Processing method for dynamically adjusting traffic light signal scheme according to traffic flow

Abstract

The invention discloses a processing method for dynamically adjusting a traffic light signal scheme according to traffic flow, which comprises the following steps of: deploying sensors including a ground induction coil, a camera and a pedestrian button at each traffic intersection, and preprocessing collected traffic flow and pedestrian flow data; and predicting trends of traffic flow, non-motor vehicle flow and pedestrian flow of each intersection in a short period, modeling traffic light states of each intersection into a dynamic system, calculating an optimal traffic light adjustment strategy, and adjusting the traffic light states in real time. By dynamically adjusting the traffic light signals, the system can more effectively respond to the actual on-site conditions including the flow of pedestrians and non-motor vehicles, more reasonable signal duration distribution is achieved, waiting time is shortened, traffic efficiency is improved, meanwhile, the pedestrians and the non-motor vehicles are considered, the safety of the system can be improved through an optimized signal control strategy, and the traffic light control method is suitable for popularization and application. The signal lamp is adjusted to provide more transition time for pedestrians and non-motor vehicles, so that the accident risk is reduced.

Description

Technical Field

[0001] The invention relates to the field of traffic signal control, and in particular to a processing method for dynamically adjusting a traffic light signal scheme according to vehicle flow. Background Art

[0002] With the acceleration of urbanization, urban traffic congestion has become a serious problem, bringing many inconveniences to people's travel. The traffic light signal control method is usually static, adjusting the changes of traffic lights according to a fixed timing cycle. However, the static method cannot flexibly respond to changes in real-time traffic conditions, resulting in unreasonable distribution of traffic light duration and widespread traffic congestion.

[0003] At present, some traffic signal control methods based on sensor data and intelligent algorithms have been proposed. For example, the traffic light duration is dynamically adjusted using vehicle flow data to reduce traffic congestion. However, most of these methods only consider the motor vehicle flow and ignore the impact of pedestrians and non-motor vehicles. In urban traffic management, the flow of pedestrians and non-motor vehicles also occupies an important position, and their behavior has a significant impact on traffic smoothness and safety. Therefore, it is necessary to design a processing method that dynamically adjusts the traffic light signal scheme according to the vehicle flow. Summary of the invention

[0004] The object of the present invention is to provide a method for dynamically adjusting a traffic light signal scheme according to vehicle flow, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for dynamically adjusting a traffic light signal scheme according to vehicle flow comprises the following steps:

[0007] Step S1, deploying sensors including ground sensing coils, cameras and pedestrian buttons at each traffic intersection to monitor the flow, speed and waiting time of motor vehicles, non-motor vehicles and pedestrians in real time;

[0008] Step S2, pre-processing the collected vehicle flow and pedestrian flow data, cleaning the data, eliminating noise and formatting it to meet the needs of subsequent analysis and processing;

[0009] Step S3, using time series analysis and machine learning models to predict the trends of vehicle flow, non-motor vehicle flow and pedestrian flow at each intersection in the short term based on real-time data and historical data;

[0010] Step S4, based on the traffic demands of motor vehicles, non-motor vehicles and pedestrians, a multi-stage decision problem is defined, and the traffic light status of each intersection is modeled as a dynamic system;

[0011] Step S5, using a dynamic programming algorithm for multi-objective optimization, the goal is to minimize the total waiting time of vehicles and pedestrians while reducing traffic congestion, and calculate the optimal signal light adjustment strategy;

[0012] Step S6, adjusting the traffic light status in real time according to the algorithm result, while taking into account the priority and traffic safety of motor vehicles, non-motor vehicles and pedestrians;

[0013] Step S7, implement the real-time feedback system, monitor the adjusted traffic conditions and effects, and optimize the traffic light control strategy based on real-time data feedback.

[0014] Furthermore, it also includes transmitting data to a central control system in real time through wireless communication technology. The central control system is responsible for processing all data and issuing traffic light adjustment commands.

[0015] Furthermore, an emergency response mechanism is further included for preferentially adjusting the signal light cycle to quickly evacuate vehicles and pedestrians when an emergency situation is detected.

[0016] Furthermore, in step S1, the camera can identify the type and number of pedestrians and non-motor vehicles through image recognition technology, and accurately adjust the corresponding traffic light signal duration.

[0017] Furthermore, in step S3, the prediction model is based on machine learning and uses a deep neural network to predict traffic flow in different time periods in the future;

[0018] In step S3, a deep neural network is used to predict traffic flow in different time periods in the future. The mathematical model is a deep neural network, and its formula is:

[0019] [\mathbf{y}=f(\mathbf{W}\cdot\mathbf{x}+\mathbf{b})];

[0020] Among them, (\mathbf{x}) is the input layer, which represents historical traffic data, (\mathbf{W}) is the weight matrix, (\mathbf{b}) is the bias term, (\mathbf{y}) is the output layer, which represents the predicted traffic flow, and (f) is the activation function, which uses the ReLU function (f(z)=\max(0,z)).

[0021] Furthermore, in step S5, the dynamic programming algorithm includes considering the impact of weather conditions on traffic flow and automatically adjusting the traffic light strategy to adapt to different climate conditions.

[0022] Furthermore, in step S5, the dynamic programming algorithm also includes optimizing the duration and switching frequency of traffic lights in time and space, so as to take into account the flow requirements of motor vehicles, non-motor vehicles and pedestrians during peak traffic hours or in the face of special events;

[0023] In step S5, a multi-objective optimization dynamic programming algorithm is used to calculate the optimal traffic light adjustment strategy. The multi-objective optimization dynamic programming formula is:

[0024] [V_i(s)=\max_{a\inA}(R(s,a)+\gamma\sum_{s'}P(s'|s,a)V_{i-1}(s'))];

[0025] Where (V_i(s)) is the maximum total reward for reaching the goal in state (s), and (\gamma) is a discount factor reflecting the importance of future rewards.

[0026] Furthermore, in step S5, the dynamic programming algorithm further includes taking into account factors for reducing environmental pollution and noise pollution, and reducing vehicle idling and emissions by optimizing traffic flow;

[0027] In the reduction of environmental and noise pollution, an additional cost function is used:

[0028] [C_{\text{env}}=\alpha\cdotE+\beta\cdotN];

[0029] Among them, (E) represents the environmental pollution cost caused by vehicle emissions, (N) represents the noise pollution cost, and (\alpha) and (\beta) are adjustment parameters.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. By dynamically adjusting the traffic light signals, the present invention can more effectively respond to the actual situation on site, including the flow of pedestrians and non-motor vehicles, achieve a more reasonable signal duration distribution, reduce waiting time, and improve traffic efficiency. At the same time, considering pedestrians and non-motor vehicles, the optimized signal control strategy can increase their safety, and reduce the risk of accidents by adjusting the traffic lights to provide more transition time for pedestrians and non-motor vehicles.

[0032] 2. The present invention can balance the needs of different traffic flows, including motor vehicles, non-motor vehicles and pedestrians, by optimizing signal light control, thereby reducing traffic congestion and cross-interference, improving road capacity, and using an emergency response mechanism. The system can quickly adjust signal lights to evacuate vehicles and pedestrians in emergency situations to ensure smooth and safe traffic. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The present invention is a flowchart of a processing method for dynamically adjusting traffic light signal solutions according to vehicle flow. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] See also Figure 1 , the present invention provides a technical solution:

[0036] A method for dynamically adjusting a traffic light signal scheme according to vehicle flow comprises the following steps:

[0037] Step S1, deploying sensors including ground sensing coils, cameras and pedestrian buttons at each traffic intersection to monitor the flow, speed and waiting time of motor vehicles, non-motor vehicles and pedestrians in real time;

[0038] Preferably, the camera can identify the type and number of pedestrians and non-motor vehicles through image recognition technology, and accurately adjust the corresponding traffic light signal duration.

[0039] Step S2, pre-processing the collected vehicle flow and pedestrian flow data, cleaning the data, eliminating noise and formatting it to meet the needs of subsequent analysis and processing;

[0040] Step S3, using time series analysis and machine learning models to predict the trends of vehicle flow, non-motor vehicle flow and pedestrian flow at each intersection in the short term based on real-time data and historical data;

[0041] Preferably, the prediction model is based on machine learning and uses a deep neural network to predict traffic flow at different time periods in the future;

[0042] In step S3, a deep neural network is used to predict traffic flow in different time periods in the future. The mathematical model is a deep neural network, and its formula is:

[0043] [\mathbf{y}=f(\mathbf{W}\cdot\mathbf{x}+\mathbf{b})];

[0044] Among them, (\mathbf{x}) is the input layer, which represents historical traffic data, (\mathbf{W}) is the weight matrix, (\mathbf{b}) is the bias term, (\mathbf{y}) is the output layer, which represents the predicted traffic flow, and (f) is the activation function, which uses the ReLU function (f(z)=\max(0,z)).

[0045] Step S4, based on the traffic demands of motor vehicles, non-motor vehicles and pedestrians, a multi-stage decision problem is defined, and the traffic light status of each intersection is modeled as a dynamic system;

[0046] Step S5, using a dynamic programming algorithm for multi-objective optimization, the goal is to minimize the total waiting time of vehicles and pedestrians while reducing traffic congestion, and calculate the optimal signal light adjustment strategy;

[0047] Preferably, the dynamic programming algorithm includes considering the impact of weather conditions on traffic flow and automatically adjusting the signal light strategy to adapt to different climate conditions;

[0048] Furthermore, the dynamic programming algorithm also includes temporal and spatial optimization of the duration and switching frequency of traffic lights, which can take into account the flow needs of motor vehicles, non-motor vehicles and pedestrians during peak traffic hours or when facing special events.

[0049] Furthermore, the dynamic programming algorithm further includes consideration of reducing environmental pollution and noise pollution, reducing vehicle idling and emissions by optimizing traffic flow;

[0050] In the reduction of environmental and noise pollution, an additional cost function is used:

[0051] [C_{\text{env}}=\alpha\cdotE+\beta\cdotN];

[0052] Among them, (E) represents the environmental pollution cost caused by vehicle emissions, (N) represents the noise pollution cost, and (\alpha) and (\beta) are adjustment parameters.

[0053] Step S6, adjusting the traffic light status in real time according to the algorithm result, while taking into account the priority and traffic safety of motor vehicles, non-motor vehicles and pedestrians;

[0054] Step S7, implement the real-time feedback system, monitor the adjusted traffic conditions and effects, and optimize the traffic light control strategy based on real-time data feedback.

[0055] The present invention also includes transmitting data to a central control system in real time through wireless communication technology, and the central control system is responsible for processing all data and issuing signal light adjustment commands.

[0056] The present invention further includes an emergency response mechanism for preferentially adjusting the signal light cycle to quickly evacuate vehicles and pedestrians when an emergency situation is detected.

[0057] In this embodiment, the steps of constructing the dynamic programming algorithm for multi-objective optimization are as follows:

[0058] The state (S) is defined as the traffic flow status at each intersection in the traffic network, including the number and flow direction of motor vehicles, non-motor vehicles and pedestrians. The state can also include time information, such as the time of day, and weather conditions that affect traffic flow.

[0059] Action (A) is defined as the possible adjustments to the signal light status, for example, the duration of traffic lights. Each action should include specific adjustments to the traffic lights at each intersection, such as the duration of red, green and yellow lights.

[0060] The reward function (R(s,a)) is defined as the immediate return obtained by taking action (a) in state (s). It is usually a negative value, representing the cost. The goal is to minimize the total cost. The cost here can be:

[0061] Waiting time: average waiting time for vehicles, non-motor vehicles and pedestrians.

[0062] Traffic congestion: vehicle density and traffic delays at intersections.

[0063] The reward function is set as:

[0064] [R(s,a)=-(\omega_1\times\text{waiting time}+\omega_2\times\text{traffic congestion})];

[0065] Among them, (\omega_1) and (\omega_2) are weight factors used to adjust the contribution of waiting time and congestion to the total cost.

[0066] The state transition probability (P(s'|s,a)) describes the probability of transitioning to a new state (s') after taking action (a) in state (s), which usually depends on the simulation results of the traffic model or historical data statistics.

[0067] The dynamic programming formula for multi-objective optimization is:

[0068] [V_i(s)=\max_{a\inA}(R(s,a)+\gamma\sum_{s'}P(s'|s,a)V_{i-1}(s'))];

[0069] Specifically, (V_i(s)) is the state value function at stage (i) at a given state (s), which represents the maximum cumulative reward that can be expected from starting from state (s) and acting according to the optimal strategy. (s) is the current state, which can be any set of variables describing the decision-making environment, such as the traffic flow and signal light status in a traffic system. (a) is the selectable action in state (s). Each action in the action set (A) can change the current state and lead to one or more new states. (R(s,a)) is the immediate reward function, which describes the immediate reward obtained by taking action (a) in state (s). According to the result of the action, for example, changing Improve traffic flow and reduce waiting time. (\gamma) is the discount factor, which usually ranges from 0 to 1. It is used to measure the importance of future rewards relative to immediate rewards. The closer the value is to 1, the more important the future rewards are in the decision-making. (\sum_{s'}P(s'|s,a)V_{i-1}(s')) is the expected discounted future reward. (P(s'|s,a)) is the state transition probability, which means the probability of transferring to state (s') after taking action (a) in state (s). (V_{i-1}(s')) is the state value function in the new state (s'), calculated according to the value of the previous stage (i-1).

[0070] For all states (s), initialize (V_0(s)), use the above dynamic programming recursive formula to update (V_i(s)) until convergence, and based on the final value function (V(s)), infer the optimal action (a^*) for each state.

[0071] Assume that traffic at a certain intersection in the traffic network is often congested during the morning rush hour on weekdays. By applying the above dynamic programming algorithm, the system can extend the green light duration during specific time periods to alleviate congestion, and optimize the signal light cycle during non-peak hours to balance the flow efficiency of vehicles and pedestrians.

[0072] The real-time processing flow of this embodiment is as follows:

[0073] Traffic cameras, geomagnetic vehicle detectors, infrared sensors, etc. are used to collect real-time traffic data about intersections, including the number of vehicles, driving speed, and pedestrian flow. The collected data is sent to the central traffic management system, and the data is used to update the traffic status(s) in real time.

[0074] Based on the collected data, the status of each intersection is updated in real time, including traffic density, vehicle type, and number of pedestrians, and the short-term traffic flow prediction model is used to estimate the changes in traffic status in the next few minutes.

[0075] Based on the current state (s), a dynamic programming algorithm is run to calculate the optimal signal adjustment (a^*), and the decision strategy is updated in real time to adjust the duration and sequence of traffic lights based on the current and predicted states.

[0076] The calculated optimal traffic light adjustment instructions are sent to the corresponding intersection signal controller to monitor the effects of the signal adjustments, such as reduced waiting time and improved traffic flow.

[0077] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0078] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamically adjusting a traffic light signal scheme according to vehicle flow, characterized in that: The following steps are involved: Step S1, deploying sensors including ground sensing coils, cameras and pedestrian buttons at each traffic intersection to monitor the flow, speed and waiting time of motor vehicles, non-motor vehicles and pedestrians in real time; Step S2, pre-processing the collected vehicle flow and pedestrian flow data, cleaning the data, eliminating noise and formatting it to meet the needs of subsequent analysis and processing; Step S3, using time series analysis and machine learning models to predict the trends of vehicle flow, non-motor vehicle flow and pedestrian flow at each intersection in the short term based on real-time data and historical data; Step S4, based on the traffic demands of motor vehicles, non-motor vehicles and pedestrians, a multi-stage decision problem is defined, and the traffic light status of each intersection is modeled as a dynamic system; Step S5, using a dynamic programming algorithm for multi-objective optimization, the goal is to minimize the total waiting time of vehicles and pedestrians while reducing traffic congestion, and calculate the optimal signal light adjustment strategy; Step S6, adjusting the traffic light status in real time according to the algorithm result, while taking into account the priority and traffic safety of motor vehicles, non-motor vehicles and pedestrians; Step S7, implement the real-time feedback system, monitor the adjusted traffic conditions and effects, and optimize the traffic light control strategy based on real-time data feedback.

2. The method for dynamically adjusting the traffic light signal scheme according to the traffic volume according to claim 1, characterized in that: It also includes real-time data transmission to a central control system via wireless communication technology, which is responsible for processing all data and issuing traffic light adjustment commands.

3. The method for dynamically adjusting the traffic light signal scheme according to the traffic flow according to claim 1, characterized in that: An emergency response mechanism is further included for prioritizing the adjustment of the signal light cycle to quickly evacuate vehicles and pedestrians when an emergency situation is detected.

4. The method for dynamically adjusting a traffic light signal scheme according to vehicle flow according to claim 1, characterized in that: In step S1, the camera can identify the type and number of pedestrians and non-motor vehicles through image recognition technology, and accurately adjust the corresponding traffic light signal duration.

5. The method for dynamically adjusting the traffic light signal scheme according to the traffic flow according to claim 1 is characterized in that: In step S3, the prediction model is based on machine learning and uses a deep neural network to predict traffic flow in different time periods in the future; In step S3, a deep neural network is used to predict traffic flow in different time periods in the future. The mathematical model is a deep neural network, and its formula is: [\mathbf{y}=f(\mathbf{W}\cdot\mathbf{x}+\mathbf{b})]; Among them, (\mathbf{x}) is the input layer, which represents historical traffic data, (\mathbf{W}) is the weight matrix, (\mathbf{b}) is the bias term, (\mathbf{y}) is the output layer, which represents the predicted traffic flow, and (f) is the activation function, which uses the ReLU function (f(z)=\max(0,z)).

6. The method for dynamically adjusting the traffic light signal scheme according to the traffic volume according to claim 1, characterized in that: In step S5, the dynamic programming algorithm includes considering the impact of weather conditions on traffic flow and automatically adjusting the traffic light strategy to adapt to different climate conditions.

7. The method for dynamically adjusting the traffic light signal scheme according to the traffic volume according to claim 1, characterized in that: In step S5, the dynamic programming algorithm also includes optimizing the duration and switching frequency of traffic lights in time and space, and taking into account the flow requirements of motor vehicles, non-motor vehicles and pedestrians during peak traffic hours or when facing special events; In step S5, a multi-objective optimization dynamic programming algorithm is used to calculate the optimal traffic light adjustment strategy. The multi-objective optimization dynamic programming formula is: [V_i(s)=\max_{a\inA}(R(s,a)+\gamma\sum_{s'}P(s'|s,a)V_{i-1}(s'))]; Where (V_i(s)) is the maximum total reward for reaching the goal in state (s), and (\gamma) is a discount factor reflecting the importance of future rewards.

8. The method for dynamically adjusting the traffic light signal scheme according to the traffic volume according to claim 1, characterized in that: In step S5, the dynamic programming algorithm further includes taking into account factors for reducing environmental pollution and noise pollution, and reducing vehicle idling and emissions by optimizing traffic flow; In the reduction of environmental and noise pollution, an additional cost function is used: [C_{\text{env}}=\alpha\cdotE+\beta\cdotN]; Among them, (E) represents the environmental pollution cost caused by vehicle emissions, (N) represents the noise pollution cost, and (\alpha) and (\beta) are adjustment parameters.

Images