JPS5827555B2 - Traffic information measurement system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a traffic information measurement system that estimates the traffic condition within a certain section of an expressway based on data obtained from passing vehicle number and passing speed measuring instruments installed at both end points of the section. The system is characterized in that the number of passing vehicles, speed harmonization, speed sum, speed square sum, etc. are stored in a register attached to the measuring instrument, and these data are processed by a central data processing unit.

Conventionally, in order to estimate the traffic condition within a section, it has been common to send passing information and speed information from both ends of the section to the center.

For example, the current traffic condition estimation system based on Nahi, which is often used, is shown in FIG.

In this system, traffic flow within a section is expressed by traffic density and spatial average speed, and the traffic volume per unit time and time average speed or representative speed measured at both ends of the section are used to estimate this.

In order to estimate the density based on this information, an extended Kalman filter is applied to the system equation and observation equation shown below.

q2(8)-U(6)ρ(6)+γ(R)
■Here, ρ: Traffic density U: Space average speed ql: Traffic volume V: Time average speed T: Unit time length 1,: Section length α Two constants η, φ, γ, Irregular variables In addition, ■～■ Formulas The subscripts 1 and 2 represent the upstream and downstream end points of the section, respectively.

As is clear from this equation, this state estimation system is nonlinear, and reliable convergence is not guaranteed.

Particularly when the traffic density is low, the denominators of the second and third terms in equation (2) become close to zero and may diverge.

FIG. 2 shows this example using actual expressway data.

On the other hand, the relationship q-ρU is used in the traffic density formula (2) of the system equation, but this relationship does not strictly hold and has a large error, resulting in poor estimation accuracy.

Furthermore, the calculation method of the nonlinear Kalman filter has drawbacks such as being complicated and requiring time for data processing.

The present invention aims to eliminate the above-mentioned drawbacks, provide a system that can easily and accurately estimate the state within an interval, and further guarantees the convergence of the estimation calculation.

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 3 is a block diagram of an expressway traffic flow monitoring system according to the present invention.

This system consists of two passing speed detectors (e.g. radar speedometers) A1 and A2 installed at both ends of the road section, and a terminal data processing device BI connected to these detectors.

B2, and a data transmission system C1 that transmits the output data of these terminal data processing devices to a central data processing device.
Further, the output data from the terminal data processing device sent by the data transmission system C is centrally processed and removed from the central data processing device for obtaining the traffic condition of a desired road section.

It is assumed that the detectors A and A2 are installed at two points on the expressway separated by a certain distance, and can respectively measure the passing of a vehicle and the speed of the passing vehicle.

During a certain time interval (usually between timing signals sent from a central data processing unit), the number of vehicles passed, speed harmonization, speed sum, square sum of speed, etc. are updated every time a vehicle passes, and these are stored in a register. Function terminal data processing device B
1. B2 has it.

Here, when data is sent to the central data processing unit at a certain time interval, the contents of the register are cleared, and the number of vehicles, speed harmony, speed sum, square sum of speed, etc. are accumulated again until the next transmission time. do.

These terminal data processing devices B1. The output of B2 is sent through data transmission system C to the central data processing unit.

In the central data processing unit, these terminal data processing units B
Using the data of 1 and B2, the traffic condition within the section at each time (the time of the normal timing signal), that is, traffic density, traffic momentum, traffic energy, spatial average speed, spatial speed variance, etc. is estimated.

We will briefly describe a method for estimating traffic density, traffic momentum, and traffic energy using the number of vehicles, speed harmonic speed sum, and speed square sum sent every unit time.

Use the following frequencies. ρ: Traffic density, number of vehicles present in a unit distance m: Traffic momentum, sum of speeds of vehicles present in a unit distance e: Traffic energy, sum of squares of vehicle speeds present in a unit distance S: speed harmony, unit Sum of reciprocals of the speed of vehicles passing in a time q: traffic volume, number of vehicles passing in a unit time p:: sum of degrees, sum of speeds of vehicles passing in a unit time γ: sum of speeds and speeds of vehicles passing in a unit time The sum of the squares of the velocity of
Then, the traffic condition within the section is expressed by the following system equation.

Furthermore, the following relationship holds between these internal states and the measured quantities at a point.

E (s)−ρ ■E
(q)-E(m) ■E
(p) −E(e) ■
■~■ are observation equations, and E (x) represents the expected value of X.

Of ■ to ■, E (q): E (!11) corresponds to formula (■) and is well known in the past.

On the other hand, ■ and ■ can be easily derived as follows.

Let the density within the entire interval be ρ, and the speed distribution function be fs(v).

The traffic density with speed V and the traffic volume are ρV and ρ, respectively.
If V, the following equation holds true.

ρ■=ρfs(v) [F
] qu=vpv @Using these formulas, the expected values of m and e are as follows.

On the other hand, S The expected value of p is as follows.

Therefore, it was found that ■~■ appear more easily than @, 0, and 0~[existence].

If ■~■ and ■~■ are respectively system equations and observation equations, the internal state can be estimated by using a linear Kalman filter.

This formula is shown below.

Here, F is the karma gain, and since it is a linear system, it is a time-invariant constant matrix.

This estimation system is stable, has good accuracy, and requires little calculation.

A comparison of this estimation with the estimation of Na1i described above is shown in FIG. 4 using the traffic density ρ as an example.

This measured data was actually taken on the expressway, and the actual traffic density was obtained by taking pictures of the section every minute.

This shows that this method is superior to the Na1i estimation method in terms of accuracy as well.

Here, device B1. The traffic volume, speed harmony, speed sum, and square sum of speed were stored in the register of B2, but instead of this, the traffic volume, speed harmony, speed, and the average of the square profit were stored.
It is also possible to reproduce this again in the central data processing unit as traffic volume, speed harmonization, speed sum and speed sum square.

This may be more preferable in terms of register capacity accuracy.

Based on the estimated value of traffic conditions within the section obtained here,
You can see the interrelationship of this.

The fifth example shows an example of plotting the estimated values of traffic density and traffic momentum.
As shown in the figure.

This represents road characteristics. Therefore, it is possible to measure road characteristics online.

On the other hand, the number of vehicles passing through the inflow ramp can be controlled by signals, gates, etc. so that the traffic conditions, traffic density, traffic momentum, traffic energy, etc. obtained here are kept constant.

As described above, according to the present invention, by using a simple arithmetic function and a storage function in the terminal data processing device, the time interval of data transmission to the solid device can be increased, and the traffic condition inside the target section can be can be reliably and accurately estimated using simple measurement data.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a conventional expressway traffic flow system, FIG. 2 is a diagram showing an example of estimated value divergence by a conventional traffic density estimating device, and FIG. 3 is a diagram showing an example of an expressway traffic flow monitoring system according to the present invention. FIG. 4 is a diagram showing a comparison of estimation accuracy between the present invention and a conventional system, and FIG. 5 is a diagram showing an example in which the present invention is used for detecting road characteristics. In the figure, A1. A2 is a vehicle, Bit BQ is a terminal data processing device, C is a data transmission system, and D is a central data processing device.

Claims (1)

[Claims] 1. A pair of measuring instruments installed at two predetermined points on the road to detect the passing of vehicles and their passing speed, and information on the sum of the reciprocals of vehicle speeds per unit time based on the output of the measuring instruments and the number of passing vehicles. A pair of terminal data processing devices that each accumulate information on the sum of A traffic information measurement system equipped with a central data processing unit.