Intelligent traffic patrol shunting safety analysis system

Abstract

The application discloses a kind of wisdom traffic patrol shunt safety analysis systems.The system obtains the traffic operation data of multiple shunt monitoring nodes in shunt path, and the vehicle entering shunt path is divided into shunt vehicle batch;According to the actual travel time of released batch, the arrival time window of to-be-released batch at each node is predicted, the in-transit shunt vehicle load of each node before its arrival is determined, and then the node queuing state is predicted and the node receivable quantity is calculated;According to the vehicle transmission ratio of to-be-released batch, the receivable quantity of each node is converted into upstream safety release amount, the shunt safety constraint node and batch safety release upper limit are determined, and the upstream shunt ratio is adjusted accordingly.The application can avoid the future storage space occupied by in-transit vehicle being repeatedly allocated, and improve the safety and adaptability of traffic shunt control.

Description

Technical Field

[0001] This invention relates to the field of intelligent traffic management and road traffic safety control technology, specifically to an intelligent traffic patrol and diversion safety analysis system. Background Technology

[0002] When traffic incidents occur on highways, urban expressways, or other roads with continuous traffic flow, such as traffic accidents, road construction, vehicle breakdowns, severe weather, or reduced capacity of local roads, traffic management departments usually select diversion entrances upstream of the traffic incident to guide some vehicles to alternative roads and, after bypassing the area affected by the traffic incident, allow the vehicles to return to their original roads or enter the target roads.

[0003] Diverted vehicles may pass through exit ramps, toll plazas, signalized intersections, road narrowing points, and merging / returning areas in sequence along their diversion paths. Due to differences in road space, traffic flow composition, vehicle dispersal capacity, and existing traffic demand at each location, the number of diverted vehicles that can be safely accommodated at each location also varies.

[0004] Existing traffic diversion schemes typically determine the diversion path and initial diversion ratio based on road traffic flow, road capacity, estimated travel time, or path length before diversion implementation. After the diversion scheme is implemented, although vehicle queuing conditions along the diversion path can be obtained through road videos, radar, or drones, the diversion strategy is usually adjusted only based on the queue length or real-time traffic flow at the node.

[0005] However, vehicles released from the upstream diversion entrance require different travel times to reach different downstream nodes. When the system adjusts the upstream diversion ratio at the current moment, vehicles that have already entered the diversion path but have not yet reached the downstream node will continue to propagate downstream. If the upstream diversion flow is determined solely based on the remaining queuing space at the downstream node at the current moment, it is easy to reallocate future queuing space already occupied by vehicles en route to newly released vehicles.

[0006] For example, an alternative road intersection may still have some queuing space, but several batches of vehicles previously released from upstream diversion entrances have not yet arrived at the intersection. Once these en route vehicles arrive, the number of vehicles queuing at the intersection may approach or exceed the safe storage capacity. If the system continues to release new diversion vehicles based on the current queuing status at the intersection, the queue may cross the safety boundary, causing queue overflow at exit ramps, intersection congestion, return node congestion, or other traffic safety issues.

[0007] Furthermore, not all vehicles entering the diversion path from the upstream diversion entrance will necessarily reach every downstream node. Some vehicles may leave, turn, or enter other branch roads midway through the diversion path. Therefore, the number of vehicles that different nodes can accept cannot be directly compared, nor can it be directly used as the number of vehicles allowed to pass through the upstream diversion entrance.

[0008] Therefore, a smart traffic patrol and diversion safety analysis system and method are needed. The system uses the estimated arrival time of the batch of diverted vehicles to be released to each diversion monitoring node as the safety analysis benchmark. It includes diverted vehicles that have been released but have not yet arrived at the corresponding node in the prediction of the node's future queuing status, and converts the number of vehicles that different nodes can accept into a unified upstream safety release amount. This determines the diversion safety constraint nodes that restrict the entire diversion path, and adjusts the upstream diversion ratio accordingly. Summary of the Invention

[0009] The technical problem to be solved by this invention is: addressing the issue that existing traffic diversion control systems typically determine the upstream diversion volume based on the current traffic status of diversion monitoring nodes, without fully considering diverted vehicles that have been released from upstream but have not yet reached downstream nodes, resulting in the repeated allocation of future remaining safety space at nodes. This invention provides a smart traffic patrol and diversion safety analysis system and method.

[0010] To address the aforementioned technical problems, this invention provides an intelligent traffic patrol and diversion safety analysis system, comprising a patrol data acquisition device, a memory, a processor, and a diversion control device.

[0011] The patrol data acquisition device is used to acquire traffic operation data of multiple diversion monitoring nodes in the diversion path. Multiple diversion monitoring nodes are sequentially linked along the travel direction of the diverted vehicles to form a diversion monitoring node chain.

[0012] The memory is used to store the shunting monitoring node chain, the node safety storage limit of each shunting monitoring node, and the computer program.

[0013] When the processor executes the computer program, it is used to classify vehicles entering the diversion path from the upstream diversion entrance into diversion vehicle batches according to a preset diversion control cycle, and to distinguish between released diversion vehicle batches and unreleased diversion vehicle batches.

[0014] The processor determines the predicted arrival time window of the batch of vehicles to be released at each diversion monitoring node based on the actual travel time of the released diversion vehicle batches at each diversion monitoring node.

[0015] For each diversion monitoring node, the processor determines the load of diversion vehicles that have been released from the upstream diversion entrance but have not yet arrived at the diversion monitoring node before the batch of diversion vehicles to be released arrives at the diversion monitoring node.

[0016] Based on the traffic operation data of the diversion monitoring node, the load of diverted vehicles en route, and the inflow and outflow of vehicles within the predicted arrival time window, the processor predicts the node queuing status when the batch of diverted vehicles to be released arrives at the diversion monitoring node, and determines the number of batches of diverted vehicles that the diversion monitoring node can accept based on the node queuing status and the node's safe storage capacity.

[0017] The processor converts the number of vehicles that the node can accept into the upstream safe release quantity corresponding to the diversion monitoring node based on the predicted vehicle transfer ratio from the upstream diversion entrance to the diversion monitoring node in the batch of vehicles to be released.

[0018] The processor determines the safety release limit of the diversion safety constraint node and the batch safety release limit of the diversion vehicle batch to be released based on the upstream safety release amount corresponding to each diversion monitoring node, and outputs the corresponding diversion control command to the diversion control device.

[0019] The diversion control device is used to adjust the diversion ratio at the upstream diversion inlet according to the diversion control command.

[0020] This invention predicts the arrival time of the batch of diverted vehicles to be released at different diversion monitoring nodes, determines the load of diverted vehicles that each node will receive before their arrival, and then predicts the queuing status of the nodes when the batch of diverted vehicles to be released arrives. Based on the vehicle transfer ratio of each node, the number of vehicles that the node can accept is converted into the upstream safe release amount, so as to realize diversion safety control that takes into account vehicle propagation delay. Detailed Implementation

[0021] The present invention will be further described below with reference to specific embodiments. These embodiments are used to illustrate the technical principles of the present invention, and are not intended to limit the scope of protection of the present invention.

[0022] Without departing from the technical concept of this invention, the specific road type of the diversion path, the number of diversion monitoring nodes, the diversion control cycle, the method of collecting traffic operation data, and the method of predicting node queuing status can all be adjusted according to the actual road conditions.

[0023] Explanation of relevant terms used in this application:

[0024] 1. Diversion Path: A diversion path is a road route used to allow some vehicles to bypass the area affected by a traffic incident. Diversion paths may include exit ramps of the affected roads, local alternative roads, intersections, toll stations, return ramps, or other road sections for vehicles to detour.

[0025] 2. Upstream diversion entrance: The upstream diversion entrance is the location where diverted vehicles leave the affected road or enter the alternative road; the upstream diversion entrance can be a highway exit, an urban expressway exit, a main-auxiliary road interchange, an intersection entrance, or a temporary diversion location set up by traffic management equipment.

[0026] 3. Diversion monitoring nodes: including: diversion exit nodes; toll station nodes; signalized intersection nodes; unsignalized intersection nodes; alternative road narrowing nodes; nodes where diverted vehicles interweave or merge with existing vehicles; and return and merging nodes where diverted vehicles re-enter the original road or target road.

[0027] Each diversion monitoring node is assigned a unique node identifier and corresponds to a certain road space range.

[0028] 4. Diversion monitoring node chain: Multiple diversion monitoring nodes are sequentially linked according to the travel direction of the diverted vehicles to form a diversion monitoring node chain.

[0029] For example, if a diversion path includes an exit ramp node N1, a local road intersection node N2, and a return merging node N3, then the diversion monitoring node chain is represented as: N1→N2→N3.

[0030] The node chain records at least the sequential order of each node and the road connection relationships.

[0031] When path branches exist, a separate chain of diversion monitoring nodes can be established for each diversion branch, or a node network with branching relationships can be established. For a batch of vehicles to be released, only the nodes it is expected to pass through are selected for the calculation of the safe release volume.

[0032] 5. Diverted vehicle batch: A diverted vehicle batch refers to the set of vehicles that enter the diverted path from the upstream diverted entrance within a preset diverted control cycle.

[0033] Each diverted vehicle batch must be associated with at least: batch identifier; batch release start time; batch release end time; and the number of vehicles released upstream.

[0034] Batches that have already entered the diversion path are called released diversion vehicle batches; batches that have not yet entered the diversion path and whose allowed release quantity is being calculated are called pending release diversion vehicle batches.

[0035] Diverted vehicle batches can be recorded using license plate recognition results, vehicle trajectory markers, vehicle-road cooperative communication markers, or according to traffic flow statistics at diversion entrances.

[0036] 6. Node safety boundary: The node safety boundary is the road position that the tail of the queue of the diversion monitoring node should not cross.

[0037] For example, the node safety boundary of an exit ramp node can be set before queuing vehicles enter the main highway; the node safety boundary of an intersection node can be set before queuing vehicles block upstream intersections, bus stops, fire lanes, or important entrances and exits; and the node safety boundary of a return and merging node can be set before queuing vehicles enter the upstream weaving area.

[0038] 7. Node Safety Storage Limit: The node safety storage limit represents the maximum number of queued vehicles that a diversion monitoring node can accommodate, provided that the tail of the queue does not exceed the node's safety boundary.

[0039] The maximum safe parking limit for a node can be determined based on the effective queuing length between the vehicle's dissipation location and the node's safe boundary, the number of queuing lanes, and the average length of vehicles occupied while queuing.

[0040] Let Li be the effective queuing length of node Ni, ni be the number of queuing lanes, and li be the average occupancy length of vehicles in the queuing state. Then, the upper limit of safe vehicle storage Mi of node Ni can be determined by the following formula:

[0041] .

[0042] The average vehicle occupancy length includes the length of the vehicle itself and the average distance between adjacent vehicles in the queue.

[0043] When the diverted vehicles include different types of vehicles such as large buses and trucks, the upper limit of safe storage capacity at each node can be determined by converting each type of vehicle into a standard vehicle unit.

[0044] The upper limit of safe parking at nodes can also be determined based on road design data, historical queuing data, on-site calibration results, or traffic simulation results.

[0045] 8. On-the-Way Diversion Vehicle Load: For a certain diversion monitoring node, the number of vehicles that have been released from the upstream diversion entrance but have not yet arrived at the diversion monitoring node at the current control time, and are expected to arrive before the batch of diversion vehicles to be released arrives at the node, constitutes the on-the-way diversion vehicle load of that node.

[0046] The same batch of released diverted vehicles may have all passed through upstream nodes, but downstream nodes may still have vehicles en route that have not yet arrived.

[0047] Therefore, the load of diverted vehicles on the way needs to be determined for each diversion monitoring node.

[0048] 9. Predicted vehicle transfer ratio: The predicted vehicle transfer ratio represents the ratio between the number of vehicles expected to arrive at a certain diversion monitoring node in a batch of vehicles to be released and the number of vehicles released at the upstream diversion entrance of that batch.

[0049] The predicted vehicle transfer ratio may differ for different diversion monitoring nodes.

[0050] For example, if 100 diverted vehicles are released upstream, of which 100 are expected to pass through the exit ramp node, 80 through the local road intersection node, and 60 to reach the return merging node, the corresponding predicted vehicle transfer ratios are 1, 0.8, and 0.6, respectively.

[0051] 10. Diversion Safety Constraint Node: The diversion safety constraint node is the node that imposes the strictest restriction on the upstream safety release volume of the batch of vehicles to be diverted among all diversion monitoring nodes.

[0052] In one implementation, the diversion monitoring node with the smallest upstream safety release amount is determined as the diversion safety constraint node.

[0053] Diversion safety constraint nodes can shift depending on the queuing status, vehicle dissipation capacity, on-the-way vehicle load, and changes in background traffic flow at each node.

[0054] The intelligent traffic patrol and diversion safety analysis system includes a patrol data acquisition device, a memory, a processor, and a diversion control device. The patrol data acquisition device, memory, and diversion control device are all communicatively connected to the processor.

[0055] The patrol data acquisition device is used to acquire traffic operation data from multiple diversion monitoring nodes; the patrol data acquisition device includes one or more of the following: roadside video equipment, millimeter-wave radar, lidar, drones, road patrol vehicles, geomagnetic detectors, vehicle-road cooperative roadside units, and vehicle trajectory data interfaces.

[0056] Roadside video equipment is installed at the entrance roads, queuing areas, or vehicle dispersal locations of diversion monitoring nodes to identify the number of vehicles entering, exiting, queuing vehicles, and the position of the end of the queue.

[0057] Millimeter-wave radar or lidar is used to obtain vehicle speed, vehicle position, headway, and queue length.

[0058] The drone flies to the diversion monitoring node that lacks fixed roadside equipment according to the patrol mission, and obtains node images or vehicle trajectory data.

[0059] Road patrol vehicles can obtain on-site video of nodes, the position of the end of the queue, and the road traffic status.

[0060] The vehicle-road cooperative roadside unit can receive vehicle identification, vehicle location, vehicle speed, and vehicle direction of travel information.

[0061] The memory is used to store: diversion paths; diversion monitoring node chains; node locations and node types of each diversion monitoring node; node safety boundaries of each diversion monitoring node; node safe storage limits for each diversion monitoring node; diversion vehicle batch data; actual travel time data; vehicle transfer ratio data; and node traffic operation data.

[0062] The processor can be a server processor, an industrial computer processor, an edge computing processor, a cloud computing platform processor, or a distributed processing system consisting of multiple processing devices.

[0063] Diversion control devices include at least one of the following: variable message signs, navigation information dissemination devices, lane indicators, ramp control equipment, and traffic signal equipment; wherein, variable message signs can disseminate detour prompts, diversion path information, or diversion control information; navigation information dissemination devices can disseminate detour suggestions to vehicle navigation terminals and control the recommended range or intensity of diversion paths according to the target diversion ratio; lane indicators can guide vehicles into corresponding diversion lanes; ramp control equipment can control the speed at which vehicles enter exit ramps or alternative roads; and traffic signal equipment can control the number of vehicles passing through diversion entrances or diversion path nodes by adjusting signal phases or green light durations.

[0064] This embodiment also provides a method for intelligent traffic patrol and diversion safety analysis, which includes the following steps.

[0065] S101, Obtain the diversion path and establish a diversion monitoring node chain.

[0066] Acquire information on road traffic incidents, road topology, and traffic diversion plans; the traffic diversion plan should include at least the upstream diversion entrance and diversion path, and select road locations that may form queue overflows as diversion monitoring nodes in the diversion path.

[0067] Select diversion monitoring nodes based on at least one of the following conditions: changes in road capacity; reduction in the number of road lanes; vehicles needing to stop for payment or inspection; vehicles being controlled by traffic signals; merging, diverging, or weaving between diverted and non-diverted vehicles; road queuing potentially affecting the upstream mainline or adjacent intersections; sequentially link multiple diversion monitoring nodes according to the direction of travel of diverted vehicles to form a diversion monitoring node chain; determine the node safety boundary and node safe parking limit for each diversion monitoring node; this step outputs the diversion monitoring node chain and the basic safety parameters of each node.

[0068] S102, Generate diversion vehicle batches

[0069] Set a diversion control cycle, which is determined based on the traffic flow change speed, diversion path length, travel time between nodes, and patrol data update frequency. For example, it can be 1 to 15 minutes. Count the vehicles that enter the diversion path from the upstream diversion entrance within each diversion control cycle. Determine the vehicles that enter the diversion path within the same diversion control cycle as a diversion vehicle batch.

[0070] For example, if a diversion control cycle is 5 minutes, then vehicles entering the diversion path between 10:00 and 10:05 constitute a diversion vehicle batch.

[0071] Record the batch identifier, release start time, release end time, and number of vehicles released upstream for each diversion batch; the number of vehicles at the diversion entrance can be determined by video counting, radar counting, license plate recognition, electronic toll collection records, or vehicle trajectory data.

[0072] S103, determine the actual travel time of the batches of vehicles that have been released for diversion.

[0073] For batches of diverted vehicles that have been released from the upstream diversion entrance, obtain the times when the vehicles in that batch passed through the upstream diversion entrance and each diversion monitoring node.

[0074] When vehicle-by-vehicle identification is possible, the passage records of the same vehicle at the upstream diversion entrance and each diversion monitoring node can be matched based on the license plate number, vehicle communication identifier, or vehicle trajectory identifier.

[0075] After matching multiple vehicles in the same diversion batch, the actual travel time interval from the upstream diversion entrance to each diversion monitoring node of that batch is determined.

[0076] For example, for a diversion vehicle batch Bk, the shortest actual travel time from the upstream diversion entrance to node Ni is Ti,k,min, and the longest actual travel time is Ti,k,max. Then the corresponding actual travel time interval for this batch is:

[0077] [Ti,k,min,Ti,k,max].

[0078] In cases where vehicle-by-vehicle identification is not possible, the actual travel time interval of diverted vehicle batches can be estimated based on the batch release time at the upstream diversion entrance, the time of increase in vehicle flow at the node, the length of the road segment, and the actual measured speed of the road segment.

[0079] S104, Predicted arrival time window for batches of vehicles to be released for diversion.

[0080] Based on the actual travel time intervals of one or more released diversion vehicle batches, determine the predicted travel time intervals from the upstream diversion entrance to each diversion monitoring node. The predicted travel time interval uses the actual travel time interval of the most recent complete diversion vehicle batch, or it can be a weighted result of the actual travel time intervals of multiple historical diversion vehicle batches.

[0081] Let the planned release time interval for the batch Br of vehicles to be released be:

[0082] [tr,s,tr,e];

[0083] The predicted travel time interval for this batch from the upstream diversion inlet to node Ni is:

[0084] [Ti,r,min,Ti,r,max];

[0085] The predicted arrival time window of the diverted vehicle batch Br at node Ni can then be determined as follows:

[0086] Wi,r=[tr,s+Ti,r,min, tr,e+Ti,r,max].

[0087] Different diversion monitoring nodes have different distances from upstream diversion entrances and different road operating conditions, thus corresponding to different predicted arrival time windows.

[0088] For example, the same batch of vehicles to be released for diversion can arrive at the exit ramp node 2 to 7 minutes after release, at the local road intersection node 10 to 16 minutes after release, and at the return merging node 20 to 28 minutes after release.

[0089] S105, acquire traffic operation data from diversion monitoring nodes.

[0090] Traffic operation data for each diversion monitoring node is acquired through the patrol data acquisition device. The traffic operation data includes at least: the number of vehicles currently in the queue; the number of vehicles entering; the number of vehicles leaving; and the flow of non-diversion vehicles. The traffic operation data also includes the position of the end of the queue, vehicle speed, lane occupancy, vehicle type, and signal control status.

[0091] The current number of vehicles in the queue can be determined by vehicle target detection and tracking, or by the difference between the cumulative number of vehicles entering the node and the cumulative number of vehicles exiting the node.

[0092] The position of the end of the queue can be determined based on the spatial distribution of the queued vehicles and is used to verify whether the current number of queued vehicles is close to the node's safe storage limit.

[0093] When the origin of a vehicle can be identified, diverted vehicles and non-diverted vehicles can be directly distinguished based on the vehicle's identity and trajectory.

[0094] When it is not possible to distinguish each vehicle individually, the flow of non-diverted vehicles can be predicted based on the background traffic flow before the diversion implementation, the detection data at the entrance of adjacent roads, or the historical traffic flow at the same time.

[0095] S106, Confirming the batches of vehicles being diverted en route.

[0096] For a batch of diverted vehicles Br to be released and any diversion monitoring node Ni, the batch of diverted vehicles in transit corresponding to node Ni is determined based on the predicted arrival time window of the batch of diverted vehicles to be released at node Ni.

[0097] Specifically, from all released diversion vehicle batches, select diversion vehicle batches that meet the following conditions:

[0098] It has already been released from the upstream diversion inlet;

[0099] Not all nodes Ni have been reached at the current control time;

[0100] The remaining vehicles are expected to arrive at node Ni before the batch of vehicles to be released for diversion, Br, arrives at node Ni.

[0101] The released diversion vehicle batches that meet the above conditions constitute the set of in-transit batches of node Ni relative to the unreleased diversion vehicle batch Br.

[0102] Because the predicted arrival times of different nodes are different, the en route status of the same batch of released diverted vehicles may be different for different nodes.

[0103] For example, a released batch Bk may have passed entirely through node N1, but only partially reached node N2, and has not yet reached node N3. In this case, batch Bk does not constitute an in-transit batch for node N1, but it can constitute an in-transit batch for nodes N2 and N3.

[0104] S107, Determine the load of vehicles being diverted en route.

[0105] For each batch of diverted vehicles en route, determine the number of vehicles in that batch that have not yet arrived at the corresponding diversion monitoring node.

[0106] Let q0,k be the number of vehicles released from the upstream diversion entrance in the already released diversion vehicle batch Bk, and pi,k be the predicted or actual vehicle transfer ratio to node Ni. Then the total number of vehicles expected to reach node Ni in this batch is:

[0107] qi,k = q0,k × pi,k.

[0108] Let qi,k,arrived be the number of vehicles that have already arrived at node Ni. Then the number of vehicles in this batch that have not yet arrived at node Ni can be determined as follows:

[0109] qi,k,remain=max{0,qi,k-qi,k,arrived}.

[0110] The number of vehicles in each batch that have not yet arrived corresponding to node Ni is merged to obtain the in-transit vehicle load of that node:

[0111] Ii,r=Σqi,k,remain.

[0112] When using interval forecasting, the lower and upper limits of the load of diverted vehicles on the way can be determined separately, and the upper limit value can be used for safety analysis.

[0113] S108, Predicting the basic queuing status of nodes before the arrival of batches of vehicles to be released for diversion.

[0114] For node Ni, the predicted time period is the time from the current control time to the expected arrival time of the diverted vehicle batch Br at node Ni.

[0115] Let mi,0 be the number of vehicles queuing at node Ni at the current control time, Ii,r be the load of diverted vehicles en route, Bi,r,before be the number of non-diverted vehicles expected to arrive at node Ni during the prediction period, and Di,r,before be the number of vehicles expected to depart from node Ni. Then, the predicted basic number of vehicles queuing before the arrival of the batch of diverted vehicles to be released can be determined by the following formula:

[0116] mi,r,pred=max{0,mi,0+Ii,r+Bi,r,before-Di,r,before}.

[0117] The predicted base queue number represents the number of queued vehicles expected to exist in the node when the batch of vehicles to be released arrives, excluding the batch of vehicles to be released and diverted.

[0118] The estimated number of non-diverted vehicles can be determined based on the current non-diverted vehicle flow rate and the length of the predicted time period, or based on historical traffic data for the same period and current traffic trends.

[0119] The estimated number of vehicles leaving can be determined based on the actual departure rate of the node in the most recent statistical periods, traffic signal timing, number of open lanes, number of toll lanes, or downstream traffic capacity.

[0120] In one implementation, the average departure rate over the most recent statistical periods is calculated, and the average departure rate is multiplied by the predicted time period length to obtain the expected number of vehicles to depart.

[0121] In another implementation, the estimated number of vehicles exiting the intersection node is calculated based on the signal cycle, effective green light time, saturation flow rate, and number of available lanes.

[0122] S109, Determine the number of nodes that can be accommodated.

[0123] Let Mi be the upper limit of the safe storage capacity of node Ni.

[0124] In addition to predicting the number of vehicles in the basic queue, it is also necessary to consider the number of non-diverted vehicles and the number of vehicles leaving the node within the predicted arrival time window of the batch of vehicles to be released for diversion.

[0125] Let Bi,r,window be the expected number of non-diverted vehicles arriving within the predicted arrival time window of the batch of vehicles to be released for diversion, and Di,r,window be the expected number of vehicles leaving. Then, the number of nodes ai,r that node Ni can accept for the batch of vehicles to be released for diversion, br can be determined by the following formula:

[0126] ai,r=max{0,Mi-mi,r,pred-Bi,r,window+Di,r,window}.

[0127] The node's acceptable number represents the number of vehicles that node Ni can receive during the arrival of a batch of vehicles awaiting release without causing the expected number of vehicles in the queue to exceed the node's safe storage capacity.

[0128] When the calculation result is less than or equal to zero, it means that node Ni no longer has safe storage space to receive batches of vehicles to be released and diverted.

[0129] At this time, the upstream safety release amount corresponding to the node can be set to zero, and control commands to reduce or pause the diversion can be output.

[0130] S110, Determine the predicted vehicle transfer ratio

[0131] For a batch of vehicles that has completed the diversion through node Ni, count the actual number of vehicles released at the upstream diversion entrance and the actual number of vehicles arriving at node Ni.

[0132] Let q0,k be the actual number of vehicles released at the upstream diversion entrance of the diversion vehicle batch Bk, and qi,k be the actual number of vehicles arriving at node Ni. Then the actual vehicle transfer ratio at node Ni for this batch is:

[0133] pi,k = qi,k / q0,k.

[0134] The predicted vehicle transfer ratio for the batch of vehicles to be released and diverted to arrive at node Ni can be obtained by using the actual vehicle transfer ratio corresponding to the most recent complete batch.

[0135] When there are significant changes in the current road operation status, such as the temporary closure of a branch road, changes in the scope of traffic control, or changes in vehicle navigation strategies, the complete batch data after the change can be used first to determine the predicted vehicle transmission ratio.

[0136] When the predicted vehicle transfer ratio is zero or lower than the preset effective ratio, it means that the batch of vehicles to be released and diverted will not pass through or will not pass through the corresponding node. This node may not participate in the determination of the safe release limit of the current batch.

[0137] S111 converts the number of nodes that can be accommodated into the upstream safe release amount.

[0138] The number of vehicles a node can accept represents the number of vehicles arriving at node Ni, while the diversion control device directly controls the number of vehicles released from the upstream diversion entrance.

[0139] Therefore, based on the predicted vehicle transfer ratio from the upstream diversion entrance to node Ni in the batch of vehicles to be released, the number of vehicles that the node can accept is converted into the upstream safe release quantity.

[0140] Let ai,r be the number of vehicles that node Ni can accept for the batch Br of vehicles to be released and diverted, and let pi,r,pred be the predicted vehicle transfer ratio. Then, the upstream safe release quantity ui,r corresponding to node Ni can be determined according to the following formula:

[0141] ui,r = ai,r / pi,r,pred.

[0142] For example, if a node can accommodate 40 vehicles waiting to be released for diversion, and it is estimated that 80% of the vehicles released from the upstream diversion entrance will arrive at this node, then the corresponding upstream safe release capacity for this node is 50 vehicles.

[0143] Through the above conversion, the vehicle acceptance capacity of different diversion monitoring nodes is uniformly converted into the number of vehicles released at the upstream diversion entrance.

[0144] S112, Determine the diversion safety constraint node and batch safety release limit.

[0145] Calculate the upstream safety release amount corresponding to each effective diversion monitoring node in the diversion monitoring node chain.

[0146] The minimum value among the upstream safety release amounts is determined as the basic safety release limit for the batch of diverted vehicles to be released:

[0147] Ur=min{u1,r,u2,r,…,um,r}.

[0148] The diversion monitoring node corresponding to the minimum value is determined as the diversion safety constraint node.

[0149] This means that when the number of vehicles released at the upstream diversion entrance exceeds the minimum value, the diversion safety constraint node is expected to reach or exceed the node's safe vehicle storage limit first.

[0150] In one implementation, the final batch safety release limit is determined based on the base safety release limit and the safety correction factor:

[0151] Qr,safe=η×Ur;

[0152] Where η is a safety correction coefficient that is greater than zero and not greater than 1.

[0153] The safety correction factor is used to reserve a safety margin for errors in vehicle arrival time prediction, changes in traffic flow, and equipment detection errors.

[0154] When the upstream safety release amounts corresponding to multiple diversion monitoring nodes are the same or the difference meets the preset proximity conditions, the node that is close to the upstream diversion entrance along the diversion vehicle's travel direction can be identified as the priority safety constraint node.

[0155] Prioritizing nodes near upstream diversion entrances helps reduce the risk of queued vehicles overflowing back onto affected roads or the upstream mainline.

[0156] S113, Determine the diversion ratio and execute diversion control.

[0157] Obtain the number of vehicles Qr expected to arrive at the upstream diversion entrance within the control period corresponding to the batch of vehicles to be released for diversion.

[0158] The target number of diverted vehicles is determined based on the maximum safe release limit for each batch:

[0159] Qr,d = min{Qr, Qr,safe}.

[0160] The diversion ratio Rr corresponding to the batch of vehicles to be released for diversion is:

[0161] Rr = Qr,d / Qr.

[0162] When the expected number of vehicles arriving at the upstream diversion entrance is zero, the diversion ratio can be set to zero or the control state of the previous control cycle can be maintained.

[0163] The processor generates a shunting control instruction based on the shunting ratio and sends the shunting control instruction to the shunting control device.

[0164] Diversion control devices can implement control in one or more of the following ways: issuing diversion prompts through variable message signs; recommending diversion routes to some vehicles through navigation information dissemination devices; guiding vehicles into diversion lanes through lane indicators; limiting the number of vehicles entering the diversion route per unit time through ramp control equipment; and adjusting the green light duration or release frequency of the diversion direction through traffic signal equipment.

[0165] S114, update the batch of vehicles to be released for diversion based on actual operation results.

[0166] After releasing the batch of vehicles to be diverted according to the adjusted diversion ratio, the batch is converted into a batch of vehicles that have been released.

[0167] The actual arrival time, actual number of arriving vehicles, and actual queuing status of the batch at each diversion monitoring node are obtained through the patrol data acquisition device.

[0168] The predicted travel time interval from the upstream diversion inlet to each diversion monitoring node is updated based on the actual arrival time.

[0169] The predicted vehicle transfer ratio of each diversion monitoring node is updated based on the actual number of arriving vehicles and the actual number of vehicles released upstream.

[0170] Based on the number of vehicles queuing before and after the arrival of this batch, the number of vehicles entering and leaving the nodes, the estimated vehicle exit capacity and non-diverted vehicle flow of the nodes are updated.

[0171] Vehicles that have not yet reached the downstream diversion monitoring nodes will be included in the load of the diversion vehicles in transit corresponding to the next batch of diversion vehicles to be released.

[0172] For the next batch of vehicles to be released for diversion, the following steps are performed again: determining the predicted arrival time window, determining the load of vehicles in transit for diversion, predicting the queuing status of nodes, determining the number of vehicles that nodes can accept, converting the upstream safe release volume, and adjusting the diversion ratio.

[0173] This forms a continuous closed loop for intelligent traffic patrol, diversion, and safety analysis.

[0174] The following example of traffic diversion security analysis will be used to explain this application:

[0175] A traffic accident occurred ahead on a highway. The traffic management platform determined that the vehicle exited the highway from the upstream exit, detoured via local roads, and then returned to the highway from the downstream entrance.

[0176] The routing path includes:

[0177] Node N1: Highway exit ramp;

[0178] Node N2: Local road signalized intersection;

[0179] Node N3: Highway return and merging node.

[0180] Nodes N1, N2, and N3 are sequentially linked according to the vehicle's direction of travel to form a diversion monitoring node chain:

[0181] N1→N2→N3.

[0182] The preset diversion control cycle is 5 minutes.

[0183] The current control time is 10:00. It is necessary to determine the safe release quantity of the diversion vehicle batch B5, which is scheduled to be released between 10:00 and 10:05.

[0184] Based on the actual travel times of previously released diversion vehicle batches, the predicted arrival time windows for batch B5 at each node are determined as follows:

[0185] Node N1: 10:02 to 10:07;

[0186] Node N2: 10:10 to 10:16;

[0187] Node N3: 10:20 to 10:28.

[0188] For node N2, the current number of vehicles in the queue is 30.

[0189] The batches of diverted vehicles B3 and B4 that have been released have not yet all arrived at node N2.

[0190] Based on the number of vehicles released upstream of batches B3 and B4, the predicted vehicle transfer ratio, and the number of vehicles that have already arrived at node N2, it is determined that there are still 25 diverted vehicles en route that will arrive at node N2 before batch B5 arrives at node N2.

[0191] It is estimated that between 10:00 and 10:10, 15 non-diversion vehicles will arrive at node N2 and 20 vehicles will leave node N2.

[0192] Therefore, the predicted base queue number of vehicles before batch B5 arrives at node N2 is:

[0193] 30 + 25 + 15 - 20 = 50 vehicles.

[0194] The maximum safe storage capacity for node N2 is 90 vehicles.

[0195] Within the predicted arrival time window of batch B5, it is estimated that 12 non-diverted vehicles will arrive at node N2, and 18 vehicles will depart from node N2.

[0196] Therefore, the number of nodes that node N2 can accept for batch B5 is:

[0197] 90 - 50 - 12 + 18 = 46 vehicles.

[0198] Based on the actual operation results of several complete batches in the past, the predicted vehicle delivery ratio for batch B5 to node N2 is determined to be 0.8.

[0199] The upstream safety release amount corresponding to node N2 is:

[0200] 46 ÷ 0.8 = 57.5 vehicles.

[0201] Calculated in the same way:

[0202] The upstream safe release quantity corresponding to node N1 is 80 vehicles;

[0203] The upstream safe release quantity corresponding to node N2 is 57.5 vehicles;

[0204] The upstream safe release quantity corresponding to node N3 is 65 vehicles.

[0205] Therefore, node N2 corresponds to the smallest upstream safety release amount, and node N2 is determined as the diversion safety constraint node for batch B5.

[0206] Assuming a safety correction factor of 0.9, the maximum safe release limit for batch B5 is:

[0207] 57.5 × 0.9 = 51.75 vehicles.

[0208] The target number of diverted vehicles can be rounded up to 52.

[0209] If 120 vehicles are expected to arrive at the upstream diversion entrance within the control period corresponding to batch B5, then the diversion ratio is approximately:

[0210] 52 ÷ 120 ≈ 43.3%.

[0211] The system uses variable message signs, navigation information dissemination devices, or ramp control equipment to allow approximately 43.3% of vehicles to enter the diversion route.

[0212] After batch B5 is released, the system continues to obtain its actual arrival time and the actual number of vehicles arriving at nodes N1, N2 and N3, in order to update the prediction parameters for the next batch B6.

[0213] Considering multiple batches of diverted vehicles en route:

[0214] In another implementation, the current control time is t0, and the batch of diverted vehicles to be released is Br.

[0215] Before the unreleased batch Br arrives at node Ni, some vehicles from the already released batches Br-3, Br-2, and Br-1 may arrive at node Ni.

[0216] in:

[0217] The total number of vehicles in batch Br-3 expected to arrive at node Ni is 40, of which 35 have arrived and 5 are still en route;

[0218] The total number of vehicles in batch Br-2 expected to arrive at node Ni is 50, of which 20 have already arrived and 30 are still en route.

[0219] The total number of vehicles in batch Br-1 expected to arrive at node Ni is 45, and there are 45 vehicles en route that have not yet arrived at node Ni.

[0220] Then the load of the diverted vehicles in transit corresponding to the batch Br to be released at node Ni is:

[0221] 5 + 30 + 45 = 80 vehicles.

[0222] When predicting the future queuing status of a node, the 80 diverted vehicles en route are included in the calculation.

[0223] This avoids judging the remaining space of a node solely based on the current number of vehicles in the queue, while ignoring vehicles that have entered the diversion path but have not yet arrived at the node.

[0224] Vehicle transfer ratio update:

[0225] In a certain diversion path, there is a branch road that exits midway between the upstream diversion entrance and node N3.

[0226] The data for the three most recent complete batches of diverted vehicles are as follows:

[0227] The first historical batch released 100 vehicles upstream, of which 60 vehicles arrived at node N3, with a vehicle delivery ratio of 0.60.

[0228] The second historical batch released 120 vehicles upstream, of which 78 vehicles arrived at node N3, with a vehicle delivery ratio of 0.65.

[0229] The third historical batch released 110 vehicles upstream, of which 77 reached node N3, with a vehicle delivery ratio of 0.70.

[0230] To improve the adaptability of the forecast results to recent traffic conditions, weights of 1, 2 and 3 were assigned to the three historical batches, respectively.

[0231] The predicted vehicle transfer ratio for the batch of vehicles to be released and diverted to arrive at node N3 is:

[0232] (1×0.60+2×0.65+3×0.70)÷(1+2+3)=0.667.

[0233] When node N3 can accommodate 40 vehicles awaiting diversion, its corresponding upstream safe release capacity is approximately:

[0234] 40 ÷ 0.667 ≈ 60 vehicles.

[0235] If a detour is temporarily closed, causing all subsequent diverted vehicles to pass through node N3, the predicted vehicle transfer ratio of node N3 can be updated to 1 based on the road control status, or the new actual vehicle transfer ratio can be used after obtaining new complete batch data.

Claims

1. A smart traffic patrol and diversion safety analysis system, comprising a patrol data acquisition device, a memory, a processor, and a diversion control device, characterized in that: The patrol data acquisition device is used to acquire traffic operation data of multiple diversion monitoring nodes in the diversion path. The multiple diversion monitoring nodes are sequentially associated along the driving direction of the diverted vehicles to form a diversion monitoring node chain. The memory is used to store the diversion monitoring node chain, the node safety storage limit of each diversion monitoring node, and the computer program; When the processor executes the computer program, it is used to: According to the preset diversion control cycle, vehicles entering the diversion path from the upstream diversion entrance are divided into diversion vehicle batches, and the batches of diversion vehicles that have been released and the batches of diversion vehicles to be released are distinguished. Based on the actual travel time of the released diversion vehicle batches at each diversion monitoring node, the predicted arrival time window of the batches to be released diversion vehicles at each diversion monitoring node is determined. For each diversion monitoring node, determine the load of diversion vehicles that have been released from the upstream diversion entrance and have not yet arrived at the diversion monitoring node before the batch of diversion vehicles to be released arrives at the diversion monitoring node; Based on the traffic operation data of the diversion monitoring node, the load of the diverted vehicles en route, and the vehicle inflow and outflow within the predicted arrival time window, the node queuing status when the batch of diverted vehicles to be released arrives at the diversion monitoring node is predicted, and the number of vehicles that the diversion monitoring node can accept for the batch of diverted vehicles to be released is determined based on the node queuing status and the node's safe storage limit. Based on the predicted vehicle transfer ratio of the batch of vehicles to be released from the upstream diversion entrance to the diversion monitoring node, the number of vehicles that the node can accept is converted into the upstream safe release quantity corresponding to the diversion monitoring node. Based on the upstream safety release amount corresponding to each diversion monitoring node, determine the diversion safety constraint node and the batch safety release limit of the batch of diverted vehicles to be released, and output the corresponding diversion control command to the diversion control device. The diversion control device is used to adjust the diversion ratio of the upstream diversion inlet according to the diversion control command.

2. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, The diversion monitoring nodes include at least one of diversion exit nodes, alternative road access nodes, and return merging nodes; Each diversion monitoring node is assigned a corresponding node safety boundary, and the upper limit of the node's safe vehicle storage is determined based on the effective queuing space between the vehicle dissipation location and the node's safety boundary.

3. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, Each batch of diverted vehicles must be associated with at least a batch identifier, a batch release start time, a batch release end time, and the number of vehicles released upstream. Vehicles that enter the diversion path from the upstream diversion entrance within a preset diversion control cycle are identified as a diversion vehicle batch.

4. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, The processor is also used for: Based on the actual passage times of the released diversion vehicle batches at the upstream diversion entrance and each diversion monitoring node, the corresponding actual travel time intervals are determined; Update the predicted travel time interval from the upstream diversion entrance to each diversion monitoring node based on the actual travel time interval of one or more released diversion vehicle batches; Based on the planned release time interval and the predicted travel time interval of the batch of vehicles to be released for diversion, the predicted arrival time window of the batch of vehicles to be released for diversion at each diversion monitoring node is determined.

5. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, When the processor determines the load of the diverted vehicles en route, it is used to: From the released diversion vehicle batches, determine the diversion vehicle batches that are en route and are expected to arrive at the diversion monitoring node after the current control time and before the batches of diversion vehicles to be released arrive at the corresponding diversion monitoring node; The number of vehicles that have not yet arrived at the diversion monitoring node is determined based on the number of vehicles released upstream in each batch of diverted vehicles, the corresponding vehicle transfer ratio, and the number of vehicles that have already arrived at the diversion monitoring node. The number of vehicles in each diversion batch that has not yet arrived at the diversion monitoring node is merged to obtain the diversion vehicle load.

6. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, The traffic operation data includes the current number of vehicles in the queue, the number of vehicles entering, the number of vehicles leaving, and the flow of non-diverted vehicles; The processor determines the predicted basic queue number of vehicles when the batch of vehicles to be released arrives at the corresponding diversion monitoring node based on the current number of queued vehicles, the load of the diverted vehicles en route, the estimated number of non-diverted vehicles arriving before the batch of vehicles to be released arrives, and the estimated number of vehicles leaving. The number of vehicles a node can accommodate is determined based on the node's safe storage capacity, the predicted base queue size, and the expected arrival and departure of non-diverted vehicles within the predicted arrival time window.

7. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, The predicted vehicle transfer ratio is determined based on the actual number of vehicles arriving at the corresponding diversion monitoring node in one or more batches of diverted vehicles and the actual number of vehicles released at the upstream diversion entrance. The upstream safety release amount is determined based on the ratio of the number of nodes that can be accommodated to the predicted vehicle transmission ratio. When the predicted vehicle transfer ratio indicates that the batch of vehicles to be released does not pass through the corresponding diversion monitoring node, the diversion monitoring node will not participate in the determination of the safe release limit of the batch.

8. The intelligent traffic patrol and diversion safety analysis system according to claim 1, characterized in that, The processor is used to determine the minimum value among the upstream safety release amounts corresponding to each diversion monitoring node as the upper limit of the batch safety release, and to determine the diversion monitoring node corresponding to the minimum value as the diversion safety constraint node; When the upstream safety release amount corresponding to multiple diversion monitoring nodes meets the preset proximity condition, the diversion monitoring node that is close to the upstream diversion entrance along the diversion vehicle travel direction is determined as the priority safety constraint node. The diversion control device includes at least one of variable message signs, navigation information dissemination devices, lane indicators, ramp control equipment, and traffic signal equipment.