CN214704951U

CN214704951U - Intelligent signal lamp based on VMD-LSTM automatic time length adjustment

Info

Publication number: CN214704951U
Application number: CN202121138519.8U
Authority: CN
Inventors: 卢海鹏; 张凯; 张龄允; 丁昱杰; 刘洪�; 龚文浩; 谈昊然
Original assignee: Nanjing University of Information Science and Technology
Current assignee: Nanjing University of Information Science and Technology
Priority date: 2021-05-25
Filing date: 2021-05-25
Publication date: 2021-11-12
Anticipated expiration: 2031-05-25

Abstract

The utility model discloses a long wisdom signal lamp when based on VMD-LSTM automatically regulated, including connecting rod, dead lever and base, the connecting rod is connected at the dead lever top, is equipped with red green signal lamp and time-recorder on the connecting rod, be equipped with the camera on the connecting rod, the inside FPGA chip that is equipped with of connecting rod, the FPGA chip is connected with camera, red green signal lamp and time-recorder respectively. The VMD-LSTM predicted traffic flow model adopted by the intelligent traffic signal system not only can more effectively process the nonlinearity and randomness of traffic flow, but also shows excellent capability in explaining long-time problems of time series prediction and has high fitting degree. The wisdom signal lamp can effectually alleviate traffic jam, has improved the current efficiency of vehicle, through the effective dispatch to the traffic flow for the urban traffic flow keeps at a stable running state, has avoided the problem of traffic jam, shortens people's trip time.

Description

Intelligent signal lamp based on VMD-LSTM automatic time length adjustment

Technical Field

The utility model relates to a traffic lights especially relates to a long wisdom signal lamp based on VMD-LSTM automatically regulated.

Background

In the coming years, with the development of Chinese economy, urban population and vehicles in China increase year by year, but at the same time, traffic jam also becomes a big problem. Accurate and real-time traffic flow prediction can be used for analyzing road planning, reducing the traffic accident rate and providing reasonable travel opinions. The traffic flow of different intersections can be predicted, so that the result can be obtained as accurately as possible, a scientific and reasonable scheme is worked out according to the result, accidents are reduced, and traffic pressure is relieved. The traditional traffic signal lamp is composed of a red light, a green light and a yellow light, the set time is long, but the traffic is always changeable, particularly when people go to work or leave work, the traffic signal lamp can not meet the requirements, and the traveling efficiency is reduced.

The traditional traffic signal lamp is composed of a red light, a green light and a yellow light, the set time is long, but the traffic is always changeable, particularly when people go to work or leave work, the traffic signal lamp can not meet the requirements, and the travel efficiency is reduced.

If the traffic flow is more in a certain driving direction of the crossroad, the traffic signal lamp of the crossroad adopts a fixed time length mode, and at the moment, the passing vehicles in the direction are difficult to pass smoothly within the specified passing time, so that a part of the vehicles stop at the crossroad, and the traffic jam is caused. Therefore, in the road section with a large traffic flow, the traffic police are required to conduct manual command, and the method is relatively backward, consumes manpower and has low efficiency.

SUMMERY OF THE UTILITY MODEL

Utility model purpose: the utility model aims at providing a long wisdom signal lamp is recalled network automatically regulated based on variational modal decomposition and length.

The technical scheme is as follows: the utility model discloses a long wisdom signal lamp when based on VMD-LSTM automatically regulated, including connecting rod, dead lever and base, the connecting rod is connected at the dead lever top, is equipped with red green signal lamp and time-recorder on the connecting rod, be equipped with the camera on the connecting rod, the inside FPGA chip that is equipped with of connecting rod, the FPGA chip is connected with camera, red green signal lamp and time-recorder respectively.

When a vehicle passes through the intersection, the special camera 1 identifies the vehicle, the FPGA chip 2 judges whether the vehicle is the special vehicle and processes the special vehicle, wherein the special camera 1 can also record the number of the vehicles passing through the intersection, the license plate number of the vehicles and other information. The short-time traffic flow prediction unit, the central information processing control unit, the corresponding signal lamp control scheme and the road signal lamp control unit of the intelligent signal lamp are all realized through a chip FPGA.

Furthermore, a solar cell panel is arranged on the upper surface of the connecting rod.

Further, the camera is a camera with a vehicle model detection function of PT-DV5120 of Winton technologies, Inc.

Further, the red and green signal lamps are LED red and green signal lamps.

Further, the timer is an LED timer.

Has the advantages that: compared with the prior art, the utility model has the advantages of it is as follows showing:

(1) the VMD-LSTM predicted traffic flow model adopted by the intelligent traffic signal system not only can more effectively process the nonlinearity and randomness of traffic flow, but also shows excellent capability in explaining long-time problems of time series prediction and has high fitting degree.

(2) Wisdom traffic signal lamp has realized the automated control to traffic signal lamp through combining internet technology for the signal lamp changes along with the change of time road conditions condition, even also made outstanding solution to complicated traffic conditions, has saved a large amount of manpowers, has reduced the crowded condition of road. The wisdom signal lamp can effectually alleviate traffic jam, has improved the current efficiency of vehicle, through the effective dispatch to the traffic flow for the urban traffic flow keeps at a stable running state, has avoided the problem of traffic jam, shortens people's trip time.

Drawings

FIG. 1 is a diagram of a smart traffic light control system;

FIG. 2 is a short-term traffic flow prediction flow diagram;

FIG. 3 is a special vehicle processing flow diagram;

fig. 4 is a schematic diagram of a smart signal lamp.

Detailed Description

The technical solution of the present invention will be further explained with reference to the accompanying drawings.

The utility model discloses a special camera is equipped with to the outside, and inside is wisdom traffic signal lamp system. The intelligent traffic signal lamp system inputs traffic flow data collected by the conventional intersection into a short-time traffic flow prediction unit, outputs a predicted value, inputs the predicted value into a central information processing control unit, integrates processing information by the central signal processing control unit and outputs a corresponding signal lamp control scheme according to a source program, and finally controls the duration of a traffic signal lamp by a road signal lamp control unit to achieve the aim of adjusting the traffic flow of a road in real time. The whole intelligent traffic signal system is in closed-loop processing, and can adjust traffic flow efficiently in real time. A specific intelligent traffic signal control system is shown in fig. 1.

Firstly, extracting original data from a traffic database, wherein the original data comprises license plate numbers, time, traffic flow, speed and longitude and latitude of intersections every 5 minutes. And extracting the time and the traffic flow of each 5-minute intersection in the original data as characteristics to be used as preprocessing data. And removing obviously abnormal data in the preprocessed data, such as negative numbers or decimal numbers appearing in the data. And for the missing data, a Newton interpolation method is adopted to complement the data. And then, normalizing the processed data, wherein the function of the normalization is to avoid the situation that the change of the decimal number is covered by the change of the decimal number when the change of the decimal number occurs in the calculation, and accelerate the convergence speed of the parameters in the solving process. Then VMD decomposes the normalized data into several subsequences, divides the training set and the testing set according to the size of the data set, and sets a proper time sliding window. And then inputting the data into an LSTM neural network for training, adjusting parameters, iterating for multiple times, and seeking the optimum. And finally, weighting and summing the prediction subsequences, carrying out inverse normalization and outputting a prediction result. A specific short-term traffic flow prediction flowchart is shown in fig. 2.

The output prediction result is input to a central information processing control unit, and the central information processing control unit compares vehicles running in the east-west direction and vehicles running in the north-south direction of the crossroad with default values. When the traffic flow of the traffic lights in the east-west direction and the traffic flow of the traffic lights in the north-south direction are predicted to be less than or equal to the default value in a certain time period, the road signal control unit keeps the original default operation of the traffic lights and does not change the duration of the traffic lights. When any predicted value of the traffic flow of the vehicle running in the east-west direction and the traffic flow of the vehicle running in the north-south direction in a certain time period is predicted to be larger than a default value, the central information processing control unit calculates the ratio of the predicted traffic flows in the two directions, the passing time in the east-west direction and the passing time in the north-south direction are divided according to the ratio, and the road signal control unit controls the passing time of the signal lamps in the two directions respectively. Where the default value is the average of the traffic flow at the same time in the last 7 days divided by 2.

The intelligent signal lamp is provided with a special camera for vehicle identification, when the camera shoots that the vehicles in the lane are special vehicles, such as ambulances, police cars, fire trucks and the like, the central information processing control unit processes images of the special vehicles, analyzes whether the special vehicle warning lamps are in a starting state or not, and further identifies whether the special vehicles are in a re-execution state or not through sound. When the execution state of the special vehicle is judged, the road signal lamp control unit can preferentially control the traffic signal lamp to pass the lane, so that the special vehicle can be ensured to pass as soon as possible; when the road is judged to have no special vehicle or the special vehicle is not in the execution state, the road signal lamp control unit controls the signal lamp to adjust the time length of the signal lamp according to the relation between the predicted value and the default value. A specific special vehicle processing flow diagram is shown in fig. 3.

This wisdom signal lamp's structural schematic diagram is shown in fig. 4, and the special camera 1 of wisdom signal lamp, LED traffic lights 3 and LED time-recorder display screen 4 are installed on the surface of wisdom signal lamp connecting rod 6, and the FPGA chip 2 of wisdom signal lamp is installed in the inside of wisdom signal lamp connecting rod 6, and solar cell panel 5 is equipped with at the back of connecting rod, and connecting rod and wisdom signal lamp dead lever 7 are connected to install the dead lever and fix on wisdom signal lamp base 8.

When a vehicle passes through the intersection, the special camera 1 identifies the vehicle, the FPGA chip 2 judges whether the vehicle is the special vehicle and processes the special vehicle, wherein the special camera 1 can also record the number of the vehicles passing through the intersection, the license plate number of the vehicles and other information. The short-time traffic flow prediction unit, the central information processing control unit, the corresponding signal lamp control scheme and the road signal lamp control unit of the intelligent signal lamp are all realized through a chip FPGA. When the weather is clear, the solar cell panel 5 can convert solar energy into electric energy, so that energy is provided for the intelligent signal lamp, and the electric energy is saved.

Claims

1. The utility model provides a wisdom signal lamp when based on VMD-LSTM automatically regulated, includes connecting rod (6), dead lever (7) and base (8), and connecting rod (6) are connected at dead lever (7) top, are equipped with red green signal lamp (3) and time-recorder (4) on connecting rod (6), its characterized in that, be equipped with camera (1) on connecting rod (6), connecting rod (6) inside is equipped with FPGA chip (2), FPGA chip (2) are connected with camera (1), red green signal lamp (3) and time-recorder (4) respectively.

2. The intelligent signal lamp based on VMD-LSTM automatic time adjustment of claim 1, characterized in that the top surface of the connecting rod (6) is provided with a solar panel (5).

3. The intelligent signal lamp based on VMD-LSTM automatic time adjustment of claim 1, wherein the camera (1) is a camera with vehicle type detection function of PT-DV5120 of Winton technologies.

4. The VMD-LSTM based intelligent signal light of claim 1, wherein the red and green signal lights (3) are LED red and green signal lights.

5. The VMD-LSTM based intelligent signal lamp as claimed in claim 1, wherein the timer (4) is an LED timer.