维交通-流向-运动匹配两维绿波-流向-引导，或/或与零维交通-均向-运动匹配零维驻波-均向-均引，或/或与一维交通-流向-运动匹配一维绿波-流向-引导，或/或与两维交通-流向-拥堵匹配两维绿波-流向-疏堵，或/或与一维交通- 流向-拥堵匹配一维绿波-流向-疏堵，或/或与两维交通-流向-混合匹配两维绿波-流向-混合，或/或与一维交通-流向-混合匹配一维绿波-流向-混合，或/或与两维交通-流向-运动混合含部分拥堵匹配两维绿波-流向-引导混合部分疏堵，或/或与一维交通-流向-运动混合含部分拥堵匹配一维绿波-流向-引导混合部分疏堵；或与绿波-流向-引导混合部分疏堵包括局部疏堵、孤波疏堵，所述局部疏堵指将局部拥堵作为小区路网配置为相应的路网疏堵信号模式，也包括所述两维绿波-流向-疏堵和一维绿波-流向-疏堵，所述孤波疏堵指使用所述孤波技术将拥堵路段的车队疏解开，也包括余时孤波，指令孤波；
    或/或与直接用所述预测方法进行绿波优化信号预测初步判断：其所预测目标的数据包括指定交通特征数据，用所述预测方法调整优化控制该交通的信号参数，根据已有指定交通特征数据和绿波参数数据组来优化出满足预测目标的交通特征数据的信号参数的优化信号初步判断，称为直接优化信号预测方法。
  12. 根据权利要求1所述交通信号智能控制架构方法，其S3决定优化信号特征在于包括，
    S302，周期优化信号初步判断，或与相位配时优化初步判断，还包括使用任何路口信号配时算法进行预测优化，韦伯斯特配时法，冲突点法，估算法，临界车道法，和周期相位配时优化信号相关的预测方法，称为周期相位配时优化信号预测方法。
  13. 根据权利要求1所述交通信号智能控制架构方法，其S3决定优化信号特征在于包括，
    S303，所述预测方法，还包括直接优化信号预测方法和周期相位配时优化信号预测方法，通过以车辆等待时间最小化或/或与车队等待长度最小化为目标的优化信号学习方法，发现路网的交通信息关注点获取的交通信息特征数据与相对应的优化信号参数数据之间的新变化关系，将该新变化关系创建为新优化信号，加进信号模式库。
  14. 根据权利要求1所述交通信号智能控制架构方法，其S3决定优化信号特征在于包括，
    S304，所述优先规则根据包括计算预测/估计各优化信号初步判断将引起的行车等待时间的减少/车队车队减少/流量增加的数据排序决定各优化信号初步判断的优先排序。
  15. 根据权利要求2所述交通信号智能控制架构方法，其S3决定优化信号特征在于包括，
    S305，所述优先规则根据所述估计数据的大小排序，排在前者优先于后者：包括指令信号模式，或与两维绿波-流向-疏堵初步判断，或与两维绿波-流向-混合初步判断，或与一维绿波-流向-疏堵初步判断，或与一维绿波-流向-混合初步判断，或与两维绿波-流向-对引初步判断，或与一维绿波-流向-对引初步判断，两维绿波-流向-引导初步判断，一维绿波-流向-引导初步判断，驻波-均向-均衡初步判断。
  16. 根据权利要求3所述交通信号智能控制架构方法，其强制孤波临时相位配时特征包括
    F201，根据预测获取的孤波车队长度，计算通过本地路口所需的时间，简称强制孤波用时；用本地相位绿灯时间减去这个强制孤波用时，得到小于0的数，则将其它相位相应量的时间借过来使用；
    根据强制孤波车队长度和预测达到前方路口的时间，预测前方路口相关相位的信号灯色，及前方路口同关相位绿灯时间中减去一个设定时间后的剩余绿灯时间，再减去这个强制孤波用时，得到剩余绿灯时间，小于0，则该绝对值就是预测借用其它相位相应量的时间；其中设定时间是假设该相位车队通过用时；
    如上预测计算设定若干个连续路口；制定所述若干路口的临时配时表。
PCT/CN2021/000212 2021-10-22 2021-10-22 交通信号智能控制架构方法 WO2023065057A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/000212 WO2023065057A1 (zh) 2021-10-22 2021-10-22 交通信号智能控制架构方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/000212 WO2023065057A1 (zh) 2021-10-22 2021-10-22 交通信号智能控制架构方法

Publications (1)

Publication Number Publication Date
WO2023065057A1 true WO2023065057A1 (zh) 2023-04-27

Family

ID=86057791

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/000212 WO2023065057A1 (zh) 2021-10-22 2021-10-22 交通信号智能控制架构方法

Country Status (1)

Country Link
WO (1) WO2023065057A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116612648A (zh) * 2023-05-15 2023-08-18 深圳市显科科技有限公司 基于情报板的道路阶段性动态交通疏堵提示方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107154156A (zh) * 2017-07-13 2017-09-12 李志远 基于自组网数据通信的智能交通控制技术
CN107665578A (zh) * 2016-07-27 2018-02-06 上海宝康电子控制工程有限公司 基于大数据研判的交通综合管控系统及方法
CN110570651A (zh) * 2019-07-15 2019-12-13 浙江工业大学 一种基于深度学习的路网交通态势预测方法及系统
JP2020013311A (ja) * 2018-07-18 2020-01-23 パナソニックＩｐマネジメント株式会社 交通信号制御装置、交通信号制御システムおよび交通信号制御方法
US20200043330A1 (en) * 2017-04-07 2020-02-06 Weiping Meng Traffic Signal Pan-Greenwave Control Method
CN111696370A (zh) * 2020-06-16 2020-09-22 西安电子科技大学 基于启发式深度q网络的交通灯控制方法
CN111710167A (zh) * 2020-08-24 2020-09-25 深圳市城市交通规划设计研究中心股份有限公司 一种基于在线交通仿真的单点优化控制方法及装置
AU2020102415A4 (en) * 2020-09-24 2020-11-12 East China Normal University An adaptive traffic signal control algorithm and system based on speed monitoring

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107665578A (zh) * 2016-07-27 2018-02-06 上海宝康电子控制工程有限公司 基于大数据研判的交通综合管控系统及方法
US20200043330A1 (en) * 2017-04-07 2020-02-06 Weiping Meng Traffic Signal Pan-Greenwave Control Method
CN107154156A (zh) * 2017-07-13 2017-09-12 李志远 基于自组网数据通信的智能交通控制技术
JP2020013311A (ja) * 2018-07-18 2020-01-23 パナソニックＩｐマネジメント株式会社 交通信号制御装置、交通信号制御システムおよび交通信号制御方法
CN110570651A (zh) * 2019-07-15 2019-12-13 浙江工业大学 一种基于深度学习的路网交通态势预测方法及系统
CN111696370A (zh) * 2020-06-16 2020-09-22 西安电子科技大学 基于启发式深度q网络的交通灯控制方法
CN111710167A (zh) * 2020-08-24 2020-09-25 深圳市城市交通规划设计研究中心股份有限公司 一种基于在线交通仿真的单点优化控制方法及装置
AU2020102415A4 (en) * 2020-09-24 2020-11-12 East China Normal University An adaptive traffic signal control algorithm and system based on speed monitoring

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116612648A (zh) * 2023-05-15 2023-08-18 深圳市显科科技有限公司 基于情报板的道路阶段性动态交通疏堵提示方法及装置
CN116612648B (zh) * 2023-05-15 2024-03-26 深圳市显科科技有限公司 基于情报板的道路阶段性动态交通疏堵提示方法及装置

Similar Documents

Publication Publication Date Title
CN110136455B (zh) 一种交通信号灯配时方法
CN108597235B (zh) 基于交通视频数据的交叉口信号参数优化及效果评估方法
Ma et al. Development and evaluation of a coordinated and conditional bus priority approach
CN111145544B (zh) 一种基于拥堵蔓延消散模型的行程时间及路径预测方法
CN104615897B (zh) 基于低频gps数据的路段行程时间估算方法
CN109872538B (zh) 基于mfd的饱和交叉口群多层边界迭代学习控制方法和装置
CN112216130B (zh) 一种车路协同环境下的应急车辆诱导方法
CN109993981A (zh) 基于全息检测的交通信号自协调控制方法
CN109902864A (zh) 一种考虑网络配载均衡的施工区交通组织方案设计方法
CN113112823B (zh) 一种基于mpc的城市路网交通信号控制方法
Xu et al. A bi-level model to resolve conflicting transit priority requests at urban arterials
CN110570672A (zh) 一种基于图神经网络的区域交通信号灯控制方法
Niittymäki General fuzzy rule base for isolated traffic signal control‐rule formulation
WO2023065057A1 (zh) 交通信号智能控制架构方法
Xie et al. Real-time traffic control for sustainable urban living
CN110164148B (zh) 一种城市路口交通灯智能配时控制方法及控制系统
Li et al. Traffic adaptive control for oversaturated isolated intersections: Model development and simulation testing
Li et al. Regional coordinated bus priority signal control considering pedestrian and vehicle delays at urban intersections
CN113408189B (zh) 基于可变元胞的城市多点循环式紧急疏散与仿真推演方法
JPH07129882A (ja) 道路交通信号制御適正化装置、およびそれを用いた道路交通流管制システム
CN113470390B (zh) 短连线交叉口边缘节点融合的多相位动态协调控制方法
Liu et al. Trade-offs between bus and private vehicle delays at signalized intersections: Case study of a multiobjective model
Wang et al. Coordinated control model for oversaturated arterial intersections
CN115063988B (zh) 一种应急车辆优先的跨交通子区信号协同控制方法
CN114613125B (zh) 一种快速路多匝道分层协同控制方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21960820

Country of ref document: EP

Kind code of ref document: A1