Dual system traffic signal control system with automatic failover function

Abstract

The application relates to the technical field of traffic signal control, and discloses a double-system traffic signal control system with automatic fault conversion function. The system comprises a state monitoring module, a fault prediction module, a window calculation module, a time period acquisition module, an overlap analysis module, a time length prediction module and a switching decision module. The state monitoring module continuously acquires main system operation parameters, which are transmitted to the fault prediction module to estimate fault occurrence time; the window calculation module calculates a time window for switching in combination with the fault occurrence time and a traffic peak time, and the window end point is the traffic peak time minus the time length required for signal stabilization; the time period acquisition module acquires a low traffic flow time period, the overlap analysis module judges whether the two time periods overlap; if the two time periods overlap, the time length prediction module calculates the total operation time length of switching; and the switching decision module compares the two time lengths, and if the two time lengths are equal, switching is triggered at the overlap start point, automatic double-system fault conversion is realized, the timing is accurate, and the traffic signal stability is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traffic signal control, in particular to a dual-system traffic signal control system with automatic fault switching function. BACKGROUND

[0002] Traffic signal system is an important regulating facility for urban road traffic operation, and its stable operation is directly related to the efficiency and order of intersection traffic. With the continuous growth of urban traffic flow, the operation reliability of single traffic signal control system has been difficult to meet the actual demand. Once the main system fails, if it cannot be switched to the standby system in time, it is easy to cause the interruption of intersection signal, vehicle congestion and even traffic accidents, which will seriously affect the urban traffic operation.

[0003] Currently, although some traffic signal control systems are equipped with standby systems, there are still many deficiencies in fault handling and system switching. The traditional fault handling mode mainly depends on the post-response, that is, the standby system switching process is started after the main system fails. Under this mode, there is a long time delay from fault discovery, manual confirmation to starting the standby system, and during this period, the traffic signal is out of control, the vehicles at the intersection are in disorder, which is easy to form a congestion bottleneck. Especially in the peak traffic period, the congestion situation will spread quickly, causing serious interference to the regional traffic operation.

[0004] Even if some systems have preliminary fault warning functions, they cannot fully consider the characteristics of traffic operation period to plan the switching time. For example, some warning systems can only predict the approximate time of failure, but they do not consider the traffic peak time and the time required for signal stability, resulting in unreasonable switching window setting. If the switching is started near the peak period, even if the switching process is completed, the signal system still needs a certain time to adjust to the stable state, and the signal fluctuation before stability will directly affect the traffic efficiency in the peak period. In addition, some systems do not fully utilize the advantages of low traffic flow period when selecting the switching period. If the switching is performed in a period with high traffic flow, the short adjustment of the signal during the switching process is easy to cause vehicle queuing, further aggravating traffic congestion. At the same time, the existing systems often ignore various time-consuming factors in actual operation when calculating the switching operation time, resulting in a large deviation between the estimated time and the actual time. If the overlapping period is shorter than the actual switching time, it will cause the interruption of the switching process, and the switching time needs to be selected again, further increasing the instability risk of the traffic signal system. The existence of these problems makes the fault switching capability of the existing dual-system traffic signal control system difficult to meet the high requirements of urban traffic on signal stability and traffic efficiency, and a system scheme is needed that can accurately plan the switching time and realize automatic and efficient fault switching. SUMMARY

[0005] The present application aims to provide a dual-system traffic signal control system with automatic fault switching function to solve the problems in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides a dual-system traffic signal control system with automatic fault switching function, which comprises:

[0007] a state monitoring module, a fault prediction module, a window calculation module, a time period acquisition module, an overlap analysis module, a time length prediction module and a switching decision module;

[0008] The state monitoring module continuously acquires the operating parameters of the main traffic signal control system and transmits them to the fault prediction module to estimate the fault occurrence time; the fault prediction module outputs the fault occurrence time to the window calculation module, which calculates the time window for switching based on the fault occurrence time and the traffic peak time, wherein the end of the time window is the traffic peak time minus the time length required for signal stabilization; the time period acquisition module acquires the low traffic flow period information, the overlap analysis module receives the time window and the low traffic flow period and determines whether there is an overlapping period; if there is an overlapping period, the time length prediction module calculates the total operation time length for switching from the main system to the backup system; the switching decision module compares the overlapping period with the total operation time length and, when the overlapping period is equal to the total operation time length, takes the start of the overlapping period as the switching time and triggers the switching action.

[0009] Preferably, the time length prediction module is configured to perform:

[0010] The device operation history data of the main traffic signal control system is continuously acquired, including the fault occurrence frequency and the system response delay time;

[0011] The start-up parameters of the backup system are acquired, including the initialization time and the signal synchronization time; the device operation history data and the start-up parameters are trained using a neural network model to generate dynamic weight coefficients;

[0012] Based on the dynamic weight coefficients, the fault occurrence frequency, the system response delay time, the initialization time and the signal synchronization time are fused and calculated to output the total operation time length.

[0013] Preferably, the window calculation module is configured to perform: when the overlapping period is greater than the total operation time length, a greedy algorithm is used to iteratively select multiple candidate sub-periods from the overlapping period, and the length of each candidate sub-period is not less than the total operation time length; the correlation score of each candidate sub-period with the traffic flow pattern is calculated, and the traffic flow pattern is generated by statistical analysis of historical traffic data; the candidate sub-period set with the highest correlation score is selected as the output.

[0014] Preferably, the switching decision module is configured to perform: receiving the candidate sub-period set; calculating a priority index of each candidate sub-period based on a real-time load rate of the primary system, the real-time load rate being provided by the state monitoring module; selecting a candidate sub-period with the highest priority index as a target period, and taking the start point of the target period as the switching time.

[0015] Preferably, the system further comprises a stability evaluation module configured to perform: obtaining a signal stability threshold of the backup system after startup and a signal decay characteristic of the primary system after shutdown; calculating a minimum overlap duration required for the backup system to reach the signal stability threshold; if the minimum overlap duration is less than or equal to the length of the candidate sub-period, selecting a candidate sub-period with the smallest deviation of the end point from the end point of the low traffic period as an optimized period; and the switching decision module taking the start point of the optimized period as the switching time.

[0016] Preferably, the duration prediction module is configured to perform: when the overlap period is less than the total operation duration, calculating a remaining operation requirement duration; the system further comprises a decay simulation module configured to generate a signal quality decay curve based on hardware characteristics of the primary system; the switching decision module compares the signal quality decay curve with a preset signal quality standard curve to determine an effective maintenance duration of the primary system after shutdown; if the remaining operation requirement duration falls within the effective maintenance duration, and the start point of the low traffic period and the sequence of the overlap period and the remaining operation requirement duration do not exceed the end point of the time window, triggering a switching action.

[0017] Preferably, the switching decision module is configured to perform: after completion of the switching action, monitoring the running state of the backup system; obtaining the end point time of the total operation duration, and tracking traffic changes in real time; when the traffic is lower than a threshold, taking the later one of the end point of the low traffic period and the end point time of the total operation duration as the primary system recovery time, and controlling the primary system to reactivate.

[0018] Preferably, the system further comprises a task scheduling module configured to perform: obtaining a predicted failure time sequence of the current batch of traffic control tasks, and generating a switching queuing sequence based on task priority; for each task, calculating its switching start time and switching end time; if the switching end time of any task is later than the end point of the time window, adjusting the queuing sequence, and triggering a recalculation module to update the time window.

[0019] Preferably, the system further comprises an adaptive control module, which is configured to perform: collecting performance indicators and traffic flow data of the backup system in real time during a monitoring period before the main system recovery time; predicting a minimum operation duration of the remaining task based on the performance indicators; if the minimum operation duration allows early recovery, calculating an optimal recovery time in combination with the signal quality decay curve; and updating the main system recovery time by the switching decision module with the optimal recovery time.

[0020] Preferably, the system further comprises a data fusion module, which is configured to perform: checking the integrity of the operating parameters collected by the state monitoring module, and fusing multi-source sensor data to correct the fault occurrence time; and inputting the corrected fault occurrence time into the window calculation module to regenerate the time window, ensuring accurate overlap determination with the low traffic flow period.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] The state monitoring module can continuously collect operating parameters of the main traffic signal control system. This continuous monitoring method can capture subtle changes in the operation of the main system in real time, discover potential operation abnormalities in a timely manner, avoid being in a passive situation due to sudden failure caused by untimely parameter monitoring, and provide real-time and accurate basic data support for subsequent fault prediction and switching planning.

[0023] The fault prediction module estimates the fault occurrence time based on the operating parameters transmitted by the state monitoring module, changes the mode of post-response in traditional fault handling, can master the time node when the main system may fail in advance, thereby reserving sufficient planning time for system switching, effectively avoiding the risk of signal interruption due to no advance preparation when the main system fails suddenly, ensuring that the traffic signal system can enter the switching preparation state before the fault occurs, and reducing the impact of the fault on traffic operation.

[0024] The window calculation module determines the time window allowed for switching based on the fault occurrence time and the traffic peak time in combination with the required duration for signal stabilization, and sets the window end point as the traffic peak time minus the required duration for signal stabilization. This design fully considers the actual requirement that the signal system needs a certain time to adjust to a stable operation state after switching, can ensure that the backup system has enough time to complete signal parameter debugging and stabilization after switching is completed, avoids the problem of decreased traffic efficiency caused by entering the traffic peak period before the signal is stabilized after switching, and guarantees the stable output of the traffic signal during the peak period, thereby providing protection for orderly traffic at intersections during the peak period.

[0025] The time period acquisition module acquires low traffic flow period information, and the overlap analysis module further judges whether the time window allowed to switch and the low traffic flow period have an overlapping period, and the switching planning combined with the traffic flow period characteristics can arrange the system switching in the period with low traffic demand. Under the low traffic flow period, the traffic pressure of the intersection is small, and even if the signal is temporarily adjusted during the switching process, it will not cause large-scale queuing congestion of vehicles, the fault tolerance of the intersection traffic operation is stronger, the influence of the switching operation on the vehicle passing is effectively reduced, and the possibility of traffic disorder during the switching process is reduced.

[0026] The time length prediction module calculates the total operation time length of switching from the main system to the standby system, the switching decision module compares the overlapping period and the total operation time length, and when the two are equal, the starting point of the overlapping period is taken as the switching time to trigger the switching action, and the accurate time length matching and time selection can ensure that the switching operation is completed in the overlapping period, which not only makes full use of the favorable conditions of the low traffic flow period, but also avoids the problem that the switching process is interrupted and needs to be re-planned if the overlapping period is shorter than the switching time length, and at the same time, it also does not waste the resources of the low traffic flow period due to the overlapping period being too long, realizes the accurate matching of the switching time and the operation time length, and improves the efficiency and reliability of the system switching. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 The timing diagram of the dual-system traffic signal control system with automatic fault switching function according to the present application;

[0028] Figure 2 The flowchart for stability evaluation and optimal period selection;

[0029] Figure 3 The flowchart for the main system recovery after switching. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0031] Please refer to Figure 1 The present application provides a dual-system traffic signal control system with automatic fault switching function, which comprises a state monitoring module, a fault prediction module, a window calculation module, a time period acquisition module, an overlap analysis module, a time length prediction module and a switching decision module. The specific implementation is as follows:

[0032] The state monitoring module continuously collects the operating parameters of the main traffic signal control system, which are transmitted to the fault prediction module in real time for estimating the time point at which the main traffic signal control system is likely to fail. The fault prediction module outputs the calculated failure time to the window calculation module, which calculates a time window for system switching in combination with the information of the traffic peak time. The end point of the time window is set as the traffic peak time minus the time required for signal stabilization, so as to avoid interference of the switching operation on the peak traffic flow. The time period collection module operates independently and is responsible for obtaining the information of the low traffic flow period in a day. The overlap analysis module receives the time window from the window calculation module and the low traffic flow period information from the time period collection module, and performs comparative analysis on the two, to determine whether there is an overlapping period. If there is an overlapping period, the time length prediction module is started immediately, to calculate the estimated total operation time length required for switching from the main traffic signal control system to the backup system. The switching decision module compares the length of the overlapping period with the total operation time length provided by the time length prediction module, and in the case where the two are equal, the switching decision module determines the start point of the overlapping period as the optimal switching time, and triggers the switching action at this moment, to start the backup system and stop the main traffic signal control system.

[0033] In Embodiment 1, the function of the time length prediction module depends on the deep analysis and modeling of the historical operation data and system parameters. The module continuously collects the equipment operation historical data from the main traffic signal control system, which constitutes the basis for analysis. The fault occurrence frequency records the number of abnormalities of the main traffic signal control system in a specific period, and the system response delay time quantifies the time interval from receiving the instruction to executing the control action of the main traffic signal control system, which dynamically reflects the health degree and performance of the main traffic signal control system. The time length prediction module also obtains the fixed starting parameters of the backup system from the system configuration library. The initialization time of the backup system refers to the time from issuing the starting instruction to the time when the backup system is available for basic services, and the signal synchronization time refers to the time required for the backup system to complete data synchronization with the current traffic signal control strategy and output stable signals.

[0034] In order to effectively fuse heterogeneous data and output a high-precision total operation time prediction value, the time prediction module introduces a designed neural network model, the input layer nodes of which correspond to four key features: fault occurrence frequency, system response delay time, initialization time, and signal synchronization time. The training process of the neural network model uses a large number of sample data accumulated in history, and the sample data contains the total operation time actually recorded under different system states. The training target is to enable the neural network model to learn the complex nonlinear relationship between the input features and the output results. Through repeated iteration and adjustment of internal weights, the neural network model can finally generate a set of dynamic weight coefficients, which are not fixed and will automatically adjust according to the current input feature values, thereby capturing the differentiated influence degree of different parameters on the total operation time under specific circumstances. In the prediction stage, the time prediction module inputs the real-time collected fault occurrence frequency, system response delay time, and preset initialization time and signal synchronization time into the trained neural network model. The model uses its internal dynamic weight coefficients to perform weighted calculation on these values. The calculation process is not a simple linear superposition, but an information fusion through multi-layer nonlinear transformation of the neural network, and finally outputs a scalar value as the predicted total operation time, which covers the total time consumption from the start of decision switching to the complete takeover and stable operation of the standby system.

[0035] The core responsibility of the window calculation module is to calculate a reasonable operation window based on the failure occurrence time provided by the failure prediction module. The end of the time window is directly determined by the traffic peak time, specifically the traffic peak time minus a fixed signal stabilization time. The signal stabilization time is a preset value that takes into account the time buffer required for the backup system to start, for its control signals to be adapted by road users, and for the expected control effect to be achieved. This calculation method aims to ensure that the entire switching operation and subsequent stabilization period are completed before the traffic flow pressure significantly increases, thereby avoiding exacerbating traffic congestion during peak hours due to system switching. The window calculation module receives the judgment results from the overlap analysis module. When the overlap analysis module determines that there is an overlapping period and the length of the overlapping period is greater than the total operation time given by the time length prediction module, it means that there are multiple potential switching time points to choose from. The window calculation module needs to select the optimal period that has the least impact on traffic flow. The window calculation module uses a greedy algorithm to solve this period optimization problem. The working mechanism of the greedy algorithm is to perform sequential scanning and local optimal selection within the given overlapping period. The algorithm starts from the beginning of the overlapping period and slides a time window with a fixed step size. The length of this time window is set to be no less than the total operation time. Each sub-interval covered by the sliding window is defined as a candidate sub-period. For each identified candidate sub-period, the greedy algorithm needs to evaluate its quality. The evaluation is based on the correlation score of the candidate sub-period with the typical traffic flow pattern. The traffic flow pattern is an abstract model generated based on long-term historical traffic data statistics. It describes the probability distribution or typical value of traffic flow intensity at different times of the day. The correlation score is calculated by comparing the predicted traffic flow in the candidate sub-period with the baseline flow in the corresponding period of the traffic flow pattern. The lower the predicted traffic flow and the smaller the fluctuations, the higher the correlation score, indicating that the system switching in this sub-period is expected to have the least disturbance to traffic operation.

[0036] The greedy algorithm will keep the candidate sub-period with the highest correlation score encountered during the traversal of the whole overlapping period. Since the overlapping period can be long, there can be multiple disjoint sub-periods that meet the length requirement and have high scores. Therefore, the algorithm will collect all the candidate sub-periods that meet the score threshold and form a candidate sub-period set. This candidate sub-period set is the output of the greedy algorithm and is passed to the switching decision module, providing multiple optimization options for the final switching time decision. The advantage of the greedy algorithm is its high computational efficiency, which can quickly find a local optimal solution within a limited time window, meeting the real-time decision speed requirement of the traffic signal control system. The accurate output of the duration prediction module and the intelligent optimization of the window calculation module complement each other. The duration prediction module provides the time reference required for the switching operation, and the window calculation module finds the execution opportunity that minimizes the impact on public transportation based on this reference. The cooperative work of the two modules constitutes the core decision logic of the system for preventive switching in non-emergency situations. The entire process relies on the accuracy of the data and the reliability of the model. The role of the data fusion module is highlighted here. The data fusion module ensures that the historical data and real-time parameters input into the duration prediction module are complete and consistent. Any data anomalies will be detected and corrected, ensuring the accuracy of the neural network model prediction results. Accurate prediction is a prerequisite for effective optimization by the window calculation module. Although the stability evaluation module is not directly involved in this embodiment, the minimum overlapping duration concept calculated by the stability evaluation module can provide an additional stability constraint for the candidate sub-period set. The switching decision module can consider both the priority index and the stability requirement in the subsequent stage to make a more comprehensive decision.

[0037] Example 2: see Figure 2The switching decision module obtains a candidate sub-period set from the window calculation module, the candidate sub-period set contains multiple potential switching windows with a length no less than the total operation duration, and the switching decision module needs to determine a specific execution time from the candidate sub-period set. The switching decision module sends a request to the state monitoring module, and the state monitoring module continuously collects real-time operation parameters of the main traffic signal control system, wherein the real-time load rate is a core index, and the real-time load rate reflects the resource occupation level of the main traffic signal control system in processing the current traffic control task. The switching decision module calculates a priority index for each candidate sub-period in the candidate sub-period set based on the received real-time load rate, and the calculation process considers the high-low change trend of the real-time load rate. A higher real-time load rate or a significant upward trend usually means that the main traffic signal control system is under greater pressure, and the risk of failure is relatively increased. In this case, the priority index of the corresponding period for preventive switching is correspondingly increased. The switching decision module compares the priority indexes of all candidate sub-periods, selects the candidate sub-period with the highest value as the target period, and determines the start point of the target period as the switching time. The switching decision module triggers a control instruction at this moment to start the conversion process from the main traffic signal control system to the standby system.

[0038] The stability evaluation module plays an independent verification and optimization role in the switching decision process. The stability evaluation module does not depend on the priority calculation of the switching decision module, but provides constraint conditions from the perspective of system reliability. The stability evaluation module reads the signal stability threshold of the standby system after startup from the system configuration, and the signal stability threshold defines the minimum quality requirement for the standby system output signal to be considered as a stable and controllable state. The stability evaluation module also analyzes the signal decay characteristics of the main traffic signal control system after shutdown, which is obtained through historical data analysis and hardware performance parameter modeling. It describes the maintenance ability and decay speed of the original signal of the main traffic signal control system on terminal devices such as traffic detectors and signal lights after the system stops working. Based on the signal stability threshold and the signal decay characteristics, the stability evaluation module calculates the minimum overlap duration required for the standby system to reach the signal stability threshold. The minimum overlap duration is a theoretical value that ensures that even in the worst case, there will be no control vacuum or significant quality decline in road traffic signals during the period from the start of switching action to the complete stabilization of the standby system signal.

[0039] The stability evaluation module compares the calculated minimum overlap duration with the actual length of each candidate sub-period in the candidate sub-period set, and filters out all candidate sub-periods whose length is greater than or equal to the minimum overlap duration. These candidate sub-periods form a subset that meets the basic stability requirement. The stability evaluation module further applies an optimization strategy in this subset. The goal of the optimization strategy is to select a time period that best matches the natural traffic low-peak period to minimize the disturbance to the traffic flow. The stability evaluation module compares the end point of each candidate sub-period with the end point of the low-traffic period, and calculates the absolute value of the time deviation between the two. The stability evaluation module selects the candidate sub-period with the smallest absolute value of the end point deviation as the optimized time period. This selection logic tends to make the switching operation and the subsequent stabilization process complete as early as possible within the low-traffic period, thereby leaving a buffer time for possible unexpected situations and allowing the traffic flow to adapt to the new system for a longer time before the peak period arrives. The stability evaluation module sends the information of the optimized time period to the switching decision module.

[0040] The switching decision module synthesizes two inputs at this time, one is the target time period selected based on the priority index calculated according to the real-time load rate, and the other is the optimization time period selected by the stability evaluation module based on the stability constraint and time period matching degree. The switching decision module has decision logic inside to process these two inputs. One implementation is to give higher weight to the stability condition, and the switching decision module preferentially adopts the optimization time period output by the stability evaluation module as the final switching execution time period. The switching decision module determines the starting point of the optimization time period as the system switching time. Another implementation is that the switching decision module performs secondary calculation, normalizes and weights the priority index and end point deviation value, and generates a comprehensive score for the candidate sub-period. The candidate sub-period with the highest comprehensive score is selected as the final execution scheme. The switching decision module issues instructions to the system at the determined switching time, controls the standby system to start and perform control right handover with the main traffic signal control system, and the whole switching process starts to execute. The work of the switching decision module does not end after triggering the switching action. The module enters the post-switching monitoring state. The switching decision module continuously receives reports from the state monitoring module about the running state of the standby system, monitors the initialization progress, signal output stability and whether an abnormal alarm occurs. The switching decision module also obtains the estimated end time of the total operation time from the time length prediction module. This time mark indicates that the standby system should have completely taken over and stably run. The switching decision module tracks the data of the traffic flow sensor in real time, observes the change of the actual traffic flow, and has a traffic flow threshold value inside, which is usually lower than the threshold value for defining the low traffic flow period, for judging whether the conditions for restoring the main system are met. When the real-time traffic flow data falls below the flow threshold value and the current time has exceeded the end time of the total operation time, it means that the switching has been completed and the traffic flow is in a non-busy state. The switching decision module starts to plan the restoration work of the main traffic signal control system. The determination of the restoration time needs to be cautious. The switching decision module compares the end point of the low traffic flow period and the end time of the total operation time, and selects the later one as the restoration time of the main traffic signal control system. This selection ensures that the standby system has sufficient time to stably run, and the restoration operation will not be performed when the traffic flow starts to rise. At the restoration time, the switching decision module triggers the reverse switching process, controls the main traffic signal control system to restart, synchronize data and take control, and makes the standby system return to standby state. The whole automatic fault conversion process forms a closed loop, and the core role of the switching decision module runs throughout, from the decision of preventive switching to the triggering of switching execution, and to the timing of system restoration. The switching decision module automatically processes according to real-time data and preset rules.

[0041] After the total operation duration is calculated by the duration prediction module, the overlap analysis module compares the total operation duration with the current identified overlap period length. When the overlap period is less than the total operation duration, it means that all switching actions cannot be completed once without interruption within the ideal low-flow window, and the system must evaluate the feasibility of segmented operation or the use of residual system functions. The duration prediction module will further calculate a residual operation requirement duration, which is equal in value to the difference between the total operation duration and the overlap period length. This difference represents the time corresponding to the switching operation that still needs to be performed after the low-flow overlap period ends. The attenuation simulation module begins to play a key role in the system. The core function of the attenuation simulation module is to simulate the quality degradation process of the output signals of the main traffic signal control system after it stops active control. The attenuation simulation module has a built-in hardware characteristic model of the main traffic signal control system. The hardware characteristic model is established based on the specific hardware composition of the main traffic signal control system, such as the performance of the central processing unit, the specifications of the signal output board, and the parameters of the communication module. The model simulates the maintenance ability and decay trajectory of the output signal parameters on terminal devices such as road detectors and signal light controllers in the absence of continuous calculation support. The attenuation simulation module runs this model to generate a signal quality decay curve. The signal quality decay curve has time as the horizontal axis and signal quality indicators as the vertical axis. The signal quality indicator is a comprehensive parameter that can cover multiple dimensions such as signal timing consistency, phase switching accuracy, and green ratio stability. The signal quality decay curve usually appears as a monotonically decreasing function of time, which visually depicts the process of the control effect of the main traffic signal control system gradually deteriorating from a stable state after it stops operating.

[0042] The switching decision module compares the signal quality decay curve generated by the decay simulation module with a preset signal quality standard curve, which defines the lower limit of acceptable signal quality and represents the minimum signal control requirement to maintain basic traffic safety and efficiency. By comparing the two curves, the switching decision module can determine a critical time point when the signal quality decay curve drops to the intersection with the signal quality standard curve. The time period from the shutdown of the main traffic signal control system to this intersection time is the effective maintenance duration. The effective maintenance duration quantifies the maximum duration that the residual control instructions of the main traffic signal control system can still effectively manage traffic without causing safety or congestion risks after stopping working. The logic judgment of the switching decision module focuses on the relationship between the remaining operation requirement duration and the effective maintenance duration. The switching decision module checks whether the remaining operation requirement duration is less than or equal to the calculated effective maintenance duration. If the condition is met, it means that although the complete switching operation cannot be completed within the low-flow overlap period, the switching can be started within the overlap period, and the remaining operation can continue within the effective maintenance duration after the main traffic signal control system stops working, without the overall traffic signal quality falling below the acceptable minimum standard. The switching decision module further verifies the time arrangement of the entire operation sequence, which consists of the start of the low traffic flow period, the overlap period, and the remaining operation requirement duration. The switching decision module calculates the total time span from the start of the low traffic flow period, through the overlap period, to the end of the remaining operation requirement duration, and checks whether this total time span exceeds the time window endpoint defined by the window calculation module. The time window endpoint is the time after the traffic peak time minus the time required for signal stabilization, which is the absolute latest time limit for the entire system to allow switching operations.

[0043] The remaining operation requirement duration falls within the effective maintenance duration, and the time span of the entire operation sequence does not exceed the time window endpoint. Only then will the switching decision module determine that the switching action is feasible. At this time, the switching decision module takes the start of the overlap period as the initiation time of the switching action, triggering the switching process of the system. The system starts to perform the first part of the switching operation within the low-flow overlap period, such as starting the basic services of the backup system, performing preliminary data synchronization, etc. After the end of the overlap period and the shutdown of the main traffic signal control system, the system continues to complete the remaining switching operation steps within the window provided by the effective maintenance duration until the backup system completely takes over and runs stably. This operation mode is equivalent to dividing the switching process into two stages, the first stage is performed within the protected low-flow period, and the second stage relies on the functional inertia of the main system to ensure the safety of the transition period.

[0044] To more accurately quantify the relationship between signal decay and time, the decay simulation module uses the following mathematical formula for calculation:

[0045]

[0046] in: Indicates time Signal quality index value at that time, This represents the time elapsed since the autonomous traffic signal control system stopped operating. This represents the initial signal quality attenuation, that is, the difference between the signal quality at the moment of shutdown and the steady-state signal quality. It is the attenuation coefficient, determined by the hardware characteristics of the main traffic signal control system, and reflects how quickly the signal quality deteriorates. This is the signal quality baseline value, representing the background noise level or minimum sustaining value when the system cannot provide effective control. It is a natural constant. The dimensions of both sides of the formula are in units of signal quality, maintaining consistency. Effective duration. It can be solved Received, among which It is the threshold defined by the signal quality standard curve.

[0047] Example 4: See Figure 3The switching decision module continuously receives reports from the state monitoring module about the operating state of the backup system, covering performance indicators such as the central processor load, memory usage, stability of the signal output port, and network communication delay of the backup system. The switching decision module also obtains the estimated total operation time end point from the time length prediction module, which is an important time reference point marking the time when the backup system should theoretically reach a completely stable state. The switching decision module activates the traffic flow tracking function and receives and analyzes real-time traffic flow data from sensors deployed on the road, including the number of vehicles passing through the detection section per unit time, average speed, lane occupancy rate, and other parameters. The switching decision module has a traffic flow threshold set internally, which is a value obtained through historical data analysis and is significantly lower than the threshold value defining the daily low traffic flow period, representing a very sparse and easily managed state of traffic flow. The switching decision module continuously compares real-time traffic flow data with this threshold value. When the system detects that real-time traffic flow has been below the set threshold for a period of time, and the current system time has exceeded the total operation time end point obtained from the time length prediction module, the switching decision module determines that the window for restoring the main traffic signal control system has appeared. The determination of the restoration time needs to consider both the natural end point of the low traffic flow period and the theoretical completion point of the switching operation. The switching decision module compares the end time of the low traffic flow period and the total operation time end point, and selects the later of the two as the restoration time of the main traffic signal control system. This strategy ensures that the backup system has enough time to complete its startup and stabilization process, avoids hasty reverse switching when the backup system is not fully ready, and also takes advantage of the lowest traffic flow period to minimize the disturbance to traffic. At the determined restoration time, the switching decision module sends control instructions to the system to start the restart program of the main traffic signal control system, and the control right is smoothly transferred from the backup system to the restored main traffic signal control system, and the backup system returns to standby state.

[0048] The task scheduling module of the system integration is responsible for handling complex application scenarios where multiple traffic control tasks need to be switched in sequence. The task scheduling module obtains the predicted failure time sequence of all pending traffic control tasks within the batch from the failure prediction module. The predicted failure time sequence includes the time point at which the main traffic signal control system subunit corresponding to each task is predicted to fail. The task scheduling module assigns a priority to each task based on pre-set rules, which may consider factors such as the importance of the intersection controlled by the task, historical failure frequency, current traffic pressure level, etc. The task scheduling module generates a queuing sequence for system switching based on the priority of the tasks, which determines the order of task execution switching operations. For each task in the queuing sequence, the task scheduling module needs to calculate its switching start time and switching end time. The task scheduling module refers to the total operation time provided by the time window prediction module for each task and combines it with the time window constraints to calculate the switching start time and switching end time. The task scheduling module checks the preliminary generated queuing sequence to verify whether there is any task whose switching end time is later than the global time window end point set by the window calculation module for the entire system. The global time window end point is calculated based on the most critical traffic peak time and is the last time limit that all switching operations must comply with. If such conflicting tasks are found, the task scheduling module will start an adjustment algorithm to optimize the queuing sequence. The adjustment algorithm may involve changing the execution order of tasks, postponing non-critical tasks to the next available window, or reorganizing resources for tasks that can be executed in parallel. The task scheduling module relies on an internally constructed task attribute table during decision-making, as shown in Table 1, which summarizes the key parameters required for switching decision-making.

[0049] Table 1: Task scheduling module task attribute table

[0050]

[0051] The task scheduling module uses the data in the table to execute conflict detection and resolution logic. When the task scheduling module identifies a conflicting task (e.g., the task with task ID T004 in the table has a switching end time of 10:55, which is later than the global time window end point of 10:50), the task scheduling module triggers the recalculation module. The recalculation module receives the adjusted task sequence information. The recalculation module may interact with the window calculation module to request a re-evaluation of the feasibility of the global time window based on new task constraints, or the recalculation module may instruct the task scheduling module to further adjust the sequence until all tasks meet the time window constraints. The task scheduling module will finally determine the conflict-free switching queuing sequence and send it to the switching decision module. The switching decision module executes the switching process of each task in sequence based on the sequence and the switching time determined for each task.

[0052] Embodiment 5: The adaptive control module starts active work at a specific monitoring period before the planned recovery time of the main traffic signal control system, the length of the monitoring period can be configured according to system configuration, for example, set to one hour before the planned recovery time. During this period, the adaptive control module collects key data at a high frequency: on the one hand, the performance indicators of the backup system, which are provided by the state monitoring module, including the central processor utilization rate, memory occupancy rate, communication port data throughput, control task queue length currently executed, and error log record of the backup system; on the other hand, real-time road traffic flow data, which is derived from roadside sensors and traffic flow detectors, including cross-section vehicle flow, average vehicle speed, vehicle queue length, lane occupancy rate and other multi-dimensional information. The adaptive control module analyzes these massive real-time data, and the analysis target is to predict the minimum operation time required for the backup system to complete all remaining control tasks it currently undertakes.

[0053] The process of predicting the minimum operation time depends on the evaluation of the processing capacity of the backup system and the complexity of the tasks. The adaptive control module establishes a task processing model inside, which converts each task in the remaining task queue into a standard calculation unit according to factors such as the number of control phases, signal cycle complexity, and coordination requirements. The adaptive control module estimates the number of standard calculation units that the backup system can process per unit time by combining the real-time monitored central processor utilization rate and memory performance of the backup system, and then calculates the minimum operation time required to complete all remaining tasks. If the predicted minimum operation time is significantly shorter than the time reserved for the planned recovery time, it indicates that the processing efficiency of the backup system is higher than expected, and there is a possibility of returning the control right to the main traffic signal control system in advance. At this time, the adaptive control module will start optimization calculation, which needs to be combined with the signal quality decay curve provided by the decay simulation module for comprehensive judgment. The signal quality decay curve of the decay simulation module describes the decay of the signal influence of the main traffic signal control system during the shutdown period.

[0054] The adaptive control module puts the idea of early resumption of the main traffic signal control system into precise calculation. The adaptive control module sets an optimization goal, which is to find a time point that enables the main traffic signal control system to improve the overall signal quality as soon as possible after resumption, while ensuring a successful handover of the backup system task. The adaptive control module takes a certain interval (for example, every minute) on the time axis from the current time to the originally planned resumption time as a candidate resumption time point. For each candidate resumption time point, the adaptive control module performs the following steps: first, it checks whether the backup system can complete the remaining tasks according to the predicted minimum operation time at the time point; then, based on the signal quality decay curve of the decay simulation module, it calculates the time required for the signal quality to recover to the preset good level after the main traffic signal control system resumes work from the candidate resumption time point; finally, it comprehensively evaluates the signal quality improvement benefit and potential operation risk brought by early resumption. The adaptive control module selects an optimal resumption time by comparing the comprehensive scores of all feasible candidate resumption time points. The optimal resumption time is a balance point that fully utilizes the surplus processing capacity of the backup system and takes into account the dynamic process of signal reconstruction after the main system resumes. The adaptive control module sends the calculated optimal resumption time to the switching decision module.

[0055] The switching decision module receives the optimal recovery time suggestion from the adaptive control module, and compares the optimal recovery time with the master system recovery time derived based on the low traffic period end time and the total operation time end time. The switching decision module has decision logic inside to evaluate the feasibility of adopting the new time. The decision logic verifies whether the optimal recovery time is still in a low traffic environment and whether it meets the basic requirements of system stability. After verification, the switching decision module updates the internal planned master system recovery time using the optimal recovery time provided by the adaptive control module. When the optimal recovery time arrives, the switching decision module triggers the system recovery process, controlling the master traffic signal control system to reactivate and switch control with the backup system. The data fusion module, as the cornerstone of system data flow, runs through all the above processes. The data fusion module continuously monitors the running parameters of the master traffic signal control system collected by the state monitoring module. The data fusion module checks the integrity of these running parameters, identifies whether there are data packet loss, sensor reading abnormal interruption or values significantly exceeding the reasonable range, etc. For the identified data anomalies, the data fusion module will start the data repair program, which may use time series interpolation, historical pattern-based filling or cross-validation and replacement using redundant sensor data. The data fusion module further fuses multi-source sensor data, including but not limited to electromagnetic induction coils, microwave radars, video cameras, GPS floating car data, etc. These sensors perceive traffic conditions from different spatial positions and different physical principles. The data fusion module uses filtering algorithms, data association techniques and feature-level fusion methods to integrate multi-source data into a consistent and reliable traffic state description.

[0056] The data fusion module uses the fused high-quality data to correct the fault occurrence time initially estimated by the fault prediction module, and the correction process is dynamic. For example, if the vibration sensor data shows that the abnormal vibration in a signal cabinet has intensified, and the temperature sensor reading has also risen, the data fusion module will integrate these information to judge that the hardware pressure has increased, and thus may advance the fault occurrence time originally estimated by the fault prediction module. The data fusion module outputs the corrected, more accurate fault occurrence time to the window calculation module. The window calculation module recalculates the time window allowed for switching based on the corrected fault occurrence time, and this recalculated time window more accurately reflects the actual risk window of the system. The overlap analysis module makes overlap judgment based on the updated time window and the low traffic period information of the period acquisition module. Since the start or end point of the time window may change due to the correction of the fault occurrence time, the judgment result of the overlap analysis module will also be more accurate.

[0057] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A dual-system traffic signal control system with automatic fault switching function, characterized in that, The system comprises a state monitoring module, a fault prediction module, a window calculation module, a time period acquisition module, an overlap analysis module, a time length prediction module, and a switching decision module; The state monitoring module continuously acquires the operating parameters of the main traffic signal control system and transmits the operating parameters to the fault prediction module to estimate the fault occurrence time; The fault prediction module outputs the fault occurrence time to the window calculation module, which calculates the time window for switching based on the fault occurrence time and the traffic peak time, wherein the end of the time window is the traffic peak time minus the time length required for signal stabilization; the time period acquisition module acquires low traffic flow period information, and the overlap analysis module receives the time window and the low traffic flow period and determines whether there is an overlapping period; If there is an overlapping period, the time length prediction module calculates the total operation time length for switching from the main system to the backup system; The switching decision module compares the overlapping period with the total operation time length, and when the overlapping period is equal to the total operation time length, the start point of the overlapping period is taken as the switching time, triggering the switching action.

2. The dual system traffic signal control system with automatic failover function according to claim 1, characterized in that, The time length prediction module is configured to perform: Continuously acquire the device operation history data of the main traffic signal control system, including fault occurrence frequency and system response delay time; Obtain the startup parameters of the backup system, including initialization time and signal synchronization time; Train the device operation history data and the startup parameters using a neural network model to generate dynamic weight coefficients; Based on the dynamic weight coefficients, the fault occurrence frequency, the system response delay time, the initialization time, and the signal synchronization time are fused and calculated to output the total operation time length.

3. The dual system traffic signal control system with automatic failover function according to claim 2, characterized in that, The window calculation module is configured to perform: when the overlapping period is greater than the total operation time length, a greedy algorithm is used to iteratively select multiple candidate sub-periods from the overlapping period, and the length of each candidate sub-period is not less than the total operation time length; Calculate the correlation score of each candidate sub-period with the traffic flow pattern, which is generated by statistical analysis of historical traffic data; Select the candidate sub-period set with the highest correlation score as the output.

4. The dual system traffic signal control system having a failure automatic switching function according to claim 3, characterized by The switching decision module is configured to perform: receive the candidate sub-period set; calculate the priority index of each candidate sub-period based on the real-time load rate of the main system, which is provided by the state monitoring module; Select the candidate sub-period with the highest priority index as the target period, and take the start point of the target period as the switching time.

5. The dual system traffic signal control system having a failure automatic switching function according to claim 3, characterized by, The system further comprises a stability evaluation module, which is configured to perform: obtain the signal stabilization threshold of the backup system after startup and the signal decay characteristics of the main system after shutdown; calculate the minimum overlapping time length required for the backup system to reach the signal stabilization threshold; if the minimum overlapping time length is less than or equal to the length of the candidate sub-period, select the candidate sub-period with the smallest deviation from the end of the low traffic flow period as the optimized period; The switching decision module takes the start point of the optimized period as the switching time.

6. The dual system traffic signal control system having a failure automatic switching function according to claim 2, characterized by The time length prediction module is configured to calculate a remaining operation requirement time length when the overlap time period is less than the total operation time length; the system further comprises a decay simulation module configured to generate a signal quality decay curve based on hardware characteristics of the main system; the switching decision module compares the signal quality decay curve with a preset signal quality standard curve to determine an effective maintenance time length of the main system after shutdown; If the remaining operation requirement time length falls within the effective maintenance time length, and the start of the low traffic volume time period and the sequence of the overlap time period and the remaining operation requirement time length do not exceed the end of the time window, a switching action is triggered.

7. The dual system traffic signal control system having a failover function according to claim 6, characterized by, The switching decision module is configured to monitor the running state of the backup system after completion of the switching action, obtain the end time of the total operation time length, and track traffic volume changes in real time; when the traffic volume is lower than a threshold value, the later of the end of the low traffic volume time period and the end time of the total operation time length is taken as a main system recovery time, and the main system is controlled to be reactivated.

8. The dual system traffic signal control system having a failover function according to claim 2, characterized by, The system further comprises a task scheduling module configured to obtain a predicted failure time sequence of the current batch of traffic control tasks, and generate a switching queuing sequence based on task priorities; for each task, a switching start time and a switching end time are calculated; if the switching end time of any task is later than the end of the time window, the queuing sequence is adjusted, and a recalculation module is triggered to update the time window.

9. The dual system traffic signal control system having a failover function according to claim 7, characterized by, The system further comprises an adaptive control module configured to collect performance indicators and traffic flow data of the backup system in real time within a monitoring period before the main system recovery time; based on the performance indicators, a minimum operation time length of the remaining tasks is predicted; if the minimum operation time length allows early recovery, an optimal recovery time is calculated in combination with the signal quality decay curve; the switching decision module updates the main system recovery time with the optimal recovery time.

10. The dual system traffic signal control system having a failover function according to claim 2, characterized by, The system further comprises a data fusion module configured to check the integrity of the running parameters collected by the state monitoring module, and fuse multi-source sensor data to correct the failure occurrence time; The corrected failure occurrence time is input into the window calculation module, the time window is regenerated, and the overlap judgment with the low traffic volume time period is ensured to be accurate.

Images