JP2003132490A - Autonomous distributed signal control system, signal control method and program for signal control - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an autonomous distributed signal control system for performing optimal control over an entire control area by autonomously calculating a control pattern by a signal controller on each of crossings in accordance with a change in the traffic status of the entire area. SOLUTION: This system is provided with signal lights 3 and 4 provided on the respective crossings in the control area and a local station 2 for controlling the signal lights 3 and 4 respectively. On the basis of the traffic status on the crossing, the local station 2 decides which of the crossings is to be made a basic crossing or a cooperative crossing. On the basis of the traffic status within a range including the adjacent crossing, the basic local station decided as a basic crossing self-independently prepares the control pattern for controlling the crossing. On the basis of the traffic status on the cooperative crossing, the cooperative local station deciding the crossing as a cooperative crossing prepares the control pattern of the cooperative crossing while being cooperative to the control pattern of one of basic crossings.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an autonomous distributed signal control system.

[0002]

2. Description of the Related Art Optimization of control patterns of signal lights is important for smooth road traffic. The control pattern of the signal light device is determined by three control parameters of cycle length, split, and offset. What is cycle length?
It is a control cycle of the signal light device. Split is the percentage of the cycle length of the time that the traffic light is blue. The offset is the time difference between the traffic lights provided at the adjacent intersections turning blue. A signal control system and a signal lamp control method for optimizing these control parameters are being developed.

As one of signal control systems, there is known a signal control system in which a large number of signal lamps are connected to a central signal controller, and the central signal controller determines a control pattern of each signal lamp. The signal control system makes a global situational judgment on the entire area to be managed and controls the entire area. In the signal control system, a method of preparing some control patterns in advance and selecting and executing one of the prepared patterns according to the measured traffic volume is the mainstream method. However, with this method, it is necessary to periodically review the patterns to be prepared according to changes in traffic conditions, and maintenance work is required to create a pattern to be prepared by conducting a traffic volume survey or the like. In addition, if control patterns are automatically generated in response to changes in traffic conditions, the amount of calculations that the central signal control unit has to perform is enormous and complicated. It is difficult to correspond to.

As another signal control system, there is known a signal control system in which a control pattern is individually set in advance for each signal lamp and each signal lamp operates according to the control pattern. However, in this signal control system, each signal lamp can only perform a predetermined operation, and cannot cope with an unexpected increase or decrease in traffic volume.

However, the autonomous distributed control capable of stably and optimally controlling all the intersections of the control area has not been put into practical use at present. It is considered that there are parameters that need to be adjusted to move all intersections stably, and that adjustment know-how is required.

[0006]

SUMMARY OF THE INVENTION An object of the present invention is to respond to changes in traffic conditions in the entire control area without having a large-scale control center facility composed of a central signal control device.
It is an object of the present invention to provide an autonomous distributed signal control system in which the signal control device at each intersection autonomously calculates a control pattern and can stably and optimally control the entire area.

Another object of the present invention is to automatically calculate a control pattern according to changes in traffic conditions, thereby eliminating the need for periodic parameter readjustment and greatly reducing maintenance costs. To provide a control system.

[0008]

[Means for Solving the Problems] Means for solving the problems will be described below with reference to the numbers and symbols used in the embodiments of the present invention. These numbers and signs are
It is added to clarify the correspondence between the description in the claims and the description in the embodiments of the invention. However, the added numbers / codes should not be used to interpret the technical scope of the invention described in [Claims].

The autonomous decentralized signal control system according to the present invention includes signal lamps (3, 4) provided at each intersection (1) of the control area, and a signal lamp (1) provided at each intersection (1). And a local station (2) for controlling the respective units 3, 4. Each of the local stations (2) determines, based on the traffic condition of the intersection where the local station (2) is provided, which one of the basic intersection and the cooperative intersection, the intersection where the own station is provided. Among the local stations (2), the key axis local station that determines that the intersection where the self is installed is the above-mentioned intersection is based on the traffic situation in the range including the adjacent intersection adjacent to the intersection where the self is installed. A control pattern for controlling the provided intersection is created independently. Among the local stations (2), the cooperative local stations that have determined that the self-provided intersection is the cooperative intersection are associated with each other, while cooperating with the control pattern of one of the basic intersections. Create a control pattern for cooperative intersections based on traffic conditions.

Each of the local stations (2) determines, based on the congestion degree of the intersection where it is provided, which one of the basic intersection and the cooperative intersection the intersection where it is provided is set. Preferably.

Further, the cooperative local station is
It is preferable that an intersection in which the self-provided intersection cooperates is selected from the adjacent intersections based on the congestion degree of the adjacent intersection.

The autonomous decentralized signal control system according to the present invention includes signal lamps (3, 4) provided at each intersection (1) of the control area and a signal lamp (3) provided at each intersection (1). 4) and a local station (2) for controlling each of them. A plurality of intersections (1) are selected as basic intersections, and the other of the intersections are selected as cooperative intersections. The control area includes a plurality of sub-areas (23,
24). Among the local stations (2), the core local station whose intersection provided with itself is the above-mentioned intersection is based on traffic conditions in a range including an adjacent intersection adjacent to the intersection provided with itself.
A control pattern for controlling an intersection where the self is provided is created independently. Among the local stations (2), the cooperative local stations whose self-provided intersections are cooperative intersections cooperate with the control pattern of one basic intersection included in the same sub-area among the basic intersections, respectively. The control pattern of the cooperative intersection is created based on the traffic situation of the cooperative intersection provided with.

At this time, it is preferable that whether each of the intersections (1) is selected as the basic intersection or the cooperative intersection is determined based on the congestion degree of the intersection (1).

In the signal control method according to the present invention, each intersection of the control area is classified into (1) one of a basic axis intersection and a cooperative intersection, and a control pattern for controlling the basic axis intersection is set to the basic axis. Creating a step based on traffic conditions in a range including an adjacent intersection adjacent to the intersection, a control pattern for controlling the cooperative intersection, while coordinating with a control pattern for controlling one of the basic intersections, Creating based on the traffic situation at the cooperative intersection.

In the signal control method according to the present invention, a step of classifying each intersection (1) of the control area into one of a basic intersection and a cooperative intersection;
The traffic situation is divided into a plurality of sub-areas (23, 24) so that each includes one basic intersection, and a control pattern for controlling the basic intersection is set to include an adjacent intersection adjacent to the basic intersection. And a control pattern for controlling the cooperative intersection,
Among the basic intersections, while coordinating with a control pattern of one basic intersection included in the same sub-area as the cooperative intersection, each of the basic intersections is created based on a traffic situation of the provided cooperative intersection. .

The signal control program according to the present invention includes a step of classifying each intersection (1) in the control area into one of a basic intersection and a cooperative intersection, and a control pattern for controlling the basic intersection. Creating a step based on a traffic situation in a range including an adjacent intersection adjacent to the key intersection, and a control pattern for controlling the cooperative intersection, while coordinating with a control pattern for controlling one of the key intersections. , Creating based on the traffic situation at the cooperative intersection.

A signal control program according to the present invention includes a step of classifying each intersection (1) of a control area into one of a basic intersection and a cooperative intersection, and each control area includes one basic intersection. Dividing into a plurality of sub-areas (23, 24) so as to include, and creating a control pattern for controlling the basic intersection based on traffic conditions in a range including an adjacent intersection adjacent to the basic intersection. And a control pattern for controlling the coordinated intersection, while coordinating with a control pattern of one of the keyway intersections included in the same sub-area as the coordinated intersection, the traffic of the coordinated intersections respectively provided. Perform the steps to create based on the situation.

[0018]

DETAILED DESCRIPTION OF THE INVENTION Referring to the accompanying drawings,
An embodiment of an autonomous distributed signal control system according to the present invention will be described.

(First Embodiment) In the first embodiment of the autonomous distributed signal control system according to the present invention, as shown in FIG. 1, a local station () is provided at each intersection 1 in the control area. LS) 2 is provided. The road connecting to the intersection 1 is called a link. At this time, the link in which the vehicle flows into the intersection may be called an inflow link, and the link in which the vehicle flows out from the intersection may be called an outflow link. The local stations provided at each intersection 1 are mutually connected by a communication line 21.

FIG. 2 shows a detailed view of one intersection 1. At the intersection 1, a local station 2, a signal light device 3,
4 and a sensor 5 are provided. The signal lamp 3 is provided for a vehicle that flows into the intersection 1 in the lateral direction, and the signal lamp 4 is provided for a vehicle that flows into the intersection 1 in the vertical direction.

The sensor 5 detects the traffic condition of the intersection 1. From the detection result of the sensor 5, the number of vehicles flowing into the intersection 1 is detected. At this time, the sensor 5 can receive the optical beacon, and if the vehicle is equipped with an on-vehicle device for the optical beacon, the sensor 5 receives the ID number of the vehicle. Its ID
The number allows the sensor 5 to identify an individual vehicle. The ID number is used to detect the flow of the vehicle, as will be described later.

The local station 2 includes a signal light device 3,
Control 4 The local station 2 has a CPU2
a is provided. The CPU 2a, according to the software (program) stored in the local station 2,
Perform the required calculations. The local station 2 is connected to another local station provided at an adjacent intersection by a communication line 21, and exchanges information necessary for controlling the signal lamps 3 and 4 via the communication line 21.

The logical structure of control performed by each local station 2 is divided into two layers, an upper layer and a lower layer, as shown in FIG. The control operation of the upper layer is the cycle T
Is performed for _each control operation of lower-order layer is performed in a short every cycle T ₂ than the period T _1. The period T ₁ is typically about 3 minutes and the period T ₂ is typically about 3 seconds.

In the control operation of the upper layer, each local station 2 first determines whether the intersection where the local station 2 is provided (hereinafter referred to as "self intersection") is a basic intersection or a cooperative intersection. The basic intersection is an intersection that serves as a reference for controlling each intersection, and the cooperative intersection is an intersection that is controlled in cooperation with the basic intersection. The intersection is
The determination of whether the intersection is a basic intersection or a cooperative intersection is made for each intersection, and therefore, a plurality of basic intersections may exist.

The local station 2, which has determined that its own intersection is a cooperative intersection, further determines which of the adjacent intersections the own intersection will cooperate with. The cooperation of the intersections is performed in a chain manner, and each cooperative intersection directly or indirectly cooperates with one basic intersection. For example, it is assumed that a cooperative intersection is directly adjacent to a basic intersection and cooperates with the basic intersection. Hereinafter, the cooperative intersection that directly cooperates with the basic intersection and cooperates with each other will be referred to as a primary cooperative intersection. Not directly adjacent to the basic intersection,
And, the cooperative intersection (hereinafter,
It is called "second cooperative intersection". ) Cooperates with the primary cooperative intersection, the secondary cooperative intersection will cooperate with the primary cooperative intersection that cooperates with the basic intersection, and thus indirectly with the basic intersection. In this way, the cooperative intersection that is not directly adjacent to the basic intersection can indirectly cooperate with the basic intersection.

A group of intersections consisting of one basic intersection and a cooperative intersection that directly or indirectly cooperates with the basic intersection constitutes a sub-area, and the control area is divided into one or a plurality of sub-areas. It Referring to FIG. 4, intersection 1
It is assumed that the intersection ₁₁ and the intersection ₁₇ are among the basic intersections. Control area is a subarea 23 comprising cornerstone intersection 1 ₁ is divided into a sub-area 24 which includes a base shaft intersections 1 _7. Cornerstone intersection 1 ₁ Intersection 1 except that in sub-area 23 is a cooperative intersection coordinated to the base shaft intersection 1 _1, the intersection 1 except cornerstone intersection 1 ₇ in the sub-area 24 is coordinated to the base shaft intersection 1 ₇ It is a cooperative intersection. Cooperative intersection in sub-area 23 is controlled base shaft intersection 1 ₁ directly or indirectly coordinated to the intersection 1 except cornerstone intersection 1 ₇ in the sub-area 24 is directly cornerstone intersection 1 ₇ Controlled indirectly or indirectly.

Judgment as to whether each intersection is a basic intersection or cooperative intersection and which cooperative intersection each cooperative intersection cooperates with are determined according to the traffic conditions of each intersection and the adjacent intersections. Is done.

In the present embodiment, the traffic conditions at each intersection 1 are evaluated using the intersection load factor λ as an index. The intersection load factor λ is an index indicating the degree of congestion at each intersection 1.
Referring to FIG. 5, the intersection load factor λ of a certain intersection is calculated by the following equation:

[Equation 1] However, λ _{1 to} λ ₄ are the inflow points 7 where vehicles flow into the intersections.
A load factor of _{1-7 4,}

[Equation 2] Q _i : Traffic flow rate measured at inflow point 7 _i [unit /
Second] S _i : Saturated traffic flow rate at inflow point 7 _i [vehicle / second] R _i : Number of remaining unprocessed vehicles measured at inflow point 7 _i [vehicle] C: Cycle length [second] α: Defined by coefficient . Here, the traffic flow rate is the number of vehicles passing per unit time. The saturated traffic flow rate is the maximum number of vehicles that the link can pass per unit time. The remaining number of vehicles left is the number of vehicles left without passing through the intersection when the traffic light changes from blue to red.
α is a parameter that sets the cycle length after which the unprocessed residue should be dealt with when the unprocessed residue occurs. α determines the degree of the influence of the remaining unhandled R on the intersection load factor, and when α is small, the influence of the remaining untreated number R on the intersection load factor becomes large.

The above-mentioned intersection load factor λ is the sum of the larger load factors at the two inflow points at which the signals become blue at the same time. The inflow point 7 ₁ and the inflow point 7 ₃ are located at positions facing each other, and the signals at the inflow point 7 ₁ and the inflow point 7 ₃ are simultaneously blue. Similarly, the inflow point 7 ₂ and the inflow point 7 ₄ are at positions facing each other, and their signals simultaneously become blue. The intersection load factor λ is defined as the sum of the larger load factor of the inflow points 7 ₁ and 7 _{3 and} the higher load factor of the inflow points 7 ₂ and 7 ₄ .

Each local station 2 calculates the intersection load factor λ of its own intersection. At the local station 2, the intersection load factor λ of its own intersection is a predetermined reference value λ.
When the above _{ST D} defined as self intersection is cornerstone intersection,
Otherwise, the own intersection is determined to be a cooperative intersection.

At this time, the movement history of the basic intersection is taken into consideration in determining the basic intersection. The intersection load factor λ of each intersection fluctuates with time, and therefore the intersection defined as the basic intersection of the intersections 1 also moves with time.
At this time, the intersections selected as the basic intersections are biased toward a plurality of specific intersections, the transitions of the basic intersections have no temporal regularity, and the intersections involved in the movement of the basic intersections are mutually predetermined. If the distance is equal to or more than, the intersections involved in the movement of the basic intersections are all defined as the basic intersections. As a result, the control area is divided into the number of intersections involved in the movement of the basic intersection. If the intersections selected as the basic intersections are not biased toward a specific intersection, if they are biased but have regularity in time, or if they are biased, the mutual distances are not more than the prescribed distance. In that case, the control area division as described above is not performed.

After determining whether the self-intersection is the basic intersection or the cooperative intersection, the local station 2 which determines that the self-intersection is the cooperative intersection is based on the intersection load factor λ of the adjacent intersection adjacent to the self-intersection. , Determine which of the adjacent intersections your own intersection will cooperate with.

This judgment is made by referring to FIG.
The intersection load factor λ of ~ 1e is 80, 20, 5 respectively.
0, 30, and 70, and the above-mentioned reference value λ
_A case where the _STD is 60 will be described as an example. The intersection load factor λ of the intersections 1a and 1e is the reference value λ described above.
_It is higher than the _STD , and the intersections 1a and 1e are defined as basic intersections. The intersections 1b, 1c, and 1d whose intersection load factor λ is lower than the reference value λ _{STD described} above are defined as cooperative intersections.

The local station 2 provided at the cooperative intersection determines which adjacent intersection the own intersection cooperates with, based on the intersection load factor λ of the intersection adjacent to the own intersection.

More specifically, it is provided at the cooperative intersection.
Local station 2 is adjacent to its own intersection
At the intersection, the intersection load factor λ is the predetermined minimum standard for cooperation.
Value λ _MINTo determine if there is anything higher than. Exchange
Difference point load factor λ is the predetermined minimum reference value λ_MINHigher than
If a local station 2 exists, the local station 2
The intersection has the highest intersection load factor λ of the adjacent intersections
It is decided to cooperate with the intersection.

In the example shown in FIG. 6, it is assumed that the cooperative minimum reference value λ _MIN is 35. The intersection load factor λ of the intersection 1a adjacent to the intersection 1b, which is a cooperative intersection, is 8
0, which is higher than 35, which is the minimum reference value λ _MIN for cooperation. The intersection 1b is determined to cooperate with the intersection 1a having the highest intersection load factor λ among the intersections adjacent to the intersection 1b. Intersection 1a is a basic intersection and intersection 1b
Are controlled in cooperation with the basic intersection 1a.

Similarly, the intersection load factor λ of the intersection 1e adjacent to the intersection 1d which is the cooperative intersection is 70, which is the largest among the intersections adjacent to the intersection 1d. Intersection 1
d is determined to cooperate with the intersection 1e.

On the other hand, if there is an adjacent intersection adjacent to the own intersection,
The intersection load factor λ is the predetermined minimum cooperative reference value λ_MIN
If there is nothing higher than the local station
N2 decides its own intersection to either (1) or (2) below.
Meru. (1) Do not cooperate with any adjacent intersections
Make a difference (2) Coordinate with any of the adjacent intersections Which one of the following (1) or (2) should be set
To determine, the local station 2 (1 ') Communicate without cooperating with any adjacent intersections
When it is assumed that control is performed by defining the control pattern of the difference point
Is the delay time t at the intersection₀ (2 ') Adjacent to each adjacent intersection
Establishing the control pattern of the own intersection so as to cooperate with the intersection
Delays that occur at the intersection
Time t₁, T_Two, ... is calculated. Delay time of (1 ')
t₀When calculating, flow from the adjacent intersection to your own intersection.
It is assumed that the traffic of incoming vehicles is uniform in time.
Based on the traffic volume, the delay time t₀Is calculated.
Delay time t₀The traffic volume used to calculate
It is a time average of the traffic volume detected by the device 5.

Further, when calculating the delay times t ₁ , t ₂ ... Of (2 ′), the traffic volume of the vehicle flowing into the intersection where the own intersection is assumed to cooperate is assumed to be uniform in time. It The assumed and determined traffic volume is a time average of the traffic volume detected by the sensor 5 described above. Furthermore, the control pattern of the intersection assumed that the own intersection cooperates is determined so that the delay time generated at the intersection assumed that the own intersection cooperates is minimized. Based on the determined control pattern, the traffic volume of the vehicle flowing from the intersection assumed to cooperate with the own intersection into the own intersection is calculated. Further, it is assumed that the traffic volume of vehicles flowing from the intersection other than the intersection where the own intersection is assumed to cooperate with each other into the own intersection is uniform in time. The assumed and determined traffic volume is a time average of the traffic volume detected by the sensor 5 described above. In this way, the traffic volume flowing into the own intersection is calculated or assumed, and the delay times t ₁ and t are based on the traffic volume flowing into the own intersection.
₂ ... Is calculated. The calculation of the delay times t ₁ and t ₂ of (2 ′) is performed for each of the adjacent intersections.

If the delay time t ₀ is the smallest among the delay times t ₀ , t ₁ , t _2, ...
Resets its own intersection as a basic intersection ((1) above). If any _{one of} the delay times t ₁ , t _2, ... Is the smallest, the local station 2 determines an intersection where its own intersections cooperate so that the delay time is the smallest.

In the example of FIG. 6, the intersection adjacent to the intersection 1c
The intersection load factor λ at the points 1b and 1d is the lowest cooperation level.
Reference value λ_MINIt is lower than 35. Set up at intersection 1c
The lost local station 2 is (1 ') at intersection 1
Without coordinating with either b or 1d, the control pattern of the intersection 1c
Intersection 1 when it is assumed that the turn is determined and controlled.
delay time t generated in c₀And (2'-b) intersection 1b
Establish a control pattern for the intersection 1c so that
Delay at intersection 1c
Time t₁And cooperate with (2'-d) intersection 1d
Assuming that the control pattern of the intersection 1c is determined and the control is performed
Delay time t at intersection 1c_Two And calculate
To do. The local station 2 has a delay time t₀Is the smallest
, The intersection 1c is one of the intersections 1b and 1d.
It is determined that it is a basic intersection that does not cooperate. Locals
Station 2 has a delay time t₁Intersects when is minimal
The point 1c is determined to cooperate with the intersection 1b, and the delay time t_TwoBut
At the minimum, intersection 1c cooperates with intersection 1d
Establish.

By the above process, the distinction between the basic intersection and the cooperative intersection of each intersection 1 and which intersection the cooperative intersection cooperates with are determined.

Hereinafter, in this embodiment, as shown in FIG.
Intersection 1₁, Intersection 1 ₇Is determined to be a basic intersection
Control area is divided into sub areas 23 and 24
Suppose Intersection 1 included in sub-area 23
Chi, intersection 1₁Intersections other than the above are basic intersections 1₁Directly to
It is a cooperative intersection that cooperates indirectly or indirectly. Similarly,
Of the intersections 1 included in the sub-area 24, the intersection 1₇
Intersections other than the above are basic intersections 1₇Directly or indirectly
It is a cooperative intersection that cooperates with.

The following control calculation is independently performed for the intersection group included in the sub-area 23 and the intersection group included in the sub-area 24. The signal lamp control is performed on the sub-area 24 independently of the signal lamp control performed on the sub-area 23. The control of the signal light device performed on the sub-area 23 and the control of the signal light device performed on the sub-area 24 are independent of each other, but are performed in the same process. Of the sub-area 23 and the sub-area 24, the control calculation of the sub-area 23 will be described below as a representative.

Each local station 2 included in the sub-area 23 determines a control target pattern of its own intersection, as shown in FIG. The control target pattern is a target value of the control pattern at each intersection, and includes a target value C _{A of} cycle length, a target value S _{A of} split, and a target value O _A of offset between adjacent intersections. Determined.

First, a local station 2 (hereinafter referred to as "basic axis LS2 ₁ ") provided at a basic axis intersection 1 ₁ )
But it defines the control target pattern of the base shaft intersection 1 _1. The basic axis LS2 ₁ determines the control target pattern of the basic axis intersection 1 ₁ as follows.

FIG. 7 is a flow chart showing a method of setting a control target pattern at a basic axis intersection. Base axis LS2
_{In No. 1} , a plurality of candidate patterns that are candidates for the control target pattern are created by combining the preset cycle length of the basic axis intersection, the split, and the candidate values of the offset between the basic axis intersection and the adjacent intersections (step S0).
1). Several of these candidate values are set in advance for setting the control target pattern.

Then, the basic axis LS2₁Is the basic intersection 1₁To
Baseline, assuming that control is performed according to the candidate pattern
Intersection 1₁Inflow point and crossing point 1₁Adjacent to
The predicted value of the delay time that occurs at the inflow point of the key intersection
Each is calculated (step S02). Step S02
In calculating the predicted value of the delay time of ₁
Intersection 1 adjacent to_Two~ 1₅The cycle length of
Intersection 1₁Is assumed to be the same as the cycle length of. Change
At the cooperative intersection 1_Two~ 1₅The split of a cooperative intersection
1_Two~ 1₅It is uniquely determined based on the number of vehicles flowing into the area.
Calculation of the predicted value of delay time is from the present time to a certain period in the future.
It is done in the middle. The calculation period of the predicted value of delay time is
Period T₁~ T_nIs divided into two equal parts. Period T₁~ T_nThat's it
Let the length be Δt. Period T₁~ T_nDelay in
The predicted values between
Will be issued.

FIG. 8 shows an intersection 1 as a basic intersection.₁Is selected
Flow of calculating the predicted value of delay time when selected
Entry point 8₁~ 8₂₀Indicates. Intersection 1 as a basic intersection₁
Is selected, (1) Cooperative intersection 1_TwoInflow point 8 where vehicle enters₁~ 8
_Three, 8₁₇ (2) Cooperative intersection 1_ThreeInflow point 8 where vehicle enters_Four~ 8
₆, 8₁₈ (3) Cooperative intersection 1_FourInflow point 8 where vehicle enters₇~ 8
₉, 8₁₉ (4) Cooperative intersection 1₅Inflow point 8 where vehicle enters₁₀~
8₁₂, 8₂₀ (5) Basic intersection 1₁Inflow point 8 where vehicle enters_Thirteen~
8₁₆ The delay time in the period T₁~ T_nFor each of
Calculated.

Calculation of the predicted value of the delay time is performed as follows. First, at time t, the inflow volume vehicle to the inlet point _{8 i} per unit time flows _Q i (t) (table /
Second) is calculated. Here, i is an integer of 1 or more and 20 or less, and t = 0 is the start time of the calculation period of the predicted value of the delay time.

Cooperative intersection 1 ₂ adjacent to the basic intersection 1 ₁
To 1 of the inflow point into _5, the base shaft intersection 1 ₁ except the intersection inflow number Q _i per unit of inlet points vehicle flows from (including outside the control area) Time _(t) is actually at its inlet point It is assumed that the time average of the measured vehicle is constant. As shown in Figure 8, if the intersection _{1 1} is selected as a key intersections, inflow volume to _Q 1 per unit of inlet points _{8 1-8 12} time _{(t) ~Q} 12 _(t) is
It is assumed to be constant regardless of time t.

Other inflow points 8_Thirteen~ 8₂₀Unit time of
Per unit Q_Thirteen(T) -Q ₂₀Calculation method of (t)
Will be described below with reference to FIG. Q_Thirteen(T)
~ Q ₂₀Out of (t), intersection 1₁Inflow point 8_Thirteen
Inflow number Q_ThirteenAn example will be described with reference to (t). Basic exchange
Difference point 1₁Inflow point 8_ThirteenAll vehicles flowing into the
Intersection 1₁Intersection 1 adjacent to_TwoOutflow point 9_ThirteenOr
It is a vehicle leaked from. First, the outflow point 9_ThirteenFrom car
Number of outflows per unit time O_Thirteen(T) is
It is calculated.

Outflow Point 9_ThirteenOutflow number O_Thirteen(T) is O_Thirteen(T) = O_Thirteen ¹(T) + O_Thirteen ^Two(T) + O
_Thirteen ^Three(T) O_Thirteen ¹(T): Inflow point 8 per unit time₁Car from
Both turn right and runoff point 9_{1 Three}Number of outflows from O_Thirteen ^Two(T): Inflow point 8 per unit time_TwoCar from
Both go straight and outflow point 9_{1 Three}Number of outflows from O_Thirteen ^Three(T): Inflow point 8 per unit time_ThreeCar from
Both turn left and runoff point 9_{1 Three}Number of outflows from Required by. O_Thirteen ¹(T) is a cooperative intersection 1_Two
Control parameters and Q ₁(T) (= constant value) and right turn rate
α₁Calculated from the product of and. Cooperative intersection 1_TwoControl para
Of the meters, the cycle length is determined by the candidate pattern.
Is assumed to be the same for cycle length and split
It Furthermore, cooperative intersection 1_TwoThe split of a cooperative intersection
1_TwoIt is uniquely determined based on the number of vehicles flowing into the area. As well
To O_Thirteen ^Two(T) is a cooperative intersection 1_TwoControl parameters
Ta, Q_Two(T) (= constant value) and straight traveling rate α_TwoProduct of
Calculated from_Thirteen ^Three(T) is a cooperative intersection 1_TwoControl of
Parameter and Q_Three(T) (= constant value) and left turn rate α_ThreeWhen
It is calculated from the product of O_Thirteen ¹(T) -O_Thirteen ¹(T)
Is a cooperative intersection 1_TwoThe influence of the control of signal lights
Q to receive₁(T) -Q_ThreeUnlike (t), general
However, it is not constant in time.

The number of inflowing units Q ₁₃ (t) per unit time at the inflow point 8 ₁₃ is calculated by Q ₁₃ (t) = O ₁₃ (t−t _t ). Here, _{t t} is the outflow point _{9 13} cooperative intersection _{1 2} are travel time to the base shaft intersection _{1 1} inflow point _{8 13, t t = L /} v L: runoff point _{9 13} and the inlet point 8 Distance v from ₁₃ : Design speed of a link connecting the outflow point 9 ₁₃ and the inflow point 8 ₁₃

Other number of inflows per unit time Q
₁₄ (t) to Q ₂₀ (t) are calculated in the same manner.

Calculated number of inflows Q ₁ (t) to Q
_{From 20} (t), the delay time t _D (i, j) in the period T _j of the inflow point 8 _i is calculated. Period T _j is time t
_The period T _j is _assumed to start at _j−1 and end at time t _j.
Inflow number q _i (j) at the inflow point 8 _i in

[Equation 3] ... Equation (1)

Further, assuming that the number of queues at the inflow point 8 _i in the period T _j is R _i (j), the number of outflow vehicles o that have reached the inflow point 8 _i from the intersection in the period T _{j i} (j) is o _i (j) = min (R _i (j) + q _i (j), p _i (j)) Equation (2) However, the maximum possible outflow number p _i (j) is delayed. The maximum outflow number of vehicles that have reached the inflow point 8 _i when the control is performed according to the candidate pattern that is the time prediction target, and that can flow out from the intersection in the period T _j ,

[Equation 4] However, P _i (t) is P _i (t) = 0 (the period when the signal at the inflow point 8 _i is red) P _i (t) = S _i (the period when the signal at the inflow point 8 _i is blue) S _i : Saturated traffic flow rate at inflow point 8 _i . The number of outflows o _i (j) is the number of queues R
When the sum of _i (j) and the inflowing number q _i (j) is large,
It is constrained by the maximum outflowable number p _i (j). On the other hand, when the sum of the number of queues R _i (j) and the number of inflowing vehicles q _i (j) is small, it is unlikely that more vehicles will flow out than the sum, so the number of outflowing vehicles o _i (j) is , The number of queues R _i
It is determined by the sum of (j) and the inflowing number q _i (j).

Further, the period T_jInflow point 8_iWaiting in
Chi queue number R_i(J), number of inflows q _i(J) and outflow
Number o_iFrom (j), the period T_{j + 1}Waiting stand in
Number R _i(J + 1) is fixed,   R_i(J + 1) = R_i(J) + q_i(J) -o_i(J). ... Formula (3) However, R_i(1) is actually measured at t = 0
The number of waiting queues. From equations (1) to (3),
Period T₁~ T_nFor each of the_i
(J) (j is an integer of 1 or more and n or less) is determined.

The predicted value t _D (i, j) of the delay time of the inflow point 8 _{i in} each period T _j is the queue number R _i.
Proportional to (j), t _D (i, j) = k · R _i (j). Predicted value t of the delay time generated at the inflow point of the basic intersection and the inflow point of the cooperative intersection adjacent to the basic intersection
The sum t _SUM of _D (i, j) is

[Equation 5] Is determined by

[0060] cornerstone LS2 _1, of the candidate patterns, those that the sum _{t SUM} predictive value of the delay time to a minimum, it is determined that the control target pattern cornerstone intersection ₁ (Step S0
3).

The basic axis LS, which defines the control target pattern of the basic intersection, further assumes the control of the basic intersection according to the control target pattern, and shows the temporal change of each traffic flow toward the cooperative intersection adjacent to the basic intersection. Predict,
The predicted time change of the traffic flow is transmitted to the cooperative LS installed at the cooperative intersection through the communication line. The predicted temporal change in traffic flow is called the inflow target profile.

Subsequently, the control target pattern of the first cooperative intersection is determined. Here, the primary cooperative intersection refers to a cooperative intersection that is directly adjacent to the basic intersection among the cooperative intersections. Further, the secondary cooperative intersection is an intersection that is not directly adjacent to the basic intersection and is directly adjacent to the primary cooperative intersection. Similarly, the n-th cooperative intersection is the n-th cooperative intersection.
An intersection that is not directly adjacent to the secondary cooperative intersection and is directly adjacent to the (n-1) th cooperative intersection. Sub area 2
For the 3, cooperative intersection ₁ 2-1 ₅ is a primary cooperative intersection, cooperative intersection _{1 6} is a second-order cooperative intersection.

FIG. 10 shows the control target pattern of the first cooperative intersection.
Show how to determine the turn. Below, the basic intersection 1₁To
Directly adjacent cooperative intersection 1_Two~ 1₅Out of the cooperative intersections
Point 1 ₅As an example, the control target pattern of the first cooperative intersection
The method of determining the zone will be described.

[0064] Cooperative intersection _{1 5} local stations provided in 2 (hereinafter. Referred to as "cooperative LS2 _5") receives the incoming target profile from the base shaft LS2 ₁ (step S11). Hereinafter, the received inflow target profile will be represented as Q _A (t).

[0065] Subsequently, cooperative LS2 ₅ includes a split defined in advance, a combination of candidate values of the offset between adjacent intersections, creating multiple candidate patterns that are candidates for the control target pattern (step S12). At this time, the cycle length of the control target pattern is 1 ₁
To match the cycle length of.

Then, cooperative exchange is performed for each candidate pattern.
Difference point 1₅Inflow link to and intersection 1₅Adjacent to
Secondary cooperative intersection, that is, secondary cooperative intersection 1₆Flow to
Incoming link and cooperative intersection 1₅To basic intersection 1₁Heading to
The sum of waiting times that occur on outflow links is calculated (
Step S13). FIG. 11 shows a waiting time calculation target.
Inflow point 10₁-10₉Is shown. Waiting time calculation
In this case, the first cooperative
Of vehicles entering the secondary cooperative intersection adjacent to the key intersection
The inflow number per unit time is the actual total of the inflow number.
The time average of the measurements is assumed to be constant. Moreover, the said
At the 1st cooperative intersection, the vehicle is directly driven from outside the control range.
When inflowing, the number of vehicles inflowing per unit time
Is the actual measured value of the vehicle entering from outside the control range.
It is assumed to be constant on a time average. Cooperative LS2₅Control
When the target pattern is determined, it is shown in FIG.
Q ₂₁~ Q₂₅Is assumed to be constant regardless of time
To be done. Cooperative LS2₅Is the inflow number Q per unit time
₂₁~ Q₂₅And inflow target profile Q_A(T)
Inflow point 10_Three-10₉Inflow unit per unit time
Number Q₂₆~ Q₂₉And that of the candidate pattern
About each, cooperation LS2₅Inflow link to and cooperation L
S2₅Immediately the sum of the waiting time generated by the outflow link from
The inflow point 10₁-10₉Calculate the sum of waiting times that occur in
Put out.

[0067] Subsequently coordination LS2 _5, as shown in FIG. 10, of the candidate patterns, in which the sum of the waiting time minimal, prescribed control target pattern cooperative intersection _{1 5} (step S14). The control target patterns of the other first cooperative intersections are also determined in the same manner.

The control target pattern for the second cooperative intersection is also determined in the same manner as for the first cooperative intersection. The inflow target profile from the first cooperative intersection to the second cooperative intersection is sent to the cooperative LS provided at the second cooperative intersection. Similar to the first cooperative intersection, the control target pattern of the second cooperative intersection is determined based on the inflow target profile. The control target pattern is similarly determined for the nth cooperative intersection, and the subarea 23 is sequentially selected.
The control target patterns are determined for all the intersections of.

Similarly for the sub-area 24, control target patterns are determined for all intersections. Figure 3
The control operation of the upper layer shown in FIG.

In the control operation of the lower layer, each local station is
Option 2 is a control pattern that minimizes the following evaluation function F.
Execution solution, that is, the cycle length that minimizes the evaluation function F
C, split S, and offset O are calculated. F = α ・ F₁(T_D1, T_D2) + β ・ F_Two(ΔC₁, ΔS₁, ΔO₁) + γ · F_Three(ΔC_Two, ΔS_Two, ΔO_Two) + δ ・ F_Four(Ρ₁, Ρ_Two) ... Formula (4) Here, α, β, γ, and δ are coefficients. F₁Is T
_D1And T_D2A monotonically increasing function for each of
Is. F_TwoIs ΔC₁, ΔS₁, And ΔO₁That's it
This is a monotonically increasing function. F_ThreeIs Δ
C_Two, ΔS_Two, And ΔO _TwoMonotonically increasing for each
It is a function to add. F_FourIs ρ₁And ρ_TwoTo each of
On the other hand, it is a monotonically increasing function. Where the lower layer control
The calculation is independent for each sub-area, similar to the control calculation of the upper layer.
Note that this is done in a standing manner.

In the equation (4), T _D1 is the delay time generated in the inflow link, and T _D2 is the delay time generated in the outflow link.

Delay time T occurring in the inflow link
_D1 is calculated from the cycle length C of the own intersection, the split S, the offset O, and the temporal change of the traffic flow from the adjacent intersection in the future, that is, the inflow prediction profile from the adjacent intersection. In order to calculate the delay time T _D1 of each local station 2, each local station 2 also has the number of queues at the intersection where it is installed and the traffic volume measured by the sensor located upstream thereof. In addition, it predicts and calculates the traffic volume that will flow to the adjacent intersection on the downstream side (outflow link side) in the future. Each of the local stations 2 transmits the predicted outflow traffic amount to the local station 2 at the adjacent intersection on the downstream side. Each local station 2 receives the predicted outflow traffic volume and the inflow prediction profile from an adjacent intersection. Each local station 2 calculates the delay time T _D1 using the received inflow prediction profile, its own cycle length C, split S, and offset O.

On the other hand, since the traffic volume flowing out to the outflow link is predictively obtained according to the cycle length C, the split S, and the offset O of the own intersection, the outflow traffic volume and the outflow prediction from the intersection adjacent to the own intersection are predicted. By comparing with the profile, the time change of the number of queues on the outflow link side is calculated. Here, the outflow prediction profile refers to a time change in which the number of queues on the inflow link of the self-intersection decreases after the present time, and without stopping the vehicles flowing in from the upstream side after the number of queues is completely eliminated. It is composed of changes in the amount of traffic that can be accepted over time. The delay time T _D2 on the outflow side is calculated by integrating the calculated time change of the number of queues on the outflow link side. In order to calculate the delay time T _D2 of each local station 2, each local station 2 has a cycle length C of its own intersection, a split S, and an offset O,
Also, based on the number of queues obtained from the sensor 5 installed at the own intersection, the traffic volume that can be accepted from the adjacent intersection on the upstream side (inflow link side) is predicted and calculated. The predicted inflowable traffic volume is transmitted to the local station 2 at the upstream adjacent intersection.

More specifically, the inflow recharge is performed as follows.
Delay time T_{D 1}And the outflow link
Delay time T_D2And are calculated. On delay
Interval T _D1And delay time T_D2And are respectively inflow links
And the number of queues of outgoing links. This style
The number of queues for incoming and outgoing links is at each intersection
The passing information of the vehicle acquired by the provided sensor 5 and the
Cycle length C, split S, and offset
It is required from the O. Sensing, as described above
The device 5 can receive the optical beacon. By optical beacon
Based on the transmitted ID number, the sensor 5 detects that the optical beacon
The vehicle equipped with the on-vehicle device can be individually identified.
Furthermore, the sensor 5 does not include an on-board device for an optical beacon.
Even if a new vehicle passes,
The body can be detected. Onboard optical beacon device
The vehicles on which they are placed can be individually identified by their ID numbers.
And are used to calculate the number of unhandled vehicles.

As shown in FIG.
-A group of vehicles including vehicle α equipped with a vehicle-mounted device on the upstream side
Intersection 1_APass the intersection 1 on the downstream side_BHeaded for
And Intersection 1 on the upstream side_ASense the sensor 5 installed in
Intellectual device 5_A, Intersection 1 on the downstream side _BSensor 5 installed in
Sensor 5_BAnd

As shown in FIG. 12 (b), sensing
Bowl 5_AIs n in front of the vehicle α₁Stand, n after vehicle α_TwoStand
Detect that the vehicle has passed. Shown in FIG.
In the example, n₁= 1 and n_Two= 3. Meanwhile, sense
Bowl 5_BIs intersection 1 on the downstream side _BWhile the traffic light is blue,
N in front of vehicle α_ThreeStand, n after vehicle α_FourVehicles pass
Detect that you have done. In the example shown in FIG.
Is n_Three= 5 and n_Four= 1. At this time, downstream
The waiting line left on the link just before the traffic light turned green
Number of rows n_AAnd immediately after the downstream signal turns blue,
Number of queues left n_BIs n_A= N_Three-N₁， n_B= N_Four-N_Two， It is calculated by.

When the delay time T _D1 of the inflow link is calculated, while the signal of the upstream intersection 1 _B is blue,
Information that n ₁ vehicles have passed in front of the vehicle α and n ₂ vehicles have passed after the vehicle α is notified from the intersection 1 _A on the upstream side to the intersection 1 _B on the downstream side. From this information, the local station at the intersection 1 _B on the downstream side can calculate the remaining unhandled numbers n _A and n _B of the inflow links. Further, the local station at the intersection 1 _B on the downstream side receives the delay time T generated on the inflow link from the queue numbers n _A and n _B.
D1 can be predicted and calculated with high accuracy.

Similarly, when the delay time T _D2 of the outflow link is calculated, while the traffic light of the intersection 1 _B on the downstream side is blue, n ₃ vehicles are in front of the vehicle α and n ₄ are behind the vehicle α. The information that one vehicle has passed is notified from the intersection 1 _B on the downstream side to the intersection 1 _A on the upstream side. From this information, the local station at the intersection 1 _A on the upstream side can calculate the queue numbers n _A and n _B of the links. Furthermore,
The local station at the intersection 1 _A on the upstream side can highly accurately predict and calculate the delay time T _D2 generated in the outgoing link from the queue numbers n _A and n _B.

In the evaluation function F, the delay times T _D1 and T
_Since the term α · F ₁ (T _D1 , T _D2 ) having _D2 as an argument is included, control of the signal light device is realized in consideration of delay times of both the inflow link and the outflow link. That is, the signal lamp of a certain intersection controls the signal lamp while optimizing both the degree of congestion at its own intersection and the degree of congestion at the intersection on the downstream side of its own intersection.

At this time, the function F ₁ has a value n ₁ at the time when the number of queues occurring in the inflow link becomes the _maximum instead of the delay time T _D1 in the inflow link or in addition to the delay time T _D1. Can be used as an argument. Similarly, instead of the delay time T _D2 on the outgoing link,
Alternatively, in addition to the delay time T _D2 , the value n ₂ at the time when the number of queues occurring in the outflow link becomes maximum can be used as an argument. Therefore, by controlling the number of queues at the time of the maximum, and controlling the number of queues that can be stored in the link without overflowing, the control to actively prevent the deadlock is implemented. It

ΔC ₁ , which is the argument of the second term of the evaluation function F,
ΔS ₁ and ΔO ₁ are the cycle length C of the execution solution of the control pattern, the split S, and the offset O, respectively, and the cycle length C _A of the control target pattern of the self-intersection determined by the above-described logical operation of the upper layer. Split S _A , offset O
_It is the difference from _A. In the evaluation function F, β · F ₂ (ΔC ₁ ,
Since the terms ΔS ₁ and ΔO ₁ ) are included, the signal light device is controlled in accordance with the control target pattern of the own intersection.

ΔC ₂ , which is the argument of the third term of the evaluation function F,
ΔS ₂ and ΔO ₂ are the differences between the cycle length C, the split S, and the offset O of the execution solution of the control pattern, and the cycle length C _A , the split S _A , and the offset O _A of the control target pattern at the adjacent intersection, respectively. is there. Evaluation function F
Includes the term γ · F ₃ (ΔC ₂ , ΔS ₂ , ΔO ₂ ), the control of the signal light device is performed according to the control target pattern of the adjacent intersection.

On the other hand, ρ which is the argument of the fourth term of the evaluation function F
₁ and ρ ₂ indicate the possibility that the speed of the vehicle group of the link exceeds the design speed, respectively, and ρ ₁ = (O−O _R1 ) · u (O−O _R1 ), ρ ₂ = (O−O _R2 ) .U (O- _OR2 ), is represented. However, O _R1 is an offset that is optimal for matching the speed of the vehicle group of the outflow link with the design speed. _OR2 is an offset that is optimal for matching the speed of the vehicle group on the inflow link to the design speed. u is
Unit step function, u (x) = 1 (x> 0), u (x) = 0 (x ≦ 0). _OR1 and _OR2 are calculated as follows.

FIG. 13 shows the setting of the speed of the vehicle group of the outflow link.
Offset O that is optimal to match the total speed_R1of
The calculation method is shown. Intersection 1 on the upstream side_AProvided in
Cal station 2 is an intersection 1 on the upstream side_AAnd upstream
Intersection 1_BThe vehicle traffic acquired by the sensor 5 installed at
Excessive information and intersection 1_BCyclic execution of control patterns
Downstream from the length C, split S, and offset O
Intersection 1_BCalculate the number of queues just before the traffic lights turn green
Put out. Intersection 1 on the downstream side_BAfter the traffic light turned blue,
Intersection 1 on the downstream side_BThe time it takes for the queue to clear
Between t_R1Then, O_R1Is O_R1= L / v-t_R1 Is. Where L is the distance between intersections and v is the setting
It is a speedometer. Intersection 1 on the upstream side_AAnd intersection 1_BOh
Husset O is O_R1Less than that shown in Figure 11
Due to the psychology of the driver, the speed of the vehicle group
Tends to rise. At this time, ρ₁= (O-O_R1) ・ U (O-O_R1), Ρ determined by₁Set offset O so that approaches 0
By changing the outflow link to the speed of the vehicle group
Can be kept small.

FIG. 14A shows the speed of the vehicle group of the inflow link.
Is an offset O that is optimal for matching the design speed.
_R2The calculation method of is shown. Intersection 1 on the upstream side_ASent by
From the incoming runoff profile, intersection 1 on the downstream side_BTo
The provided local station 2 is the intersection 1
_BYou can calculate the remaining number of units left before the traffic light turns green.
At this time, your own intersection 1_AEven though the queue of
Time required t_R2Then, O_R2Is O_R2= L / v-t_R2 Is. Where L is the distance between intersections and v is the setting
It is a speedometer.

As shown in FIG. 14B, the timing at which the intersection 1 _B on the downstream side becomes blue is offset O.
_If it becomes faster than the timing determined by _R2 , the speed of the vehicle group tends to increase due to the psychology of the driver. Similar to the outflow link, ρ ₂ = (O−O _R2 ) · u (O−O _R2 ), that is, ρ ₂ is set close to 0 to bring the speed of the vehicle group of the inflow link close to the design speed. You can

Since the evaluation function F includes the term δ · F ₄ (ρ ₁ , ρ ₂ ) having the above-mentioned ρ ₁ and ρ ₂ as arguments, the speed of the vehicle group is appropriately controlled. be able to.
Note that the argument of F ₄ is not limited to include both ρ ₁ and ρ _2, and only one of them can be the argument of F ₄ .

Coefficients α, β, γ, δ of each term of the evaluation function F
Different values are used for the basic intersection and the cooperative intersection. When calculating the execution solution of the control pattern at the basic intersection, β is set to 1.0 and α, γ, and δ are set to 0, and the evaluation function F is calculated. As a result, the evaluation function F is β · F
₂ (ΔC ₁ , ΔS ₁ , ΔO ₁ ). The basic intersection is substantially controlled by a control pattern close to the control target pattern of the basic intersection. That is, the basic intersection is controlled independently. Since the control target pattern is determined so that the delay time at the basic intersection and the adjacent intersection is positively reduced,
After all, the control parameter of the basic intersection is determined so that the delay time between the basic intersection and the adjacent intersection is positively reduced.

On the other hand, the execution solution of the control pattern of the cooperative intersection
When calculating, β and γ are small and α is large.
The evaluation function F is set and calculated. The evaluation function F is α
・ F ₁(T_D1, T_D2)
Becomes α ・ F₁(T_D1, T_D2) Is emphasized
By doing so, a cooperative intersection can be coordinated with an adjacent intersection.
It is controlled while trying. As a result, directly at the basic intersection
At the 1st cooperative intersection adjacent to
The signal lights are controlled accordingly. Furthermore, the second cooperative intersection
Then, the signal lights are controlled in coordination with the first cooperative intersection.
Be seen. Similarly, at the nth cooperative intersection, the n-1st cooperative cooperative
The signal lights are controlled in cooperation with the key intersection. This
Due to such a chain of cooperative relations, the intersection to be controlled is
Controlled directly or indirectly in coordination with the key intersection
It Through such coordinated control, the entire signal control system
As a result, stabilization and optimization are achieved.

At this time, the calculation of the evaluation function F performed by the local station 2 is performed according to the equation (4) although the setting values of the parameters α, β, γ and δ are different. That is, the control patterns of the basic intersection and the cooperative intersection are both calculated based on the equation (4).
As a result, the software for calculating the control pattern between the basic intersection and the cooperative intersection is shared.

Coefficients α, β, γ, δ of each term of the evaluation function F
Can be changed according to regional characteristics and time of day. As a result, optimal control is realized according to regional characteristics and time zones. For example, the value of δ is increased at the road where it is expected that the speed of the vehicle is excessive, or at night. Accordingly, δ · F ₄ (ρ ₁ , ρ ₂ ) for the value of the evaluation function F
The influence of is increased, and the control of the signal light device is performed with an emphasis on prevention of overspeeding of the vehicle group between the intersections.

The signal control system described above can easily expand its control area as described below. FIG. 15 shows a method of expanding the control area. Consider the case where an intersection 1 ₁₁ adjacent to the intersection ₁₇ is added to the control area as shown in FIG.

[0093] (1) the intersection _{1 11} includes a local station _{2 11} newly established, the establishment of sensor 5 ', the signal lamp device (not shown) and is provided. (2) The local station 2 ₁₁ installed at the intersection 1 ₁₁ is connected to the communication line 21. (3) to the local station 2 _11, other existing local station 2 identical software and is installed. As a result, the local station 2 ₁₁
Performs the same operation as the other existing local station 2. (4) to the local station 2 _11, the order of lighting the signal lamp device, yellow time, total red time, the maximum number of seconds the green time,
The set value considering the safety after operation, such as the minimum number of seconds, is entered. (5) In addition to the local station 2 _11, the direction of the intersection 1 ₇ local station 2 ₇ existing is provided, the distance of the link connecting the intersections 1 ₃ and intersections 1 _7, the design speed is set. (6) On the other hand, in the existing local station 2 ₇ , the direction of the intersection 1 ₁₁ in which the local station 2 _{1 1} is newly established, the distance of the link connecting the intersection 1 ₇ and the intersection 1 ₁₁ and the design speed are set. . During the setting input, a local station 2 ₇ performs control of the signal lamp device in the control pattern of a fixed set in hardware, software for control is stopped. After (7) setting the input complete, the software of the local station 2 ₇ is restarted and the local station 2
₇ restarts the autonomous distributed control with the new setting. (8) Furthermore, the local station ₂₁₁ is activated,
Autonomous distributed control of the control area including the intersection 1 ₁₁ is started.

As described above, when a new intersection is added, a signal control device is newly installed at the intersection and a new signal control device is installed at another intersection without performing the parameter adjustment work by the traffic volume survey. The control area can be expanded simply by connecting to a common communication line with the installed signal control device.

As described above, in the signal control system of the present embodiment, control is performed while the coordinated intersections are directly or indirectly coordinated with each other at the axis intersections that are controlled independently. , Stabilization and optimization as a whole are realized. At this time, a plurality of basic intersections are determined according to the traffic situation, and the control area is divided into sub-areas each including one basic intersection, so that the traffic volume is larger and the need for optimization is greater. More optimal control is performed centering on the basic intersection.

Further, the basic intersection is determined on the basis of the intersection load factor λ, and the signal control is performed centering on the intersection having a large traffic volume. As a result, the entire area is coordinated around the most congested intersection in the controlled area, so that the congestion is eliminated and the traffic volume in the entire area is leveled.

(Second Embodiment) FIG. 16 shows a second embodiment of the autonomous distributed signal control system according to the present invention. In the second embodiment, a master station 22 which gives a global control guide to each local station 2 is connected to the communication line 21. Each local station 2
Determines the control pattern of its own intersection based on the control guideline, and thereby optimizes the overall control.
Other configurations of the autonomous distributed signal control system according to the second embodiment are the same as those of the first embodiment, and description thereof will not be given.

Logical structure of control performed in the second embodiment
Is the upper layer, the middle layer and the
It is divided into a total of three lower layers. Upper layer, middle layer, and
And the control calculation of the lower layer are₁₁, Cycle T
₁₂, And period T_ThirteenIt is done every time. Where the upper layers
Control calculation cycle T₁₁Is the cycle T of the control operation of the middle layer
₁₂The period T of the control operation of the middle layer which is longer and longer₁₂Is subordinate
Layer control calculation cycle T_ThreeLonger than. Cycle T₁₁Is the source
The pattern is about 15 minutes, and the cycle T_TwoIs typically 3
Minutes, and the cycle T_ThreeIs typically around 3 seconds
It

In the control operation of the upper layer, a policy control guideline for the entire control area is established. The control guideline includes a target traffic flow rate of each link in the control area and its allowable range, and a control guideline pattern and its allowable range. The target traffic flow rate is a target value of the number of vehicles passing per unit time. The maximum value and the minimum value that can be allowed for the target traffic flow rate are set, and the allowable range of the target traffic flow rate is set. The control guideline pattern is a solution that is easily calculated and is considered to be close to the optimum solution of the control pattern performed at each intersection.The control guideline cycle length, which is a guideline for the cycle length of the entire control area, It is expressed by a control guide split which is a guide of split, a control guide offset which is a guide of offset of each link, and their respective allowable ranges.

The master station 22 formulates the control guideline. The master station 22 transmits the established control guideline to each local station 2 via the communication line 21. As will be described later, each local station 2 calculates a more optimal control pattern execution solution with reference to the received control guideline, and controls the signal lamps 3 and 4. By referring to the control guideline when calculating the execution solution of the control pattern, global optimization of the entire area can be achieved.

More specifically, the control operation of the upper layer is performed as follows. First, the OD demand forecast is calculated. For the prediction of OD demand forecast, in the control area,
A pair of origin and destination is defined. OD demand forecast is
It is composed of the predicted value of the traffic demand from the starting point to the destination, which is calculated for each of the sets, from the present time to a predetermined certain time later. Typically, the predicted value of the traffic demand for 15 to 30 minutes from the present time is used as the OD demand forecast.

For the calculation of the OD demand forecast, the above-mentioned uplink information of the optical beacon is used. When the vehicle runs in the control area, the ID number of the vehicle can be identified by the uplink information of the optical beacon received by the sensor 5. From the ID number, the starting point and the destination of each vehicle traveling in the control area are specified. Information on the starting points and destinations of a plurality of vehicles is collected and registered in the past value database. Past value database is OD for each day of the week and time
Manage demand. Using the past value database,
Forecast OD demand for 15 to 30 minutes in the future.

Then, the target traffic flow rate Q _k of each link is set from the calculated OD demand forecast. According to the OD demand forecast, it is assumed that each vehicle passes through the link which is the shortest route, and the traffic volume per unit time passing through each link is obtained, which is set as the target traffic flow rate. However, when the target traffic flow rate exceeds the processing capacity of the intersection, the processing capacity Q _k ′ of the intersection is set as the upper limit value Q _k ^max of the target traffic flow rate. If there is a detour, the traffic flow rate of the link to be the detour is corrected and the detour traffic volume is added.

Here, the processing capacity Q at the intersection^k'Is the following
Ask like. Q_k’= G_k× S_k G_k: Time ratio at which the traffic light on the link k side of the target intersection is blue
rate S_k: Saturated traffic flow rate of link k G_kIs the target intersection of each link connecting to the target intersection.
Flow rate Q_kCalculated from the ratio of

The target traffic flow rate Q _k and the upper limit Q _k ^max of each link calculated in this way are transmitted to each local station 2 and used for calculating the control pattern.

Subsequently, a control guide pattern is created by the following process. First, a plurality of control guideline candidate patterns that are candidates for control guideline patterns are created. In order to create the control guideline candidate pattern, a candidate pattern length of several patterns is prepared as a candidate for the cycle length of the control guideline candidate pattern, and a candidate split of several patterns is prepared as a candidate for the split of the control guideline candidate pattern,
Several candidate pattern offsets are prepared as candidates for the control guide candidate pattern offset. The candidate offset is expressed as a ratio (unit is%) to the candidate cycle length,

Each of the candidate cycle length, the candidate split, and the candidate offset is given as an arithmetic progression having a certain value as a tolerance. For example, 60s, 70 every 10s
10 candidate cycle lengths of s, ..., 150 s are prepared, and four candidate offsets of 0%, 25%, 50%, and 75% are prepared for every 25%. The prepared candidate cycle length, candidate split, and candidate offset are combined to create a plurality of control guideline candidate patterns.

At this time, in order to reduce the amount of calculation required for formulating the control guideline pattern, the control guideline candidate pattern is created under the following conditions. First, among the prepared candidate cycle lengths, only those having a required cycle C _min or more at the intersection having the largest traffic load factor λ are used for the combination. Here, the required cycle C _min is

[Equation 6] L: A value defined by the total lost time (the time when the signal is yellow and the time when the signals in all directions are red). The intersection load factor λ is a value defined by [Equation 1] described in the first embodiment, and represents the degree of congestion at the intersection, as described above.

Each local station 2 calculates the intersection load factor λ of the intersection where it is installed and notifies the master station 22 of it. From the notified intersection load factor λ, the master station 22 calculates the required cycle C _min of the intersection having the maximum intersection load factor λ.

Of the prepared candidate cycle lengths, the master station 22 creates only the above-mentioned control guideline candidate patterns by combining only the required cycle C _min or more of the intersection having the largest traffic load factor λ. A cycle length shorter than the required cycle C _min is obviously not optimal in consideration of handling the traffic at the intersection with the largest traffic load factor λ. A control guideline candidate pattern including such a cycle length is not created, and thus the number of control guideline candidate patterns is reduced.

Secondly, the control guideline candidate offsets of links whose distances are shorter than a predetermined reference length are all 0, and a plurality of candidate offsets are not prepared for that link. A large offset value for a link with a short distance is not preferable in terms of traffic control. The control guideline candidate offsets of links whose distance is shorter than a predetermined reference length are all set to 0, and the number of control guideline candidate patterns is reduced.

Thirdly, among the links connecting the intersections, the control guideline candidate offset of the link whose target traffic flow rate is smaller than a predetermined reference value is a certain value which is a function of the length of the link and the cycle length. To be Links with less traffic have less impact on the overall traffic control.
Therefore, the offset of such a link is determined by a simple rule, and the number of control guideline candidate patterns is reduced.

Fourth, when a closed loop is composed of a plurality of links, as shown in FIG. 18A, the sum of the offsets of the links forming the closed loop is 100.
The control guideline candidate offset is set to be a multiple of%. A control guideline candidate offset in which the sum of the offsets of the links forming the closed loop is not a multiple of 100% as shown in FIG. 18B is not adopted. In the optimal control of traffic, the sum of the offsets of the links forming the closed loop is optimally a multiple of 100%. Therefore, a set of offsets in which the sum of the offsets of the links forming the closed loop is not a multiple of 100% is not included in the control guideline candidate pattern. As a result, the number of control guideline candidate patterns is reduced.

With respect to each of the created control guideline candidate patterns, of all the links included in the control area, the sum of the delay times generated in the main links having the target traffic flow rate higher than the predetermined reference value is calculated and calculated. The smallest total sum of the delay times is defined as the control guideline pattern. At this time, only the links having a target traffic flow rate higher than a predetermined reference value are targeted for the delay time, thereby reducing the amount of calculation when formulating the control guideline pattern.

Further, the master station 22 calculates
The allowable range of the specified control guideline pattern is determined. Master
Station 22 uses the control pointer size of the control pointer pattern.
Curl length C^TPermissible range ΔC^TSet each intersection 1
Control pointer split S_j ^TFor each
Surrounding ΔS_j ^TAnd the control guideline offset O for each link _k
^TFor each, the allowable range ΔO_k ^TDetermine. Tolerance
Range ΔC^TIs the tolerance of the candidate cycle length, which is an arithmetic sequence.
Determined to match. 1 as in the example above
Every 10 seconds, 60s, 70s, ..., 150s
Allowable range ΔC when complementary cycle length is prepared
^TIs 10 s, which is equal to the tolerance of the candidate cycle length. same
The allowable range ΔS_j ^TIs within the tolerance of the candidate split.
It is determined to match, and the allowable range ΔO_k ^TOff the candidate
Defined to match set tolerances.

When calculating the control pattern of each intersection,
The cycle length of all intersections 1 in the control area is C_TInside
Hearty C^T-ΔC^TTo C^T+ ΔC^TDetermined by
Be done. Furthermore, the split at each intersection 1 is S_j ^TCentered around
And S_j ^T-S_j ^TTo S _j ^T+ ΔS_j ^TFixed in the range of
And the offset of each link is O_j ^TCentered on
O_j ^T-O_j ^TTo O_j ^T+ ΔO_j ^TDefined in the range of
It

With the above, the control operation of the upper layer is completed.

The control operation of the middle layer is performed by each local station 2. In the control calculation of the middle layer, as shown in FIG. 17, each local station 2 determines whether the self-intersection is a basic intersection or a cooperative intersection, and further determines the self-intersection as a cooperative intersection. The local station 2, which has been determined to exist, determines which of the adjacent intersections the own intersection will cooperate with. A process in which each local station 2 determines whether the self-intersection is a basic intersection or a cooperative intersection, and which local station 2 determines that the self-intersection is a cooperative intersection, one of the intersections to which the self-intersection is adjacent The process of deciding whether to cooperate with is the same as that performed in the first embodiment. No detailed explanation will be given.

Similar to the first embodiment, a group of intersections each including one basic intersection and a cooperative intersection that directly or indirectly cooperates with the basic intersection constitutes a sub-area,
The control area is divided into one or a plurality of sub areas.
With reference to FIG. 19, in the following, it is determined that the intersections 1 ₁ and _{17 of the} intersections 1 are basic intersections, and the control area is a sub-area 23 including the basic intersection 1 ₁ ,
The description will be given assuming that the sub-area 24 is divided into the sub-area 24 including the basic intersection ₁₇ .

The control calculation performed below is performed in the sub-area 2
The intersection group included in 3 and the intersection group included in the sub-area 23 are independently performed. The signal lamp control is performed on the sub-area 24 independently of the signal lamp control performed on the sub-area 23. The control of the signal light device performed on the sub-area 23 and the control of the signal light device performed on the sub-area 23 are independent of each other, but are performed in the same process. Of the sub-area 23 and the sub-area 24, the control calculation of the sub-area 23 will be described below as a representative.

Each local station 2 included in the sub-area 23 defines the control target pattern of the intersection where it is installed. The control target pattern is the target value of the control pattern at each intersection, and is the target value C _U of the cycle length, the target value S _{U of the} split, and the target value O _U of the offset between the adjacent intersections. Determined by Tsuto.

In determining the control target pattern, first, the basic axis
The control target pattern at the intersection is determined, and then the
The control target pattern of the cooperative intersection is decided in cooperation with the axis intersection.
Is determined. FIG. 20 shows the control target pattern of the basic intersection.
It is a flow chart which shows a setting method. Sub area 23
Axis crossing 1₁Local station
2 (hereinafter, “basic axis LS2₁". ) Is the control described above
Set the pointer pattern and its allowable range to the master station.
Control guide pattern received from
Based on the allowable range, the control that becomes a candidate for the control target pattern
Creating a plurality of target candidate patterns (step S2
1). The control pointer pattern has a control pointer cycle length C.^T
And its allowable range ΔC^T, Control pointer split length S^TToso
Allowable range ΔS^T, Control pointer offset O^T _Two,
O^T _Three, O^T _Four, O^T ₅And its allowable range ΔO^T _Two, Δ
O^T _Three, ΔO^T _Four, ΔO^T ₅And are included. here
And O^T _Two, O ^T _Three, O^T _Four, O^T ₅Is shown in FIG.
As shown in the figure, each of the basic intersections 1 ₁And next to it
Each cooperating intersection 1 that touches_Two~ 1₅Link 11 connecting to
_Two~ 11 ₅Is the control guideline offset defined for
It Allowable range ΔO^T _Two, ΔO^T _Three, ΔO^T _Four, And ΔO
^T ₅Is the control pointer offset O^T _Two,
O^T _Three, O^T _Four, And O^T ₅Is the allowable range.

[0123] cornerstone LS2 _1, the control guidelines cycle length ^{C T}
Centered on the control guideline cycle length C ^T , the five cycle lengths C ^T −2 · ΔC, C ^T −ΔC, C ^T , C ^T + ΔC, and C ^T + 2 · ΔC are the control target pattern cycles. Make it a candidate for the chief. However, ΔC = ΔC ^T / 4, ^is, at equal intervals in the range of from ^{C T -ΔC} T to ^{C T} + [Delta] C ^T, is defined five cycle length candidate.

[0124] Furthermore, the base shaft LS2 ₁ is centered on the control pointer split ^{S T,} 5 single split ^S T -2 · ^ΔS in the vicinity of the control pointer split ^{S T, S T -ΔS, S} T, S T + ΔS , And S ^T + 2 · ΔS are candidates for the cycle length of the control target pattern. However, ΔS = ΔS ^T / 4, ^is, at equal intervals in the range of from ^{S T -ΔS} T to ^{S T} + ^{ΔS T,} defined five cycle length candidate.

Further, the basic axis LS2₁Is link 11_Two~ 1
1₅For each of the^T _iTo
Center and its control guideline offset O^T _iIs near
5 offsets O^T _i-2 / ΔO_i, O^T _i-ΔO_i, O^T _i, O^T _i
+ ΔO_i, O^T _Two+ 2 · ΔO_i Each link of the control target candidate pattern 11_Two~ 11₅Oh
It is a candidate for Husset. Where i is 2 or more and 5 or less
It is an integer. Furthermore, ΔO_i= ΔC^T/ 4, And O^T _i-ΔO^T _iTo O^T _i+ ΔO^T _iRange of
Thus, five offset candidates are defined at equal intervals.

[0126] cornerstone LS2 _1, the candidate and has been cycle length, split, and a combination of offset, to create multiple control target candidate pattern. Cycle length above,
From the combination of split and offset, 5 ⁶ control target candidate patterns will be created.

Then, as shown in FIG. 20, the basic axis LS2 ₁ is the basic axis intersection 1 on the assumption that the basic axis intersection 1 ₁ is controlled according to the above-mentioned control target candidate pattern.
₁ of the inlet points, and the predictive value of the delay time occurring in the inflow point cooperating intersection adjacent to the base shaft intersection 1 ₁ respectively calculated (Step S22). The process of calculating the predicted value of the delay time is the process performed in the determination of the control target pattern of the basic axis intersection in the first embodiment (step S0 of FIG. 7).
It is the same as 2) and its detailed description will not be given.

[0128] cornerstone LS2 ₁ among the control target candidate pattern, what the sum _{t SU M} predictive value of the delay time to a minimum, it is determined that the control target pattern of the base shaft intersection _{1 1} (step S23).

The basic axis LS2 ₁ further predicts and predicts the time change of each traffic flow toward the cooperative intersection adjacent to the basic axis intersection 1 ₁ on the assumption that the basic axis intersection 1 ₁ is controlled according to the control target pattern. The time change of the traffic flow is transmitted to the local station 2 installed at the cooperative intersection through the communication line 21. The predicted temporal change in traffic flow is called the inflow target profile.

Then, the control target pattern of the first cooperative intersection is determined. As described above, the primary cooperative intersection refers to a cooperative intersection that is directly adjacent to the basic intersection among the cooperative intersections.

FIG. 22 shows the control target pattern of the first cooperative intersection.
Show how to determine the turn. Below, the basic intersection 1₁To
Directly adjacent cooperative intersection 1_Two~ 1₅Out of the cooperative intersections
Point 1 ₅As an example, the control target pattern of the first cooperative intersection
The method of determining the zone will be described.

[0132] Cooperative intersection _{1 5} local stations provided in 2 (hereinafter. Referred to as "cooperative LS2 _5") receives the incoming target profile from the base shaft LS2 ₁ (step S31). Hereinafter, the received inflow target profile will be represented as Q _A (t).

[0133] Further, coordination LS2 ₅ receives the aforementioned control guidance pattern from the master station 31, the control guidance pattern received, creating multiple control target candidate patterns that are candidates for the control target pattern (step S32).

[0134] The cycle length for all control target candidate pattern created is determined to match the cycle length of the control target pattern of the base shaft intersection 1 ₁ defined as described above.

[0135] Further, the control guidance split ^{S T} cooperative intersection _{1 5} defined in a pattern centered, its control guidelines below five in the vicinity of the offset ^{S T} offset ^{S T -2 · ΔS, S T} -ΔS, S ^T , S ^T + ΔS, S ^T
+2 · [Delta] S becomes the respective split control target candidate pattern cooperative intersection 1 _5. Here, ΔS is calculated as ΔS = ΔS ^T / 4, using the allowable range ΔS ^T of the control guide split.

Furthermore, the cooperation defined in the control guideline pattern
Intersection 1₅Offset O of each link connected to_i ^TInside
As a heart, the control guideline offset O_i ^TBelow near
5 offsets O_i ^T-2 / ΔO_i, O_i ^T-ΔO_i, O_i ^T, O_i ^T
+ ΔO_i, O_i ^T+ 2 · ΔO_i However, cooperative intersection 1₅Control target candidates for each link connecting to
It becomes each offset of the pattern. Where ΔO_iIs the control
Allowable range ΔO of each link specified in the guideline offset
_i ^TUsing, ΔO_i= ΔO_i ^T/ 4, It is calculated by.

[0137] Then, for each control target candidate pattern, and the inflow link to the cooperative intersection 1 _5, cooperative intersection 1 ₅
The sum of the waiting times generated at the secondary cooperative intersection adjacent to, that is, the inflow link to the secondary cooperative intersection 1 ₆ and the outgoing link from the cooperative intersection ₁₅ to the basic intersection 1 ₁ is calculated (step S33). The process of calculating the sum of waiting times performed in step S33 is the same as the process of calculating the sum of waiting times in the first embodiment (step S13), and a detailed description thereof will not be given.

[0138] Subsequently coordination LS2 _5, as shown in Figure 22, of the control target candidate pattern, what the sum of the waiting time minimal, prescribed control target pattern cooperative intersection 1 ₅ (step S14 ). The control target patterns of the other first cooperative intersections are also determined in the same manner.

The control target pattern for the secondary cooperative intersection is also determined in the same manner as for the primary cooperative intersection. The inflow target profile from the first cooperative intersection to the second cooperative intersection is sent to the cooperative LS provided at the second cooperative intersection. Similar to the first cooperative intersection, the control target pattern of the second cooperative intersection is determined based on the inflow target profile and the control guideline pattern. Similarly, the control target pattern is determined for the nth cooperative intersection, and the control target pattern is sequentially determined for all intersections in the controlled area.

Similarly, for the sub-area 24, the control target patterns of the basic intersection and the cooperative intersection are determined,
The control operation of the middle layer shown in FIG. 20 is completed as described above.

In the control operation of the lower layer, each local station 2 executes the control pattern which minimizes the following evaluation function F ′, that is, the cycle length C, the split S and the offset O which minimize the evaluation function F ′. To calculate. F ′ = α · F ₁ (T _D1 , T _D2 ) + β · F ₂ (ΔC ₁ , ΔS ₁ , ΔO ₁ ) + γ · F ₃ (ΔC ₂ , ΔS ₂ , ΔO ₂ ) + δ · F ₄ ( ρ ₁ , ρ ₂ ) + ε · F ₅ (ΔC ₃ , ΔO ₃ ) + ζ · F ₆ (Q _k , Q _k ^max , Q _k ^a ). Expression (4) ′ The evaluation function F ′ defined by the expression (4) ′ is the same as the evaluation function F described in the first embodiment with ζ · F ₆ (Q _k , Q _k
^max , Q _k ^a ) are added. Functions F _{1 to} F ₅
The action of is as described in the first embodiment.

Q which is the argument of the sixth term of the evaluation function F ′
_k is the target traffic flow rate set in the above-mentioned upper layer.
Further, Q _k ^max is the upper limit value of the target traffic flow rate set in the above-mentioned upper layer. Furthermore, Q _k ^a is a traffic flow rate of each link is achieved by executing a control pattern obtained in the lower layer.

The function F ₆ increases as the deviation between the achieved traffic flow rate Q _k ^a and the target traffic flow rate Q _k increases. Further, when the upper limit value Q _k ^max is set from the upper layer, the function F ₆ is increased when the difference between the achieved traffic flow rate Q _k ^a and the upper limit value Q _k ^max becomes large. . In this way, the target traffic flow rate set in the upper layer is reflected in the control pattern in the lower layer to achieve the optimal distribution of the traffic flow considering the entire area.

The lower layer control operation is completed.

Coefficients α, β, γ of each term of the evaluation function F ′,
δ, ε and ζ can be changed by the system operator operating the master station 22. In response to the operation of the system operator, the master station 22 instructs the local station 2 provided at the intersection connected to the link to be preferentially run to increase the weighting coefficient α. The waiting times T _D1 and T _D2 generated on the link are reduced, and priority travel is realized. Further, the master station 22 instructs the local station 2 provided at the intersection connected to the link to be restricted to set the weighting coefficient α to 0 according to the operation of the system operator. This prevents traffic from flowing into the link.

Furthermore, the coefficients α, β of each term of the evaluation function F ′,
γ, δ, ε and ζ can be changed according to the occurrence of the accident area. The master station 22
The traffic condition acquired by the sensor 5 at each intersection 2 is monitored over the entire control area. Master station 22
Shows that the amount of traffic that has flowed into an intersection has not increased, but there is still unresolved traffic, and the outflow side of that intersection is green time, but continues for a certain period of time.
When it is detected that the outflow amount of the vehicle is smaller than a predetermined value, it is determined that the intersection is an accident intersection. The master station 22 is a local station 2 provided at an intersection connected to a link flowing into the accident intersection.
To instruct the weighting factor α to be zero. This prevents traffic from flowing into the accident area. Further, the master station 22 separates the accident intersection from the control area, selects the basic intersections in the remaining control areas, and cooperates with the cooperative intersections centering on the basic intersections.

Further, the coefficients α, β of each term of the evaluation function F ′,
[gamma], [delta], [epsilon], and [zeta] can be changed according to the notification of the traveling of the emergency vehicle. Master station 22
Upon receipt of the notification of the traveling of the emergency vehicle, the local station 2 provided at the intersection 1 connected to the link on which the emergency vehicle travels is instructed to increase the weighting coefficient α. Further, the local station 2 provided at the intersection 1 connected to the link in which the emergency vehicle does not travel is instructed to reduce the weighting coefficient α. As a result, the green time of the link on which the emergency vehicle travels becomes longer, and the green time of the link on which the emergency vehicle does not travel becomes shorter, so that the emergency vehicle can run smoothly.

The autonomous distributed signal control system according to the second embodiment can easily expand its control area as in the first embodiment. In this case, steps (1) to expansion of the control area described in the first embodiment
In addition to (8), the setting of the master station 22 is changed. At this time, even if the master station 22 is temporarily stopped due to the setting change of the master station 22, each local station 2 can continue the control by cooperation between the local stations 2. Even if the master station 22 is stopped,
Data exchange between the local stations 2 selects a crossing point. Further, ε, which is the coefficient of the evaluation function F,
And ζ are set to 0, and the master station 22 does not give the control guideline pattern or the target traffic flow rate,
The local station 2 is the master station 22
The control can be continued independently from.

In the autonomous distributed signal control system according to the second embodiment, the control guideline created by the master station 22 is shown in each local station 2. Each local station 2 determines the control pattern of each intersection with reference to the indicated control guideline. At this time,
The basic crossing is controlled according to a control pattern that refers to the control guideline and is determined based on the measured traffic conditions, and the cooperative crossing is controlled while directly or indirectly cooperating with the basic crossing. Therefore, stabilization and optimization of the entire control area is realized.

At this time, a plurality of basic intersections are determined according to the traffic situation, and the control area is divided into sub-areas each including one basic intersection, so that the traffic volume is larger and the optimization is optimized. More optimal control is performed centering on the key axis intersection, which is greatly required.

[0151]

According to the present invention, the signal control device at each intersection can autonomously operate according to the change in the traffic condition of the entire control area without having a large-scale control center facility composed of the central signal control device. Provided is an autonomous distributed signal control system capable of calculating a control pattern and stably and optimally controlling the entire area.

Further, according to the present invention, by automatically calculating a control pattern according to changes in traffic conditions, periodical parameter re-adjustment is unnecessary, and maintenance cost can be greatly reduced. Will be provided.

[Brief description of drawings]

FIG. 1 shows an autonomous distributed signal control system according to a first embodiment of the present invention.

FIG. 2 is a detailed view of the vicinity of an intersection 1.

FIG. 3 shows a logical structure of a control operation performed in the autonomous distributed signal control system according to the first embodiment.

FIG. 4 shows subareas 23 and 24 defined by dividing a control area.

FIG. 5 is a diagram illustrating a method of calculating an intersection load factor λ.

FIG. 6 is a diagram illustrating a method of determining a cooperation destination of a cooperative intersection.

FIG. 7 shows a method of determining a control target pattern at a base-axis intersection in the first embodiment.

FIG. 8 is an inflow point 8 ₁ for which a delay time is calculated.
Show the 8 _20.

FIG. 9 shows a method for calculating the number of inflowing units Q ₁₃ (t) per unit time to the inflow point 8 ₁₃ .

FIG. 10 shows a method of determining a control target pattern of a cooperative intersection in the first embodiment.

Figure 11 is a diagram for explaining a method of calculating delay times of the inflow link and outflow link for cooperative intersection 1 _5.

FIG. 12 shows a method of estimating the number of queues using an optical beacon.

FIG. 13 is a diagram showing a vehicle group having a speed exceeding the speed at an outflow link.
Optimal offset O to prevent overload _R1The calculation method of
You

FIG. 14 is a diagram showing a group of vehicles in the inflow link exceeding the speed.
Optimal offset O to prevent overload _R2The calculation method of
You

FIG. 15 shows a method of expanding a control area of the autonomous distributed signal control system according to the first embodiment.

FIG. 16 shows a second embodiment of an autonomous distributed signal control system according to the present invention.

FIG. 17 shows a logical structure of control operation performed in the autonomous distributed signal control system according to the second embodiment.

FIG. 18 shows a method of determining a control guideline candidate offset for a link forming a closed loop.

FIG. 19 shows subareas 23 and 24 defined by dividing a control area.

FIG. 20 shows a method for determining a control target pattern at a base-axis intersection in the second embodiment.

Figure 21 illustrates a link ₁₁ 2-11 _5, the control pointer offset ^O _T 2 ^{~ O} _{T 5} stipulated in each link ₁₁ 2-11 _5.

FIG. 22 shows a method of determining a control target pattern of a cooperative intersection in the second embodiment.

[Explanation of symbols]

1: Intersection 2: Local station 3, 4: Signal light device 5, 5 ': Detector 7 _{1 to} 7 ₄ , 8 ₁ to 8 ₂₀ , 10 ₁ to 10 ₉ : Inflow point 9 ₁₃ : Outflow point 21: Communication line 22 : Master stations 23, 24: Sub area

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Shigetaka Hosaka             2-1-1 Niihama, Arai-cho, Takasago, Hyogo Prefecture             Takasago Laboratory, Mitsubishi Heavy Industries, Ltd. (72) Inventor Range Hiroyuki             2-1-1 Niihama, Arai-cho, Takasago, Hyogo Prefecture             Takasago Laboratory, Mitsubishi Heavy Industries, Ltd. F term (reference) 5H180 AA01 BB02 BB04 BB15 EE03                       JJ02 JJ06 JJ11 JJ16

Claims (9)

[Claims] 1. A signal light device provided at each intersection of a control area, and a local station provided at each of the intersections and controlling the signal light device, respectively. Each of the local stations is provided by itself. Based on the traffic conditions of the intersection, it is determined whether the intersection where the self is installed is a basic intersection or a cooperative intersection, and, among the local stations, the intersection where the self is installed is determined. Based on the traffic situation in the range including the adjacent intersection adjacent to the intersection where the self is provided, the basic local station that determines that it is the basic intersection self-independently creates a control pattern for controlling the intersection where the self is provided. Then, of the local stations, it is determined that the self-provided intersection is the cooperative intersection. The coordinated local station cooperates with the control pattern of one of the basic intersections to create a control pattern of the cooperative intersection based on the traffic situation of the provided cooperative intersection. Control system. 2. The autonomous distributed signal control system according to claim 1, wherein each of the local stations determines an intersection where the self is provided based on the degree of congestion of the intersection where the self is provided, as the base intersection. And an autonomous decentralized signal control system for determining which of the above-mentioned coordinated intersection is to be used. 3. The autonomous distributed signal control system according to claim 1, wherein the cooperative local station determines an intersection where the intersection provided by itself cooperates based on the congestion degree of the adjacent intersection as the adjacent intersection. Autonomous distributed signal control system to choose from among. 4. A signal light device provided at each intersection of a control area, and a local station provided at each of the intersections and controlling each of the signal light devices, wherein a plurality of the intersections are selected as a basic intersection. The other of the intersections is selected as a cooperative intersection, the control area is divided into a plurality of sub-areas each including one basic intersection, and among the local stations, the intersection where the self is provided is the basic intersection. The core local station, which is an intersection, self-independently creates a control pattern for controlling the intersection where the self is provided, based on traffic conditions in a range including an adjacent intersection adjacent to the intersection where the self is provided, and the local station Among the stations, the cooperative local station whose own intersection is a cooperative intersection is
Autonomous decentralized type that creates a control pattern for a cooperative intersection based on the traffic situation of the coordinated intersections in which each of the basic intersections is included in the same sub-area while cooperating with the control pattern of the basic intersection. Signal control system. 5. The autonomous distributed signal control system according to claim 4, wherein whether the intersection is selected as the basic intersection or the cooperative intersection is determined based on a congestion degree of the intersection. Type signal control system. 6. A step of classifying each intersection of the control area into one of a basic intersection and a cooperative intersection, and a control pattern for controlling the basic intersection, the range including an adjacent intersection adjacent to the basic intersection. And a control pattern for controlling the cooperative intersection, while coordinating with a control pattern for controlling one of the basic intersections based on the traffic situation of the cooperative intersection. A signal control method comprising: a step of creating. 7. A step of classifying each intersection of the control area into one of a basic intersection and a cooperative intersection, the control area being divided into a plurality of sub-areas so that each of the control areas includes one basic intersection. Dividing, a step of creating a control pattern for controlling the basic intersection based on traffic conditions in a range including an adjacent intersection adjacent to the basic intersection, and a control pattern for controlling the cooperative intersection, the basic intersection Of the coordinated intersections, the control pattern of one of the basic intersections included in the same sub-area is created based on the traffic condition of each of the provided coordinated intersections. 8. A step of classifying each intersection in the control area into one of a basic intersection and a cooperative intersection, and a control pattern for controlling the basic intersection, the range including an adjacent intersection adjacent to the basic intersection. And a control pattern for controlling the cooperative intersection, while coordinating with a control pattern for controlling one of the basic intersections based on the traffic situation of the cooperative intersection. A signal control program that executes the steps to create. 9. A step of classifying each intersection of the control area into one of a basic intersection and a cooperative intersection, wherein the control area is divided into a plurality of sub-areas so that each of the control areas includes one basic intersection. Dividing, a step of creating a control pattern for controlling the basic intersection based on traffic conditions in a range including an adjacent intersection adjacent to the basic intersection, a control pattern for controlling the cooperative intersection, the basic intersection Among them, a signal control program for executing the step of creating based on the traffic situation of each of the coordinated intersections, while coordinating with the control pattern of one basic intersection included in the same sub-area as the coordinated intersection. .