JPH06266993A - Highway traffic controller - Google Patents

Abstract

PURPOSE:To obtain a signal control system in which reliability can be improved at the time of preventing the occurrence of traffic jam by predicting traffic volume by imparting a function which mutually communicates information with the other sub-system to the sub-system which predicts the traffic volume and controls a traffic signal. CONSTITUTION:In a sub-system 1 which manages one or several inter-sections, information from a traffic volume measuring device 10 is inputted to an information traffic volume predicting device 11, and the predicted value of the feature traffic volume and occupancy ratio is obtained and outputted to a signal parameter computer 12. The signal parameter computer 12 transmits a preliminarily set signal adjustment pattern to a traffic signal controller 13 by using the transmitted information related with the traffic volume, and the traffic signal controller 13 controls a traffic signal 14. Also, an information storage device 18 records the information measured by the traffic volume measuring device 10, or preserves the information necessary for the traffic volume predicting device 11 and the signal parameter computer 12. The above mentioned devices are connected with a communication equipment 15, and the transmission and reception of the necessary data is performed by communication with the outside system.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traffic control system and apparatus for road traffic, and more particularly to signal control for preventing traffic congestion by predicting traffic volume.

[0002]

2. Description of the Related Art As a conventional traffic control system, Japanese Patent Laid-Open No.
As described in No. 4-51399, there is a method of measuring the traffic volume on the road, predicting the time of occurrence of the traffic jam from the information, and adjusting the traffic volume of the road connected to the traffic jam occurrence prediction point with a signal parameter. In order to control a wide range of traffic, a hierarchical structure is used in which a central center device is connected to a terminal traffic light control device via a sub-center device on the way.

A method of predicting the time of occurrence of the congestion is described in The Institute of Electrical Engineers of Japan, Volume C, 107, No. 10, page 896 (198).
The historical average pattern method described in 7) is used. The past average pattern method is one of the points on the road network.
Focusing on the daily traffic volume (or occupancy rate), taking the average of the past data at that point as a reference time series pattern, and comparing the reference pattern with the time series pattern of traffic volume up to the present time , It seeks future traffic patterns. For example, in order to take into consideration that the traffic patterns change greatly depending on the day of the week,
The data was classified such as for Saturday and for special days, and the standard data was sought for each classification. In addition, the standard data for the past 1 hour and the data for 1 hour measured on the day are summed, the ratio of the two values is calculated, and the standard data for 5 minutes and 10 minutes ahead to be predicted from the ratio is obtained. I was looking for it by multiplying.

[0004]

The conventional system configuration has a hierarchical structure in which the functions are completely separated, and its control method determines the signal parameter from the traffic volume measured at each intersection with reference to past data. Therefore, when the traffic volume increases, the whole traffic flow becomes unsuitable, and there is a problem that the occurrence of traffic congestion cannot be avoided.

Further, since the signal parameter is calculated by the central unit, its processing capacity and the amount of communication with the traffic signal control device at each intersection become enormous, which requires communication amount and communication cost. Further, when an abnormality occurs in the central device, it becomes impossible to control the entire lower system properly, which is a big problem in terms of reliability.

Further, in the conventional example, there is a lack of a viewpoint of considering an environment such as an event such as an exhibition or a phenomenon in which the traffic volume changes greatly such as road construction, and exhaust gas and noise.

[0007] The present invention relates to signal control for preventing traffic congestion by predicting traffic volume and improving reliability.

[0008]

According to the present invention, each unit or a unit for integrally managing several intersections is a subsystem, and each subsystem has a traffic volume measuring unit for measuring a traffic volume and a traffic volume measuring unit. As a subsystem, a road traffic control device including a traffic volume prediction unit that predicts traffic volume, a signal parameter calculation unit that obtains a signal parameter of a traffic signal, a traffic signal control unit that switches the display of the traffic signal, and an information storage unit that stores information , Has a function of connecting a plurality of the subsystems to a common communication path and allowing the subsystems to mutually communicate information with other subsystems by communication means in each of the subsystems for communicating information.

Further, each subsystem has a preliminary function of a main means and a status monitoring means. The status monitoring means has a self-abnormality judging function for judging an abnormality of the subsystem itself and also detects an abnormality occurring in the system. An abnormality detecting means is provided.

In order to manage the traffic demand, the traffic volume that is currently occurring in relation to the event or the traffic volume that is predicted to occur in the future is used as input data for the system. For event traffic, a system is provided to control the traffic volume according to the request from the system. At this time, information for adjusting traffic demand is provided not only to the event site but also to vehicles around the site and points closely related to the site through the information display board and the in-vehicle device. Furthermore, in order to take into account changes in traffic capacity, data such as schedule and scale of road construction will be used as input data to the system.
On the other hand, in order to consider the environment, a means for estimating the exhaust gas amount / noise emitted from the vehicle is provided, and the exhaust gas amount / noise is used as an evaluation index value for control.

[0011]

Each subsystem predicts the traffic volume from the traffic volume predicting means based on the traffic volume flowing into each intersection obtained by the traffic volume measuring means, and from this prediction data, the signal parameter calculating means uses the optimum traffic signal. Optimal traffic control is possible because the signal parameters are determined and the traffic signal display is switched by the traffic signal control means. Note that the traffic volume is predicted by using a neuro computer or a multivariate analysis method.

Further, since each subsystem has a communication means, information is provided one after another up to subsystems in which the upper limit of the passing traffic amount can be given when calculating the signal parameter of each intersection and the constraint condition due to the passing traffic upper limit value can be eliminated. Congestion can be eliminated by controlling so that it will be transmitted. At this time, by storing the data such as the traffic volume received from the other subsystems in the information storage means, it is possible to know the surrounding traffic condition and determine the signal parameter of the intersection managed by the user.

Further, when an abnormality occurs, the presence / absence of an abnormality in the system can be detected by transmitting the information to the abnormality detection means by the self-abnormality determination means of the subsystem or by checking the abnormality detection means, and a substitute backup A highly reliable traffic control system can be realized by operating the functions.

Regarding the management of the traffic demand including the event, the predicted value considering the event traffic can be obtained by adding the data from the event to the normal traffic predicted value. Furthermore, for changes in the environment such as traffic volume, exhaust gas volume, and noise due to construction, etc., it is used at any time when calculating signal parameters.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIG. In the subsystem 1 that manages one or several intersections, the information from the traffic volume measuring device 10 is input to the traffic volume predicting device 11 to obtain the predicted values of future traffic volume and occupancy rate, and the signal parameter calculating device 12 Output to. The signal parameter calculation device 12 sends a preset signal adjustment pattern to the traffic signal control device 13 using the sent information regarding the traffic volume, and the traffic signal control device 13 causes the traffic signal 14
To control. Further, the information storage device 18 is the traffic volume measuring device 1
The information recorded at 0 is recorded, and necessary information is stored in the traffic volume prediction device 11 and the signal parameter calculation device 12.
The above-mentioned devices are connected to the communication device 15, and each of them transmits / receives necessary data through communication with an external system. In the figure, the connections between the devices are drawn using the same means, but it is also possible to connect the individual devices separately.

The traffic volume measuring device 10 uses an ultrasonic wave detector, a sensor using a loop coil, a microwave, or an existing device using an image processing technique by various methods. Further, as the traffic signal 14 and the signal controller 13,
Published by Traffic Engineering Research Society, "Traffic Signal Control Technology",
An apparatus as described in p.12-p.20 (1983) is used. Here, the traffic signal control device 13 controls one or more traffic signals. Furthermore, as a communication method for connecting each device in the figure, published by the Traffic Engineering Research Society,
"Traffic Signal Control Technology," p.22-p.25 (198)
The method and apparatus described in 3) are used.

The traffic volume predicting device 11 is a traffic volume measuring device 10.
In response to the input from, the predicted values of the future traffic volume and occupancy rate, for example, after 10 minutes and 1 hour, are obtained and sent to the signal parameter calculation device 12. In the present embodiment, as the input data from the traffic volume measuring device 10, for example, the passing traffic volume and the occupancy rate every 5 minutes are assumed. However, this time interval may have any value, and the interval may not be constant. The traffic volume prediction unit 16 used in the traffic volume prediction apparatus 11 uses a neuro computer. Although FIG. 1 depicts a configuration in which the traffic volume prediction device 11 and the signal parameter calculation device 12 are separated, they may be built in the traffic signal control device or built in the central device. The information storage device 18 uses a medium such as a magnetic tape, a magnetic disk, an optical disk, or a semiconductor storage device.

The overall system configuration is such that an upper subsystem 81 and one or more lower subsystems 82 are hierarchically connected as shown in FIG.
Only two connected to each other via a network,
Alternatively, several forms are conceivable, such as a mixture of the two. Hierarchical structures such as higher subsystems of higher subsystems and lower subsystems of lower subsystems are also conceivable. Regarding each subsystem constituting the entire system, not only each subsystem is uniform, but subsystems including different elements can be mixed together at the same time.

The traffic forecasting unit 16 using a neuro computer will be described with reference to FIG. The traffic volume prediction unit 16 is composed of a neural network 27, an input switching unit 26, and a learning mechanism 20. The learning mechanism 20 includes a learning control mechanism 24 for controlling each element during learning, an input signal generating mechanism 21, a teacher signal generating mechanism 23, a parameter changing mechanism 22 for changing the weight of the neural network, an output of the neural network 27 and a teacher signal. It comprises a matching section 25 for matching with the teacher signal from the generation mechanism 23. The neural network 27 used here is a Ramel-Hart type neural network composed of an input layer 30, an intermediate layer 31, and an output layer 32 as shown in FIG. 3, and the output of the neuron 33 of each layer has no feedback in its own layer and is input. Information is transmitted in a feedforward manner from the layer 30 to the output layer 32. The operation at this time will be described below.

(1) Recalling Process Recalling process is a process for obtaining output data from input data. The input data of the neural network is input by the input switching unit 26. FIG. 4 shows the configuration of the neuron 33 that constitutes the neural network 27. The input signal x (i) of the neuron 33 is multiplied by the weight value w (i) in the multiplier 41, and the product of the input signal x (i) input to the neuron in the adder 42 and the weight value w (i). Are added (sum of products operation). The result of the product-sum calculation is input to the function converter 43. The function converter 43 uses SIGM
Continuously differentiable function conversion such as the OID function is performed, and the data after the function conversion is output from the neuron 33.

A specific processing operation will be described with reference to FIG. In the figure, an input layer 30 composed of a plurality of neurons 33
Enter the data in. The output of the neuron 33 of the input layer 30 is input to the intermediate layer 31 composed of a plurality of neurons 33 through the product-sum operation, and the output of the neuron 33 of the intermediate layer 31 becomes the input of the neuron 33 of the output layer 32 in the same manner. The output of the neuron 33 of the output layer 32 becomes the final output of the neural network 27. When the neural network 27 is unlearned, the weight w (i) of the neuron 33 is indefinite, so that the output of the output layer 32 is indefinite with respect to the data input to the input layer 30. Therefore, it is necessary to learn and optimize the weight. The learning method for determining the weight of the neuron 33 will be briefly described below.

(2) Learning Process The learning process is a process for determining the weight w (i) of the neuron 33 that constitutes the neural network 27. As input data to the neural network 27 during learning, the input switching unit 26 inputs the data from the input signal generating mechanism 21. As a method of determining this load w (i),
For example, it is described in detail in the second section "Error Back Propagation Learning Method" of the paper "Possibilities of Neurons" described in "Journal of the Institute of Electronics, Information and Communication Engineers", Vol. 71, No. 11, pages 1241 to 1247. ing.

The operation will be briefly described with reference to FIG.
In advance, the input layer 30 of the neural network 27
A pair of data to be input to and data to be output from the output layer 32 for the data is measured. Then, one of the data is input to the input layer 30 from the input signal generating mechanism 21, and the recall process is performed. On the other hand, the data 51 desired to be output corresponding to the input data is generated from the teacher signal generating mechanism 23 and input to the matching section 25.
In the matching section 25, the weight of the output layer 32 is changed so that the difference between the teacher signal 51 and the output 50 of the neural network 27 tends to decrease. After changing the load of the output layer 32, similarly, the load of the intermediate layer 31 is changed.

First, data such as the passing traffic volume and the occupancy rate for each set time is measured. Then, data for the set time interval, for example, data from 8:00 to 11:00 is input to the input layer 30. Furthermore, data such as the day of the week, the weather, and the presence / absence of events are also input as input layer data. When the time-series data of traffic volume and occupancy rate following the data input to the input layer at this time, for example, the data from 11:00 to 12:00 is teacher data, the output of the output layer 32 at that time is unlearned. Since it is not a series of data, the weights of the output layer 32 and the intermediate layer 31 are changed so as to reduce the difference between this output value and the teacher data. By repeating this operation at different times and by repeatedly learning with various data with different dates, if you input some partial time series data and data such as the day of the week and the weather, you will be able to obtain subsequent time series data. .

The learning method is not a method of predicting the data of the following time from the partial data of one day, but the entire data of one day is used as input data, and the data of day of the week, weather, etc. It is also possible to use a method of learning similar past data as teacher data or, as another learning method, a method of feeding back the output data of the output layer to the input layer and considering the correlation in the time series.

Now, a method of predicting the traffic volume at a certain point on the road network from the upstream and downstream traffic volumes using a neural network will be described below with reference to FIG. First, the traffic 1401 in FIG. 14 is the traffic to be predicted. Therefore, traffics 1402, 1404, 1405, and 1403 are traffics upstream of them, and traffics 1406, 14
07 and 1408 are downstream traffic. Then, as the input of the neural network 27, similar to the method of predicting the traffic volume of one point 1401 described above, the traffic volume of a certain set time interval is used.
The data of 2, 1404, 1405, and 1403 are also input. After that, as the teacher data and the learning method, learning can be performed by using the same method as the method of predicting the traffic volume at one point, and the prediction data can be obtained by recalling the network. Also, traffic 14
On the other hand, when the congestion occurs due to traffic congestion, the traffic volume on the downstream side is also affected. Therefore, the data of traffics 1406, 1407, and 1408 on the downstream side are similarly given as inputs to the neural network for prediction. To use.
Further explaining with reference to FIG.
4, 1505 are traffic 1401 and 140 of FIG. 14, respectively.
2,1403, the horizontal axis represents time and the vertical axis represents traffic volume. In this example, 1503, 1504
The data of a certain time width in the data of 1505 is set as past data, and the neural network 1502
Input to the input layer of the
1401 future data 1501 is given. This is learned by staggering the time or staggering the days. If necessary data is input to the neural network 27 for which learning has been completed at a time point at which prediction is desired, the data is recalled inside the network and a predicted value is output.

An example of a method using a model formula created by using a multivariate analysis method as the traffic volume prediction method described above will be described below. First, variables such as daily traffic volume (occupancy rate), day of the week, weather, weekdays / holidays, and presence / absence of events are taken as explanatory variables constituting the model formula. Here, with regard to the traffic volume, as a quantitative value, for example, Kawaguchi Sosho "Introduction to Multivariate Analysis 1", Morikita Publishing Co., 1973, pp. 3 to p. 3
The prediction model formula is obtained by using the method of multiple regression analysis described in No. 0. In addition, all of these explanatory variables are qualitative values, for example, Kawaguchi Tosho "Introduction to Multivariate Analysis 1", Morikita Shuppan, 1973, pp. 3 to p. 30,
The prediction model formula is obtained by using the method of multiple regression analysis described in. In addition, all of these explanatory variables are qualitative values, for example, Kawaguchi Sosho "Introduction to Multivariate Analysis 1", Morikita Publishing Co., 1973, p.93-p.100,
It is also possible to obtain the prediction model formula by using the method of quantification analysis 1 class described in. In addition, Toshiro Haga, Shigeru Hashimoto “Statistical Analysis Program Course 2,
Regression analysis and principal component analysis ", Nikkan Giren Publishing Co., Ltd., 1980,
By using the method of integrating the regression analysis and the quantification theory 1 described in Chapter 5, all the explanatory variables with traffic volume as a quantitative explanatory variable and other data as a qualitative explanatory variable. Can also be incorporated into one predictive model equation. Here, as an explanatory variable of the traffic volume, it is possible to model the variation curve of the entire day by using the terms with a high degree, or by using only the terms with a relatively low degree, a certain part of the day It is also possible to model the variation curve of the time zone. By using the model formula created in this manner and substituting the explanatory variable at the time of prediction, the predicted value can be obtained.

By the method described above, the traffic volume or occupancy rate can be predicted. Hereinafter, a method of determining the signal parameter in the signal parameter calculation device 12 using the predicted value will be described. Signal parameters are described in "Traffic Signal Control Technology," published by Japan Society for Traffic Engineering, p.37-p.8.
1 (1983), but the present invention is different in that a predicted value is used instead of the measured amount. Taking one cross-shaped intersection as an example, a traffic volume measuring device 10 and a traffic volume predicting device 11 are installed on the inflow path to the intersection, and predictive values are received from each. The signal parameter calculation device 12 obtains the degree of saturation from the saturated traffic flow rate, which is the traffic capacity of each inflow path obtained in advance, and this predicted traffic volume. Cycle length and split are calculated from this saturation. Signal control can be realized by sending this parameter to the traffic light control device 13 and changing the display of the traffic light 14 under the control. Thus, by applying the above method at each intersection, more effective signal control becomes possible.

Next, an example using a method different from the above-mentioned signal parameter calculation method will be described. Consider a case where an intersection A and an intersection B, which are different in the number of handled vehicles, are adjacent to each other as shown in FIG. At this time, the number of handled vehicles at the intersection A (that is, the intersection passage capacity) is smaller than that at the intersection B. Then, each intersection shall calculate and apply the signal parameter according to the traffic volume at that time so that there is no delay with respect to the inflow traffic volume. This can be realized by using the above-mentioned conventional method. At the intersection A in FIG. 6 where the traffic volume is increasing now, after adjusting the signal parameters,
It is assumed that the upper limit for handling traffic 60 has been reached. at this point,
Recalculate the signal parameters at intersection B upstream of traffic 60. In other words, as a constraint in this recalculation, the traffic 60 must not exceed the critical value (passing traffic upper limit) of the traffic 60.
The signal parameters are calculated so as to limit the traffic volume of the traffic 61, the traffic 62, and the traffic 63 that affect the. For that purpose, it is necessary to consider the saturation situation of the intersection B at that time, the respective contribution ratios of the inflow traffic volume that influences the increase of the traffic volume of the traffic 60 such as the traffic 61, 62, 63. Then, the parameters are determined based on the knowledge of the signal control. As an example, if the knowledge that “limit traffic in proportion to the margin of traffic handling ability” is given, traffic 6
The traffic volume is restricted by reducing the green time of the traffic light from 1, 62, 63 which has a sufficient traffic handling capacity. In addition, as another example, when "prioritize flows with high traffic volume", the parameters are calculated so that the main flow of traffic, that is, the flow with higher traffic volume is prioritized and the traffic volume with less traffic volume is limited. To do. In addition, as another example, if the knowledge of "prioritize traffic on roads with high importance" is given, national roads are preferred over prefectural roads and prefectural roads over city roads, regardless of the amount of traffic flowing there. The direction to prioritize is determined based on the index of importance given as the attribute of the road, and the parameters are calculated.

Limiting the passing traffic in this manner means waiting for traffic in that direction. And if the delayed traffic does not exceed the processing capacity of the intersection, the problem can be solved only at these two intersections. However, if the traffic in the restricted direction is close to the critical value of the road, the restriction may cause the intersection to go beyond saturation. In this case, the signal parameter of the intersection further upstream is similarly recalculated.

This will be described with reference to FIG.
When reaches the upper limit of the passing traffic volume, the parameter for limiting the traffic volume of the traffics 61, 62, 63 at the intersection B is recalculated. As a result, traffic 63 is restricted. At this time, if the traffic 63 has reached the upper limit of the passing traffic volume, with respect to the intersection C, which is the intersection on the upstream side,
Similarly, the signal parameters are recalculated. In this way, the control using the upper limit of passing traffic is successively propagated to the intersections on the upstream side, whereby control for preventing the occurrence of traffic congestion can be realized.

At this time, there may be a request to recalculate signal parameters from a plurality of directions at a certain intersection at the same time. If they contradict each other, if the policy of "priority of main traffic flow" is given as knowledge of signal control at the intersection, the problem can be solved by prioritizing the request from the direction of heavy traffic flow. Can be achieved.

Up to now, only the above-mentioned upper limit value of passing traffic has been considered as a constraint condition when calculating the signal parameter at an intersection, but other than that, information such as the presence or absence of an accident or event is also a constraint. It can be used as a condition. If an accident occurs downstream of the intersection,
Until it is processed, the point cannot be passed, or the amount of traffic passing through is extremely reduced. Therefore, as a constraint, the parameters are calculated to reduce the traffic to the point where the accident occurred as much as possible. When an event occurs downstream of the intersection, it can be regarded as a parking lot that absorbs or discharges traffic. Therefore, more traffic can be passed when absorbing, and passing traffic must be reduced when discharging, but the signal parameters are calculated as a constraint at the intersection by giving the absorbed emission. be able to.

Next, another embodiment will be described with reference to FIG.
In the own subsystem 100, the traffic amount measuring device 10, the traffic amount predicting device 11, the traffic signal control device 13, the traffic signal 14, the communication device 15 and the information storage device 18 are used.
8 or the signal parameter calculation device 12 to the communication device 15
Through the other subsystem 200. By adopting this configuration, the signal parameters can be intensively calculated by the other subsystems 200, and there is no need to communicate the upper limit value of passing traffic. It should be noted that the information storage devices 18 of the own subsystem 100 and the information storage devices 18 of the other subsystems 200 do not have to be two at the same time.

Another embodiment will be described with reference to FIG. A configuration is also possible in which the information storage device 18, the traffic volume prediction device 11, and the signal parameter calculation device 12 are realized in another subsystem 300 via the communication device 15. By adopting this configuration, it becomes possible to perform complicated calculation processing collectively in another system, and to apply sensing and control in a system near the intersection. Further, the number of information storage devices 18 of the own subsystem 300 and the information storage devices 18 of other subsystems need not be two at the same time.

Another embodiment will be described below. In this embodiment, as shown in FIG. 11, the control command device 111 and the input reception device 1 are added to the configuration of the subsystem 1 of FIG.
12 is included. Input reception device 112
Can be realized by, for example, a keyboard, a mouse, a CRT or the like used in a general computer,
It has the function of receiving commands from the operator. The control command device 111 interprets a command input from the input reception device 112, changes internal parameters, outputs stored information, and the like to devices in other subsystems via the own subsystem or the communication device 15. The request is sent.

For example, through the input reception device 112, the operator selects a certain process from a plurality of processes prepared in advance. Then, in the control command device 112, the value of the parameter is changed to the signal control device 13 in the subsystem 1 based on the information stored in the program stored in the information storage device 18 or the information about the selected processing. To issue a command or to measure the traffic volume 10
It issues a command to output the measured information.

Here, the entire system can be hierarchically constructed by using the subsystem 1 having this function as an upper subsystem. On the other hand, there may be a plurality of subsystems having this function in the entire system.

Another embodiment will be described below. In this embodiment, as shown in FIG. 12, the state monitoring device 121 is included in the configuration of the subsystem 1 of FIG. The state monitoring device 121 can be realized as an infinite loop process in a general computer and has a function of monitoring the system state. Specifically, the status monitoring device 121
Sends a signal to each device that configures its own subsystem and other subsystems asking whether or not they are operating normally. Upon receiving the signal, each device replies the current status. If normal, a normal signal is sent to the state monitoring device 121, and if abnormal, a signal indicating the content is sent to the state monitoring device 121. The state monitoring device 121 interprets the transmitted signal and does nothing if there is no abnormality. If an error is returned from any of the devices, an alarm to that effect is issued, and at the same time, a command is issued to substitute the function of the device with the error for another subsystem.

This will be described with reference to FIG. Several subsystems are connected via a communication path 2000. The subsystem A131 uses the traffic volume data predicted by the traffic volume prediction apparatus 1312 as the signal parameter calculation apparatus 131.
Send to 3. The signal parameter calculator 1313 uses the data to calculate the signal parameter and sends the result to the signal controller 1314. This is subsystem B
Similar processing is performed inside. Then, in the subsystem C133, the state monitoring device 1332 is supposed to monitor the state of the entire system.

Here, consider a case where an abnormality occurs in the signal parameter calculation device 1313 of the subsystem A 131. The abnormality is detected by the status monitoring device 1 in the subsystem C133 via the communication device 1311 and the communication device 1331.
Detected at 332. Therefore, the state monitoring device 13
At 32, a command is issued to the signal parameter calculation device 1323 of the subsystem B132 so that the subsystem B132 can perform the function. Signal parameter calculation device 1
323 receives the command, receives the information from the traffic volume predicting device 1312 of the subsystem A131 via the communication device 1321, and calculates the signal parameter in order to substitute the function of the signal parameter calculating device 1313 in which the abnormality has occurred. , And activates a process for returning the result to the traffic light control device 1314 of the subsystem A 131 via the communication device 1321.

In this way, when the receiving side is ready, the status monitoring device 1332 outputs the output destination to the traffic volume predicting device 1312 of the subsystem A 131 via the communication device 1311 to the signal parameter calculating device of the subsystem B 132. 1323 is issued to the traffic signal control device 1314 of the subsystem A 131, and the communication device 1
Command to receive output from subsystem B 132 via 311. As a result, it is possible to realize a mechanism for substituting the function of the device in which an abnormality has occurred in one subsystem for another subsystem.

By using the method described above, it is possible to quickly and automatically take emergency measures when an abnormality occurs. Here, as for the method of monitoring the state, as described above, each device may voluntarily report the state without issuing a request from the state monitoring device 1332. In that case, it can be realized by using either method of sending the presence / absence of abnormality periodically or sending only when an abnormality occurs. On the other hand, there may be a plurality of subsystems C having this function.

Further, another method of substituting for functions when an abnormality occurs in the traffic volume measuring device 1315 of FIG. 13 will be described below. If the traffic volume measuring device 1315 is abnormal, the measuring function itself cannot be delegated to another subsystem unless there is a spare device there.
However, it is possible to complement that function well in the whole. As the method, there is a method of estimating the current traffic volume based on the past measurement data stored in the information storage device 1316. This is the traffic forecasting device 13
12 can be used as it is. That is, the state monitoring device 1332 that detects the abnormality of the traffic volume measuring device 1315 is
The traffic volume prediction apparatus 1312 is instructed to switch the prediction method from the method of predicting using data measured online to the method of predicting using past data. By doing so, the traffic volume measuring device 1
Even if an error occurs in 315 and data is not correctly sent to the traffic volume prediction apparatus 1312, the traffic volume prediction apparatus 1312 supplements its function, so that the entire system can be operated without stopping. .

An embodiment of the system for managing traffic demand will be described below with reference to FIG. The event management device 1601 is provided in a parking lot management center or the like at the event site, outputs the current value and predicted value of the traffic volume generated in association with the event, and controls the generated traffic volume from the outside. input. The predicted value can be obtained by creating a model from past data, or can be set by the operator based on experience. Further, the information providing device 1603 provides the driver with detour information such as an information display board on the road or an in-vehicle device.

On the other hand, the demand management amount calculation device 1606 receives the predicted traffic volume as an input and handles it by signal control.
If you want to control event traffic and suppress demand, present a detour to the driver to change demand, or combine them, and then control event traffic or normal traffic, Outputs the control amount and detour. For example, the upper limit traffic volume that can be handled by signal control is set as a boundary, and if it is smaller than that, all is covered by signal control, and when it is more than that, the large amount is shared by the set ratio of event traffic and normal traffic and suppressed.

The overall flow will be described with reference to FIG. First, from the event management device 1601 and the traffic volume measuring device 1602,
The current value / predicted value and the traffic volume are input to the traffic volume prediction device 1604. The traffic volume prediction device 1604 obtains a predicted value in consideration of event traffic, and passes it to the signal parameter calculation device 1605. At this time, the demand management amount calculation device 16
According to 06, if the predicted value is within the range covered by signal control, the signal is changed based on the calculated signal parameter.

If the signal control exceeds the corresponding range, in addition to the signal control, the excess amount is divided into the event traffic suppression amount and the normal traffic suppression amount. Event management device 16
In 01, the number of vehicles leaving the car is limited by the control value. on the other hand,
For indirect control of indicating a detour for normal traffic, the change in traffic demand is considered by inputting data such as the estimated number of vehicles that can be controlled by the detour information or the ratio to the traffic to the traffic volume prediction device 1604. To do.

Although the embodiment of the system for managing the traffic demand has been described above, it is also possible to predict the future situation further by using a simulator online for this and to execute the control based on the information. Is. That is, a simulation is executed based on past data and data input online up to that time, and measures are taken in advance based on the predicted traffic conditions. As a result, there is no time delay and effective countermeasures can be taken. By using the simulator offline, it is possible to evaluate the event operation plan in advance.

Next, with reference to FIG. 17, a method of considering a change in road capacity will be described. In FIG.
Similar to traffic volume measuring device 1602, information providing device 160
3. The traffic volume prediction device 1604 and the signal parameter calculation device 1605 use the above-described method. Construction management device 17
01 is a function that outputs information about the road capacity that changes with the set construction and its time zone, and that receives the regulation amount as an input. The road capacity management amount calculation 1706 has a function similar to the demand management amount calculation 1606 of FIG. 16, and the predicted traffic volume is input, and the traffic volume is reduced by controlling the scale of construction or reducing it by signal control. Measures such as increasing the capacity, presenting a detour to the driver to change the demand, or combining them, etc. are taken, and based on the result, the scale of construction can be suppressed In the case of controlling the normal traffic, the detour is output. For example, considering that the road capacity has been reduced due to construction, the upper limit traffic volume that can be handled by signal control is set as a boundary, and if it is less than the upper limit traffic volume, it will be covered entirely by signal control. Measures will be shared by the set ratios using the respective methods of capacity increase due to and traffic reduction due to normal traffic control. At this time, if there are construction regulations that cannot be dealt with immediately, all measures will be taken by normal traffic control. Note that the overall flow of FIG. 17 is the same as that described with reference to FIG. Further, although road construction has been taken as an example here, this can be similarly applied to an event that changes road capacity, such as a road event.

The method for managing traffic demand and the method for considering changes in road capacity have been separately described above with reference to FIGS. 16 and 17, but these two methods are the demand management amount calculation device 1606 and the road capacity management. If the two functions of the quantity calculation device 1706 are unified, they can be realized at the same time.

Here, a traffic control method considering the environment will be described below. First, various average speeds of an automobile, exhaust gas amount and noise level at that time are measured for each vehicle type, and are grasped as a relationship between speed and exhaust gas and a relationship between speed and noise level, respectively. Then, sensors for measuring traffic volume, speed, and vehicle type using image processing and microwaves are installed on the road to measure the current traffic situation. If the vehicle type cannot be measured, it is also possible to estimate the number of vehicles of each vehicle type by using data such as a large vehicle mixture ratio measured in advance. Using the data obtained at this stage, the exhaust gas amount and noise at that time can be estimated from the relationship between the speed and the exhaust gas / noise described above. However, this is only the resulting value. So, to take these data into account,
Several control methods with different parameters and methods are prepared, and the predicted value of the measured traffic data is given as an input to obtain the exhaust gas amount and noise respectively. This can be achieved by using a simulator. That is, the simulator has the passing traffic volume and the control algorithm, and the traffic volume / average speed for each vehicle type can be obtained by actually simulating. Then, the exhaust gas amount and noise are obtained from the data. For example, as an evaluation function, we use weighted sums of congestion length, exhaust gas, and noise, and add them together, and substitute the values obtained for each method into that evaluation function. Select the method with the smaller value and apply it. As the evaluation function, not only the above-described linear sum form but also various variations such as multiplication and logarithm can be considered. Although the method described above uses the simulation online, the control method can be evaluated in advance by using the simulator with the past traffic volume data as an input and offline.

[0053]

By controlling several subsystems as a whole, the processing load can be dispersed, and more reliable and higher-level control can be realized. In addition, since the traffic volume can be predicted with higher accuracy, it is possible to perform signal control on the first move and the first move, so that appropriate traffic control can be realized. Furthermore, when an intersection becomes saturated with the upper limit of the passing traffic as a constraint, signal parameters of the upstream intersections are recalculated, and each intersection is given knowledge about signal control to prevent congestion. It is possible to control signals as much as possible. In addition, by considering and controlling factors that greatly affect actual traffic such as events and construction, it is possible to build a comprehensive integrated road traffic control system from demand management to appropriate control. Furthermore, by considering the environmental impact such as exhaust gas and noise in the control, not only the smoothness of traffic but also the sophistication of the road system as a social system can be achieved.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a subsystem that constitutes a traffic control system.

FIG. 2 is a block diagram of a neuro computer used as a method of predicting traffic volume.

FIG. 3 is an explanatory diagram showing a configuration of a neural network in the neurocomputer.

FIG. 4 is an explanatory diagram of a configuration of neurons forming a neural network.

FIG. 5 is an explanatory diagram of a learning method of a neuro computer.

FIG. 6 is a diagram showing an intersection and a traffic flow when explaining a signal control method.

FIG. 7 is a diagram showing an intersection and a traffic flow when explaining a signal control method.

FIG. 8 is an explanatory diagram when a system is hierarchically configured by using subsystems.

FIG. 9 is an explanatory diagram when the functions of the subsystem are distributed to some subsystems.

FIG. 10 is an explanatory diagram when the functions of the subsystems are distributed to some subsystems.

FIG. 11 is an explanatory diagram when the subsystem has a function of receiving an input from an operator.

FIG. 12 is an explanatory diagram when the subsystem has a function of monitoring the state.

FIG. 13 is an explanatory diagram when the subsystem has a function of monitoring the state.

FIG. 14 is a diagram showing an intersection and a traffic flow when explaining a traffic volume prediction method.

FIG. 15 is a diagram showing inputs and outputs when explaining a traffic volume prediction method using a neural network.

FIG. 16 is an explanatory diagram of a method of managing traffic demand.

FIG. 17 is an explanatory diagram of a method that considers a change in road capacity.

[Explanation of symbols]

1 ... Subsystem 10,1602,1604 ... Traffic amount measuring device, 11,1312,1322 ... Traffic amount predicting device, 1
2, 1323, 1605 ... Signal parameter calculation / calculation device, 13, 1314, 1324 ... Traffic light control device, 14
... Traffic light, 15, 17, 1311, 1321, 1331
... communication device, 16 ... traffic volume prediction system, 18 ... information storage device, 20 ... learning mechanism, 21 ... input generating mechanism, 22 ... parameter changing mechanism, 23 ... teacher signal generating mechanism, 24 ... learning control mechanism, 25 ... matching Section, 26 ... input switching section, 27, 1502 ... neural network, 30 ...
Input layer, 31 ... Intermediate layer, 32 ... Output layer, 33 ... Neuron, 41 ... Weight, 42 ... Product-sum operator, 43 ... Function converter, 50 ... Neural network output, 51 ... Teacher data, 60 ... Intersection B From the road to intersection A, 61 ...
Incoming traffic to intersection B, 62, 63 ... Incoming traffic to intersection B, 81 ... Upper subsystem, 82 ... Lower subsystem, 111 ... Control command device, 112 ... Input acceptance device, 1
21 ... Status monitoring device, 131 ... Subsystem A, 132
... Subsystem B, 133 ... Subsystem C, 1313
... Signal parameter calculation device, 1332 ... Condition monitoring device,
1401, 1402, 1403, 1404, 1405
1406, 1407, 1408 ... Transportation, 1501, 15
03, 1504, 1505 ... Traffic volume data, 1601 ...
Event management device, 1603 ... Information providing device, 1606
... demand management amount calculation device, 1701 ... construction management device, 1706
… Road capacity management amount calculation device.

Claims (14)

[Claims] 1. A traffic volume measuring means for measuring traffic volume information relating to a traffic volume on a road, a traffic volume predicting means for predicting a traffic volume based on the traffic volume information from the traffic volume measuring means, and the traffic volume prediction. Signal parameter calculation means for obtaining a signal parameter of a traffic signal at an intersection from the means, traffic signal control means for controlling display switching of the traffic signal based on information from the signal parameter calculation means, and information storage means for storing the traffic information A subsystem is a road traffic control device having a plurality of subsystems, a plurality of the subsystems are connected to a common communication path, and information is communicated between the subsystems by communication means in each of the subsystems, thereby receiving from other subsystems. The traffic volume information stored in the information storage means, and the traffic volume information of its own subsystem and the traffic of the other subsystem. Road traffic control apparatus characterized by determining a signal parameter of said intersection from at least one of the traffic information of the information. 2. The road traffic control device according to claim 1, wherein a plurality of subsystems communicating with each other is a lower subsystem, and one or more subsystems are upper subsystems to form a hierarchical structure. Road traffic control device. 3. The road traffic control device according to claim 1, wherein an input means for inputting an operation command is provided in the subsystem. 4. The road traffic control device according to claim 1, wherein the subsystem is provided with a status monitoring unit including a self-abnormality determination unit and an abnormality detection unit, and the self-abnormality of the subsystem is determined and the abnormality is detected. A road traffic control device comprising: a status monitoring means for transmitting a self-abnormality determination result to the detection means, or for monitoring the status of the system to detect the presence or absence of an abnormality in the system, and outputting the monitoring result. 5. The road traffic control device according to claim 4, wherein when the status monitoring means detects an abnormality of a certain means in the certain subsystem, another abnormality detection means is the same as the other abnormality generation means. A road traffic control method characterized in that it is also used as a means in the system. 6. The road traffic control device according to claim 1, wherein a neurocomputer or a multivariate analysis is used for the traffic volume predicting means, and the traffic volume and day attributes are input to the traffic volume predicting means. A road traffic control device characterized by using at least one of information such as weather, presence / absence of event, and presence / absence of accident. 7. The road traffic control device according to claim 1, wherein the signal parameter calculation means calculates using at least information about an accident or an event around an intersection. 8. The road traffic control device according to claim 1, wherein an upper limit value of the passing traffic volume between intersections at a certain time point is determined, and the signal parameter calculation means calculates using the upper limit value. Road traffic control device. 9. The road traffic control device according to claim 1, wherein the signal parameter calculation means obtains a traffic clearance amount at each of the intersections, and the traffic clearance amount matches a road capacity of the intersection, or When the value is close to the road capacity, the road traffic control device causes the intersection near the intersection to recalculate the signal parameter so as to suppress the inflow of traffic to the intersection that has reached the road capacity. 10. The road traffic control device according to claim 1, wherein an upper limit value of the traffic handling amount is obtained by the signal control,
A road traffic control device comprising means for controlling an event that changes a road capacity when traffic exceeding it occurs. 11. The road traffic control device according to claim 1, wherein the traffic flow predicting means substitutes for the traffic flow measuring means when an abnormality occurs in the traffic flow measuring means. apparatus. 12. The road traffic control device according to claim 1, wherein the input information of the traffic volume predicting means is at least based on a change in traffic volume from a parking lot or information to a road information display board or an in-vehicle device. A road traffic control device characterized by using information for changing traffic volume or road capacity, exhaust gas volume, and noise. 13. A traffic volume measuring means for measuring information on traffic volume on a road, a traffic volume predicting means for predicting traffic volume based on information from the traffic volume measuring means, and a signal from a traffic signal from the traffic volume predicting means. Among the signal parameter calculation means for obtaining parameters, the traffic signal control means for controlling the switching of the traffic signal display based on the information from the signal parameter calculation means, and the information storage means for storing information, the traffic volume measuring means and the traffic volume The prediction means, the traffic light control means, the traffic light, the information storage means, and the communication means are their own subsystems, the signal parameter calculation means and the communication means are other subsystems, and the other subsystems and a plurality of the above By connecting its own subsystem to a common communication path, the own subsystem can share other subsystems via the communication means. Road traffic control device, characterized by. 14. A traffic volume measuring means for measuring information on traffic volume on a road, a traffic volume predicting means for predicting traffic volume based on the information from the traffic volume measuring means, and a signal from a traffic signal from the traffic volume predicting means. Of the signal parameter calculation means for obtaining parameters, the traffic signal control means for controlling the switching of the traffic signal display based on the information from the signal parameter calculation means, and the information storage means for storing information, the traffic volume measurement means and the traffic signal control Means, the traffic light, the information storage means, the communication means, and its own subsystem,
The signal parameter calculation means, the traffic volume prediction means, and the communication means are other subsystems, and the other subsystems and a plurality of the own subsystems are connected to a common communication path, and the own subsystem is the A road traffic control device characterized in that the other subsystem is shared via a communication means.