Semi-intelligent prediction-optimization based traffic signal optimization method, system, and device

Abstract

This invention discloses a traffic signal optimization method, system, and device based on semi-intelligent prediction and optimization. The invention includes the following steps: collecting traffic flow data at intersections at different time periods; designing several signal control schemes based on the traffic flow data to form a signal scheme set, where each signal control scheme includes different phase sequences, signal cycles, and green light ratios; constructing a signal scheme optimization model and a model loss function; training the signal scheme optimization model based on the traffic flow data and the model loss function to obtain a signal scheme selection optimization model; and using the signal scheme selection optimization model to select the optimal signal control scheme from the signal scheme set, generating a sequence of optimal signal control schemes for the intersection across multiple time periods. This invention can significantly improve the accuracy of signal scheme selection across multiple time periods, thereby reducing vehicle delays at intersections and improving vehicle traffic efficiency.

Description

Technical Field

[0001] This invention belongs to the field of urban road traffic technology, and relates to a traffic signal optimization method, system and device based on semi-intelligent prediction-optimization. Background Technology

[0002] Urban traffic control systems are a crucial factor influencing traffic flow and play a significant role in promoting urban development. Single-intersection signal control methods mainly fall into three categories: time-of-day (TOD) timing control, sensor-based control, and adaptive control.

[0003] Multi-period timing control is mainly divided into two aspects: multi-period division and signal timing. In terms of signal timing optimization, most researchers currently construct optimization models by considering multiple performance indicators (intersection delay, intersection capacity, number of stops, queue length and fuel consumption, etc.) as optimization objectives, and combine them with constraints to optimize parameters such as intersection cycle length, green light duration and phase difference.

[0004] The above multi-period timing control has a prerequisite: that is, the flow rate of the period to be controlled is available (usually the historical average or obtained through prediction methods). Under such circumstances, the problem to be solved in each period of multi-period timing control can be regarded as a two-stage problem of first predicting the flow rate and then formulating the optimal signal scheme for the intersection based on the flow arrival status.

[0005] In practice, the optimal signaling scheme formulation process for this two-stage problem typically involves selecting the most suitable signaling scheme from a limited set of schemes for the current traffic conditions (e.g., the signaling scheme that minimizes intersection delays). Previous technical solutions for this two-stage problem almost always proceeded separately, i.e., traffic prediction was performed first, followed by scheme selection. This invention integrates these two processes, providing a semi-intelligent prediction-optimization-based traffic signal optimization method, system, and device. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a traffic signal optimization method, system, and device based on semi-intelligent prediction and optimization.

[0007] In a first aspect, the present invention provides a traffic signal optimization method based on semi-intelligent prediction and optimization, comprising the following steps:

[0008] S1. Collect traffic flow data at intersections at different times;

[0009] S2. Based on traffic flow data, design several signal control schemes to form a signal scheme set. Each signal control scheme includes different phase sequence, signal cycle and green ratio.

[0010] S3. Construct a signal scheme optimization model and model loss function;

[0011] S4. Train the signal scheme optimization model based on traffic flow data and model loss function to obtain the signal scheme selection optimization model;

[0012] S5. Use the signal scheme selection optimization model to select the optimal signal control scheme from the signal scheme set and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

[0013] Secondly, the present invention provides a traffic signal optimization system based on semi-intelligent prediction-optimization, comprising:

[0014] Data acquisition module: used to collect traffic flow data at intersections at different times;

[0015] Scheme Design Module: Used to design several signal control schemes based on traffic flow data, forming a signal scheme set. Each signal control scheme includes different phase sequence, signal period, and green ratio.

[0016] Model building module: used to build signal scheme optimization models and model loss functions;

[0017] Model training module: used to train the signal scheme optimization model based on traffic flow data and model loss function, so as to obtain the signal scheme selection optimization model;

[0018] Scheme selection module: Used to select the optimal signal control scheme from the set of signal schemes using the signal scheme selection optimization model, and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

[0019] Thirdly, the present invention provides a computer device, the computer device comprising:

[0020] One or more processors;

[0021] Memory, used to store one or more programs;

[0022] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-described traffic signal optimization method based on semi-intelligent prediction-optimization.

[0023] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions that, when executed by one or more processors, cause the one or more processors to perform the steps in the above-described method.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] 1. This invention constructs a loss function driven by an optimization objective function and reconstructs the optimization objective based on this loss function, achieving excellent results and outperforming existing conventional methods that first predict traffic flow and then optimize the signal scheme.

[0026] 2. This invention can be applied to situations where traffic arrivals change in reality, and the greater the change in traffic arrivals, the more obvious the advantages of the proposed model become.

[0027] 3. The framework proposed in this invention can be easily transferred to any two-stage traffic problem that requires prediction followed by optimization, demonstrating strong method transferability. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of an intersection in an embodiment of the present invention.

[0029] Figure 2 This is a schematic diagram of a traditional prediction and optimization framework.

[0030] Figure 3 This is a schematic diagram of the SPO prediction and optimization framework in an embodiment of the present invention.

[0031] Figure 4 This is a schematic diagram of the signal timing dual-loop structure in an embodiment of the present invention.

[0032] Figure 5 This is a schematic diagram of the data structure for each step of SPO prediction and optimization in an embodiment of the present invention.

[0033] Figure 6 This describes the selection of signal schemes for intersection 1 at different time periods under different methods in this embodiment of the invention.

[0034] Figure 7 This describes the selection of signal schemes for intersection 2 at different time periods under different methods in this embodiment of the invention. Detailed Implementation

[0035] The technical solution of the present invention will be further explained clearly and in detail below with reference to the accompanying drawings and specific examples.

[0036] This application provides a traffic signal optimization method based on semi-intelligent prediction and optimization, including the following steps:

[0037] S1. Collect traffic flow data at intersections at different times;

[0038] S2. Based on traffic flow data, design several signal control schemes to form a signal scheme set. Each signal control scheme includes different phase sequence, signal cycle and green ratio.

[0039] S3. Construct a signal scheme optimization model and model loss function;

[0040] S4. Train the signal scheme optimization model based on traffic flow data and model loss function to obtain the signal scheme selection optimization model;

[0041] S5. Use the signal scheme selection optimization model to select the optimal signal control scheme from the signal scheme set and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

[0042] In a preferred example, the specific steps for forming the signal scheme set are as follows:

[0043] Considering the arrival characteristics of traffic flow at different control periods at the intersection, a subset of optional phase sequences is determined, which in turn determines the organization structure and release order of each direction at the intersection.

[0044] Values ​​are taken around the optimal period for a specific traffic flow scenario to determine a subset of selectable signal periods; the optimal period is calculated according to the Webster signal timing method.

[0045] Taking into account the actual situation, a subset of green ratios is determined, and green ratio schemes that do not conform to the actual traffic conditions are deleted. The green ratio determines the release time weight for different phases.

[0046] After determining all feasible subsets of phase sequence, signal period, and green ratio, the three are combined to obtain the final set of signal schemes.

[0047] In a preferred embodiment, the signal control scheme employs a dual-ring structure for timing design. This structure divides the phases of each flow direction at the intersection into two sets of coordinated phase rings, ensuring the safe passage of traffic flows from different directions by setting phase barriers. During the scheme generation process, the green light duration for each phase must meet the following basic constraints:

[0048] (1) Phase independence constraint: Each phase must ensure the shortest green light time to meet the basic needs of pedestrians crossing the street and clearing the queue of vehicles, while being limited by the longest green light time to avoid excessive waiting time for vehicles in a certain direction;

[0049] (2) Phase combination constraints: The coordination requirements between consecutive phases within the same ring must be met to ensure that the green light duration allocation for key flow directions matches traffic demand;

[0050] (3) Barrier constraints: The two phase groups that cross the barrier need to maintain a time balance to ensure the fairness and efficiency of the allocation of right-of-way in each direction at the intersection.

[0051] Furthermore, considering the time-varying characteristics of traffic demand in each direction at the intersection, the phase duration is not fixed but dynamically adjusted based on real-time traffic flow data while satisfying the aforementioned constraints. Specifically, by analyzing the vehicle arrival rate, queue length, and saturation index of each approach lane, an iterative optimization algorithm is used to determine the optimal green light duration allocation for each phase in the current time period, enabling the signal timing scheme to adaptively respond to changes in traffic demand.

[0052] In a preferred example, the signal scheme optimization model is specifically as follows:

[0053] Traditional methods optimize each time period independently, neglecting the need for a smooth transition between adjacent time periods. Frequent scheme switching leads to driver adaptation difficulties and decreased control effectiveness. Therefore, an optimization model is constructed with the joint objectives of minimizing total daily delay and minimizing scheme switching penalties, introducing a scheme difference penalty term between adjacent time periods. Accordingly, for a single control period in a day, the signal scheme that best matches the traffic flow arrival time within that period should be selected as the signal control scheme for that period. Based on this, a binary decision variable is defined. If signal scheme m is chosen as the signal control scheme for control period h, then the decision variables will be... Set to 1, otherwise 0. The signal scheme selection model can be represented by the mathematical model M1:

[0054] Optimization objective M1:

[0055]

[0056] Constraints:

[0057]

[0058]

[0059] Where h represents a certain time period of the day, H represents the total number of time periods of the day; m represents a certain signal scheme, and M represents the total number of schemes in the phase sequence scheme set; This represents the total vehicle delay at the intersection during time period h, using phase sequence scheme m. yes The function, This indicates the traffic flow of different vehicle types at each entrance of the intersection during time period h. The function representing the difference between schemes in adjacent time periods; This is a penalty coefficient used to control the frequency and magnitude of scheme switching; is the time value weighting coefficient for the k-th type of vehicle, reflecting the differences in social costs among different types of vehicles.

[0060] The optimization objective M1 is to minimize the total vehicle delay at the intersection, which means minimizing the sum of the total vehicle delay and the scheme switching penalty caused by the signal scheme selected in all time periods. At the same time, the first constraint ensures that a phase sequence scheme is assigned in each time period, and the second constraint ensures the domain of the decision variables.

[0061] In the optimization objective M1, the coefficients of the decision variables are: , By predicting traffic first Then, using the delay formula, and considering that for a specific signal scheme, traffic flow is the most critical factor affecting the total vehicle delay, and that the historical traffic flow and delay formula can yield precise values ​​of vehicle delay for different time periods and different signal schemes, it is possible to directly predict vehicle delays for each signal scheme at different times of the next day based on historical delay data for each signal scheme. .

[0062]

[0063] in, The initial loss function is... For the number of signal schemes, This represents the actual vehicle delay under the i-th signal scheme. It is the predicted vehicle delay under the i-th signal scheme.

[0064] In a preferred example, the model loss function minimizes the mean squared error (MSE) while reducing the prediction error between every two signal control schemes, so as to directly predict the total intersection vehicle delay under all signal optimization schemes.

[0065] In a preferred example, the model loss function for:

[0066]

[0067] in, It represents the sum of squares of the difference between the predicted and actual total vehicle delays for each of the two options; and Let represent the actual value and the predicted value of the total vehicle delay at the intersection for the i-th scheme, respectively.

[0068] also, The model references the Webster delay model, which is suitable for intersections with low saturation (typically <0.9). The specific calculation formula is as follows:

[0069]

[0070]

[0071] First item For uniform delay, this term represents the delay caused when vehicles arrive at a uniform rate. It assumes that vehicles queue evenly during the red light and disperse evenly during the green light.

[0072]

[0073] Second item For random delays (oversaturation delays), this term takes into account the randomness of vehicle arrivals and the additional delays caused by temporary oversaturation (i.e., slightly more vehicles arriving than available in a given period).

[0074]

[0075] Third item This is a correction term (simulation correction), an empirical correction term derived from fitting simulation data. Its purpose is to better match the theoretical curve with the simulation results observed by Webster in the laboratory, correcting the deviation of the first two terms to within a 5%-15% error range.

[0076] in: The average delay per vehicle (seconds / vehicle); The signal period duration (in seconds); Green light ratio (i.e., effective green light time) With cycle duration The ratio, ) Arrival flow rate (vehicles / second); Saturation (i.e., flow rate) With traffic capacity The ratio, = ,in (This is the saturation flow rate). Saturation flow rate (vehicles / second);

[0077] In a preferred example, a signal scheme optimization model is trained based on traffic flow data and a model loss function to obtain a signal scheme selection optimization model M2, which can be expressed as:

[0078]

[0079] Constraints:

[0080]

[0081]

[0082] Where h represents a certain time period of the day, H represents the total number of time periods of the day; m represents a certain signal scheme, and M represents the total number of schemes in the phase sequence scheme set; This represents the total vehicle delay at the intersection during time period h, using phase sequence scheme m. yes The function, This represents the traffic flow vector for each approach lane of the intersection during time period h. In optimization model M2, the value to be predicted is... Typically, the prediction objective is to minimize the sum of the squared errors (MSE) between the predicted and actual values. Observing the optimization model M2 reveals that if simultaneously... Overestimating or underestimating the same value does not affect the choice of phase scheme. The predicted value serves to guide the selection of the signal scheme, but is not used as a parameter of the signal scheme. Therefore, the loss function is derived from the structure of the optimization model.

[0083] In a preferred example, for a single control period within a day, the signal scheme selection optimization model should select the signal control scheme that minimizes the total vehicle delay at the intersection during that period as the optimal signal control scheme. Since the signal scheme selection optimization model is integrated into the loss function of the prediction process and directly predicts the delays of different signal schemes, the optimization results can directly generate signal schemes for different time periods. This invention can significantly improve the accuracy of signal scheme selection across multiple time periods, thereby reducing intersection vehicle delays and improving vehicle traffic efficiency.

[0084] To verify the practical application effect of the proposed solution, two intersections were selected for case analysis. Both intersections are typical four-phase signal-controlled intersections, such as... Figure 1 As shown, lane-level traffic flow data with a 3-minute granularity was obtained through video detection from August 3, 2020 to August 31, 2020. During data processing, traffic flow data for Saturdays and Sundays were deleted, and the remaining data was reconstructed into a 21-day time series. The data from the first 20 days was used as historical prediction data, and the data from the 21st day was used as validation data.

[0085] The following sections optimize the signal schemes for two intersections using five different methods, and compare the scheme selections for each method across 24 time periods to analyze their consistency with the theoretically optimal scheme. Furthermore, the sections compare the absolute error of total vehicle delay with the theoretically optimal scheme to analyze the delay differences between the two schemes. Figure 2 and Figure 3 These are schematic diagrams of the traditional prediction and optimization framework and the SPO prediction and optimization framework in the embodiments of the present invention; Figure 4 This is a schematic diagram of the signal timing dual-loop structure in an embodiment of the present invention; Figure 5This is a schematic diagram of the data structure for each step of SPO prediction and optimization in an embodiment of the present invention.

[0086] Figure 6 and Figure 7 The data for the third set of data demonstrates the selection of schemes under the three proposed schemes and two control schemes. FS represents the fixed scheme, where the first scheme from the set of 12 schemes is selected for all 24 control periods of the day; BS represents the theoretically optimal scheme. It can be seen that the schemes generated by FS for all 24 periods are not optimal, and their errors compared to the theoretically optimal scheme are significantly greater than those of the other schemes. R1 and R2 selected the optimal scheme for 15 control periods, but R2's total vehicle delay error was smaller than R1's. R3 selected the optimal scheme for 21 periods, and compared to FS, R1, and R2, it had the smallest total vehicle delay error.

[0087] This application also provides a traffic signal optimization system based on semi-intelligent prediction-optimization, including:

[0088] Data acquisition module: used to collect traffic flow data at intersections at different times;

[0089] Scheme Design Module: Used to design several signal control schemes based on traffic flow data, forming a signal scheme set. Each signal control scheme includes different phase sequence, signal period, and green ratio.

[0090] Model building module: used to build signal scheme optimization models and model loss functions;

[0091] Model training module: used to train the signal scheme optimization model based on traffic flow data and model loss function, so as to obtain the signal scheme selection optimization model;

[0092] Scheme selection module: Used to select the optimal signal control scheme from the set of signal schemes using the signal scheme selection optimization model, and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

[0093] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0095] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0096] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0097] The above description is merely a preferred embodiment of the present invention. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make many possible variations and modifications to the technical solutions of the present invention using the methods and techniques disclosed above, or modify them into equivalent embodiments with equivalent changes, without departing from the scope of the technical solutions of the present invention. Therefore, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall still fall within the protection scope of the technical solutions of the present invention.

Claims

1. A traffic signal optimization method based on semi-intelligent prediction-optimization, characterized in that, Includes the following steps: S1. Collect traffic flow data at intersections at different times; S2. Based on traffic flow data, design several signal control schemes to form a signal scheme set. Each signal control scheme includes different phase sequence, signal cycle and green ratio. S3. Construct a signal scheme optimization model and model loss function; S4. Train the signal scheme optimization model based on traffic flow data and model loss function to obtain the signal scheme selection optimization model; S5. Use the signal scheme selection optimization model to select the optimal signal control scheme from the signal scheme set and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

2. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 1, characterized in that, The specific steps for forming the signal scheme set are as follows: Considering the arrival characteristics of traffic flow at different control periods at the intersection, determine the selectable subset of phase sequence; Values ​​are taken around the optimal period for a specific traffic flow scenario to determine the selectable subset of signal periods; Taking into account the actual situation, determine the optional subset of green ratios and delete green ratio schemes that do not conform to the actual traffic conditions; After determining all feasible subsets of phase sequence, signal period, and green ratio, the three are combined to obtain the final set of signal schemes.

3. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 2, characterized in that, The optimal period is calculated based on the Webster signal timing method.

4. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 1, characterized in that, The signal control scheme satisfies the dual-loop structure signal timing method.

5. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 1 or 4, characterized in that, The model loss function aims to minimize the prediction error between any two signal control schemes, so as to directly predict the total vehicle delay at the intersection under each signal optimization scheme.

6. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 5, characterized in that, The model loss function for: in, For the number of signal schemes, It represents the sum of squares of the difference between the predicted and actual total vehicle delays for each of the two options; and Let represent the actual value and the predicted value of the total vehicle delay at the intersection for the i-th scheme, respectively.

7. The traffic signal optimization method based on semi-intelligent prediction-optimization as described in claim 1, characterized in that, For a single control period during the day, the signal scheme selection optimization model selects the signal control scheme that minimizes the total vehicle delay at the intersection during that period as the optimal signal control scheme.

8. A traffic signal optimization system based on semi-intelligent prediction-optimization, characterized in that, include: Data acquisition module: used to collect traffic flow data at intersections at different times; Scheme Design Module: Used to design several signal control schemes based on traffic flow data, forming a signal scheme set. Each signal control scheme includes different phase sequence, signal period, and green ratio. Model building module: used to build signal scheme optimization models and model loss functions; Model training module: used to train the signal scheme optimization model based on traffic flow data and model loss function, so as to obtain the signal scheme selection optimization model; Scheme selection module: Used to select the optimal signal control scheme from the set of signal schemes using the signal scheme selection optimization model, and generate a sequence of optimal signal control schemes for multiple time periods at the intersection.

9. A computer device, characterized in that, The computer device includes: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the semi-intelligent prediction-optimization-based traffic signal optimization method as described in any one of claims 1-7.

10. A computer-readable storage medium storing computer instructions, characterized in that, When the computer instructions are executed by one or more processors, the one or more processors cause the processors to perform the steps of the method according to any one of claims 1-7.

Images