Method and apparatus for enhancing vehicle navigation using lane level road congestion

Abstract

A method of enhancing vehicle navigation in conjunction with lane level road congestion includes: collecting vehicle data from a fleet; identifying a lane-level distribution of a plurality of vehicles operating on a road including a plurality of lanes, the plurality of vehicles including a host vehicle; comparing the lane level distribution with non-congested lanes of the road; and identifying whether a lane-level congestion occurs in at least one of the plurality of lanes.

Description

Technical Field

[0001] The present disclosure relates to autonomous vehicle navigation systems. Background Art

[0002] Many customer-facing applications require lane-level congestion assessment, including navigation, road anomaly detection, and accident reporting. Current systems assess road network congestion, but lack lane traffic distribution information, making it difficult to identify lane congestion with high fidelity.

[0003] Therefore, while current systems and methods for autonomous vehicle navigation achieve their intended purposes, a need exists for new and improved systems and methods for identifying lane-level road congestion. Summary of the invention

[0004] According to some aspects, a method is provided for collecting vehicle data from a fleet; identifying a lane-level distribution of a plurality of vehicles operating on a road comprising a plurality of lanes, the plurality of vehicles including a host vehicle; comparing the lane-level distribution to non-congested lanes of the road; and identifying whether lane-level congestion occurs in at least one of the plurality of lanes.

[0005] In another aspect of the present disclosure, the method includes: identifying at least one vehicle ahead of the host vehicle, moving slower than the host vehicle, defined as a lead actor vehicle, the lead actor vehicle defining a vehicle traffic deceleration; increasing a deceleration profile of the host vehicle based on a planned deceleration of the lead actor vehicle; and

[0006] In another aspect of the present disclosure, the method includes operating a sensor of a host vehicle to detect a lead vehicle; and upon initial detection of the lead vehicle, initiating deceleration of the host vehicle.

[0007] In another aspect of the present disclosure, the method includes: modifying a following distance between the host vehicle and the lead vehicle; and continuing to follow the lead vehicle at the following distance until the host vehicle reaches a deceleration of vehicular traffic ahead of the host vehicle.

[0008] In another aspect of the present disclosure, the method includes: performing lane-level load balancing to alleviate a situation where a non-congested lane identified by a host vehicle becomes congested; and creating a plurality of lane routing options for a plurality of vehicles on a road, the plurality of vehicles including the host vehicle.

[0009] In another aspect of the present disclosure, the method includes: applying an equal lane routing strategy to create multiple lane routing options, wherein the best selection of the multiple lane routing options is identified by an algorithm; applying a utility function metric, wherein equal weights are assigned to each of the multiple lane routing options; randomizing the multiple lane routing options and assigning probabilities to each of the multiple lane routing options; and assigning a vehicle among a plurality of vehicles to one of the multiple lane routing options.

[0010] In another aspect of the present disclosure, the method includes: applying a proportional lane routing strategy to create multiple lane routing options, wherein the multiple lane routing options are identified by an algorithm (which may include one of the Dijkstra algorithm or the Bellman-Ford algorithm); applying a utility function metric, including identifying a design capacity and a current traffic volume for each of the multiple lane routing options; assigning different weights to different options of the multiple lane routing options based on the available traffic volume; and randomly assigning a probability to a vehicle among multiple vehicles and an option among the multiple lane routing options.

[0011] In another aspect of the present disclosure, the method includes: identifying a decelerating vehicle defined as a lead vehicle in front of the host vehicle; and decelerating the host vehicle while maintaining a following distance from the lead vehicle.

[0012] In another aspect of the present disclosure, the method includes performing a trade-off analysis between a minimum travel time and customer comfort, wherein the minimum travel time has a plurality of lane changes to be completed to minimize the impact of one or more road congestion locations in front of the host vehicle, and maximizing the customer comfort by minimizing the number of lane changes and using a utility function for the trade-off, the utility function combining the travel time or estimated arrival time with the customer comfort.

[0013] In another aspect of the disclosure, the method includes applying a process to identify a system steady-state estimate.

[0014] According to some aspects, a method for enhancing vehicle navigation in conjunction with road congestion includes: analyzing a current distribution of multiple vehicles in front of or near a host vehicle traveling on a road having multiple lanes; evaluating vehicle clustering of at least one lane of the multiple lanes in front of the host vehicle; comparing the analysis results and the vehicle clustering results with a non-congested vehicle lane profile to distinguish a lane-level distribution of multiple vehicles in front of or near the host vehicle; using the lane-level distribution to generate a prediction of at least one congested lane in front of the host vehicle; and performing a trade-off analysis that provides an estimated travel time (TT) for the host vehicle to a predetermined destination, the trade-off analysis including: predicting a total amount of lane change operations required for the host vehicle until the predetermined destination is reached to minimize interactions with at least one congested lane.

[0015] In another aspect of the present disclosure, the method includes predicting at least one congested lane using a process including a Markov Chain process.

[0016] In another aspect of the disclosure, the method includes determining a system steady-state estimate.

[0017] In another aspect of the present disclosure, the method includes combining a first lane selection function and a second lane selection function, including: selecting a driving lane for the host vehicle in the first lane selection function based on habits saved or input by a host vehicle user and a comfort level of the host vehicle user; and developing lane selection in the second lane selection function based on planning data, the planning data being developed to prepare the host vehicle to exit at an upcoming road exit.

[0018] In another aspect of the present disclosure, the method includes: identifying a target lane for the host vehicle to transition into before reaching a road exit; and selecting an operation between adjusting the speed of the host vehicle and moving the host vehicle into the target lane as quickly and timely as possible to exit the exit.

[0019] In another aspect of the present disclosure, the method includes providing a user of a host vehicle with a visual indication of road segment congestion using a plurality of road segments respectively having different colors and line widths, the plurality of road segments collectively indicating a non-congested condition, a severely congested condition, and a congested condition, the host vehicle automatically adjusting a driving path based on the visual indication of the road segment congestion.

[0020] In another aspect of the present disclosure, the method includes: instructing a first component or component A to issue a lane-level congestion notification to a user or operator of a host vehicle; providing a second component or component B to perform lane-level congestion-aware navigation; combining with a third component or component C to perform lane-level load balancing; and including a fourth component or component D for maintaining a lane-level dynamic following distance.

[0021] According to some aspects, a method for enhancing vehicle navigation in conjunction with lane-level road congestion includes: providing personalized lane guidance to road users based on detected traffic distribution of lanes on the road to maximize traffic flow through congested areas; presenting lane profile data to road users for reference or use by an autonomous driving vehicle as part of vehicle behavior planning; presenting a road map depicting a predetermined portion of a road for vehicle travel and having multiple lanes; if a congested area is identified in the predetermined portion of the road, depicting one or more congestion symbols on the road map; and presenting a recommended vehicle driving path into one of a plurality of lanes used to represent recommended driving lanes to minimize driving interference in the congested area.

[0022] In another aspect of the present disclosure, the method includes: if a congested area is identified in a predetermined portion of the road, projecting at least one travel delay symbol and at least one estimated arrival time (TT) symbol on a road map; and identifying multiple lane changes to maximize the vehicle's passage through the congested area.

[0023] In another aspect of the present disclosure, the method includes: identifying a lane selected from a first lane, a second lane, and a third lane as a recommended driving lane on a predetermined portion of the road; and visually presenting a direction indicator, such as presenting an arrow pointing to the recommended driving lane on each of the first lane, the second lane, and the third lane that is not identified as the recommended driving lane.

[0024] Further areas of applicability will become apparent from the description provided herein.It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are for illustration purposes only and are not intended to limit the scope of the present disclosure in any way.

[0026] Figure 1 is a system diagram of a system for incorporating lane-level road congestion to enhance vehicle navigation according to an exemplary aspect;

[0027] Figure 2 yes Figure 1 The road section diagram of the system;

[0028] Figure 3 Is Figure 1 System-generated map representation of;

[0029] Figure 4 Is Figure 1 System-generated road lane representation;

[0030] Figure 5 is based on Figure 2 Modified road segment map to identify comfortable routes with minimal lane changes;

[0031] Figure 6 is based on Figure 2 a modified road segment map to identify vehicle routes with improved arrival times;

[0032] Figure 7 is used for Figure 1 a combined map of the system’s roads and lanes;

[0033] Figure 8 It is a system flow chart that provides dynamic following distance;

[0034] Fig. 9 is a flow chart of a system for providing lane level speed limits; and

[0035] Fig.10 yes Figure 1 System block diagram of the system. DETAILED DESCRIPTION

[0036] The following description is merely exemplary in nature and is not intended to limit the present disclosure, application, or uses.

[0037] refer to Figure 1 , a system 10 for enhancing vehicle navigation in conjunction with lane-level road congestion can be divided into two system portions. In a first system portion 12, for example located in a cloud computing network, a large amount of vehicle data is collected and analyzed to evaluate whether vehicle road lane congestion occurs in front of a first vehicle or host vehicle 14, which, according to some aspects, defines an autonomous vehicle (AV), and at least one second vehicle 16, which can define a leading vehicle in front of the host vehicle 14. In an evaluation portion 18, three performance evaluations are obtained. The data required for performing the evaluation portion 18 can be provided by a navigation provider 20, which can include road conditions, route conditions, and global positioning system (GPS) data. The calculations performed in the evaluation portion 18 can be performed using a computer 22, which can be located in a cloud-based system or in the host vehicle 14.

[0038] In the first of the three performance evaluations, the current lane-level distribution of multiple vehicles in front of or near the host vehicle 14 is analyzed, and a vehicle clustering evaluation is performed on at least one lane in front of the host vehicle 14. The results of the first performance evaluation are compared with the non-congested vehicle lane profile to distinguish the lane-level distribution of vehicles in front of or near the host vehicle 14.

[0039] In the second of the three performance evaluations, lane-level congestion prediction is performed, which can be performed using a process such as a Markov chain process. Note that other processes may also be used within the scope of the present disclosure. Lane-level congestion prediction is performed to determine a system steady-state estimate.

[0040] In a third of the three performance evaluations of the evaluation portion 18, a trade-off analysis is performed to provide an estimated travel time (TT) or estimated time of arrival (ETA) for the host vehicle 14 to a predetermined destination, which includes a prediction of the total amount of lane change maneuvers required by the host vehicle 14 until the predetermined destination is reached. The total amount of lane change maneuvers is applied to ensure that a predetermined level of vehicle user comfort is maintained during travel of the host vehicle, as will be described in greater detail herein.

[0041] A separate device such as a processor, a software component, an onboard computer, or a remote computer (hereinafter collectively referred to as computer 22) can be used to perform control of the system to enhance vehicle navigation 10 in conjunction with lane-level road congestion. Computer 22 can be deployed in the main vehicle 14 with perception capabilities or can be located remotely. Computer 22 is a non-universal electronic control device that has a pre-programmed digital controller or processor for storing data such as control logic, software applications, instructions, computer code, data, lookup tables, etc., memory or non-transitory computer-readable media, and a transceiver or input / output port. Computer-readable media include any type of media that can be accessed by a computer, such as read-only memory (ROM), random access memory (RAM), hard drive, compact disk (CD), digital video disk (DVD), or any other type of memory. Non-transitory computer-readable media do not include wired, wireless, optical or other communication links that transmit temporary electrical signals or other signals. Non-transitory computer-readable media include media that can permanently store data and media that can store data and then rewrite, such as rewritable optical disks or erasable storage devices. Computer code includes any type of program code, including source code, object code, and executable code.

[0042] The results of the three performance evaluations described above are forwarded to a map 24. The map 24 may include crowd-sensing dynamic data 26 and crowd-sensing static maps 28. The data collected or generated by the first system portion 12 may be stored in and retrieved from a cloud computing network 30.

[0043] After the data generation and collection by the first system portion 12, additional analysis is performed by the second system portion 32 located in the host vehicle 14. The first system portion 12 is in data communication with the second system portion 32. For example, data from the group-sensing static map 28 is transmitted to the lane static onboard map 34. For example, the output of the perception module 36 that collects data using sensors of the host vehicle 14, the group-sensing dynamic data 26, and the request data of the static onboard map 34 are forwarded to the evaluation and selection module 38. The evaluation and selection module 38 also receives the request data from the map 24.

[0044] The evaluation and selection module 38 performs a variety of functions. The first function of the evaluation and selection module 38 is to define a route evaluation 40, wherein a route selection based on detected traffic is updated. The second function of the evaluation and selection module 38 is to define a first lane selection function and a second lane selection function, including a first lane selection function 42, wherein a lane of travel for the host vehicle 14 is selected based on a saved or inputted habit of a user or passenger of the host vehicle 14, and the comfort of the operator of the host vehicle 14 is selected. In the second lane selection function 44, a lane selection is developed in the second lane selection function based on planning data, and the planning data is developed to prepare the host vehicle 14 for exiting at an upcoming road exit. The third function of the evaluation and selection module 38 is to define behavioral adaptations, which include headway adaptation 46, lateral lane change adaptation 48, speed profile adaptation 50, and hazard mitigation 52. The data collected or analyzed by the second system portion 32 is forwarded to the vehicle control module 54 as vehicle control instructions.

[0045] Also provided in the second system portion 32 is a vehicle navigation system 56 that communicates with the map 24. The vehicle navigation system 56 provides a notification unit 58 that provides system notifications to the operator of the host vehicle 14, a route selection unit that allows the operator to identify a preferred route, and a vehicle navigation system 56 that communicates with the map 24. Figure 1 A load balancing unit 62 is shown, and a motion suggestion unit 64 for identifying suggested vehicle speeds and inter-vehicle following distances, etc.

[0046] The road and route intelligence unit of the vehicle navigation system 56 provides up-to-date road and route data to the user or operator of the host vehicle 14. The data includes data regarding identified enabled roads 66 that define roads available for vehicle travel, identified driver interventions 68 to inform the operator of the host vehicle 14 when and if it may be necessary, for example, for the driver to take over operation of the automated vehicle, automated vehicles that define the host vehicle 14, traffic dynamics 70 that provide the latest changes occurring on the road, including, for example, accident reports, and road conditions 72 that provide the latest changes in road conditions that may occur, for example, due to weather conditions, etc.

[0047] refer to Figure 2 And still refer to Figure 1 , an exemplary road segment diagram 74 identifies how an exemplary road portion in the travel path of the host vehicle 14 is depicted. By applying the large amount of lane data received and analyzed and the lane level participation, a lane level congestion status can be provided for use by the operator of the host vehicle 14 and the vehicle navigation system 56. The exemplary road may include a first lane 76, a second lane 78, a third lane 80, and a fourth lane 82. Each lane may be divided into multiple segments, which may include a first or A segment 84, a second or B segment 86, a third or C segment 88, a fourth or D segment 90, and a fifth or E segment 92.

[0048] The segments of the lane set may be further specified, for example, an identifier 94 defined as X(A, 1) may be assigned to the first or A segment 84 of the first lane 76, an identifier 96 defined as X(A, 2) may be assigned to the first or A segment 84 of the second lane 78, an identifier 98 defined as X(A, 3) may be assigned to the first or A segment 84 of the third lane 80, and an identifier 100 defined as X(A, 4) may be assigned to the first or A segment 84 of the fourth lane 82.

[0049] A visual indication of road segment congestion may be provided to the operator of the host vehicle 14. For example, the first or A segment 84 of the first lane 76 may be presented as a lane segment 102 having a first color or line width indicating a non-congested condition, the second or B segment 86 of the first lane 76 may be presented as a lane segment 104 having a second color or line width indicating a severe congested condition, the third or C segment 88 of the first lane 76 may be presented as a lane segment 106 having a second color or line width indicating a severe congested condition, the fourth or D segment 90 of the first lane 76 may be presented as a lane segment 108 having a third color or line width indicating a congested condition, and the fifth or E segment 92 of the first lane 76 may be presented as a lane segment 110 having a congested condition. The third color or line width indicates a congested condition. The host vehicle 14 may automatically adjust the travel path, or the operator of the host vehicle 14 may manually adjust the vehicle travel path in view of the visual indication provided by the road segment map 74.

[0050] According to some aspects, the system 10 for enhancing vehicle navigation incorporating lane-level road congestion can be divided into multiple components. For example, the system can be divided into component A for notifying a user or operator of lane-level congestion notification, component B involving lane-level congestion-aware navigation, component C involving lane-level load balancing, and component D involving lane-level dynamic following distance. The individual components are discussed below.

[0051] refer to Figure 3, based on the detected traffic distribution of lanes on the road, personalized lane guidance is provided to road users to maximize traffic flow through congested areas. Lane profile data can be presented to users or operators other than the host vehicle 14 for reference by the AV or use as part of behavior planning. Component A of the system combines lane-level road congestion to enhance vehicle navigation 10, for defining lane-level congestion notifications, which can be presented as a road map 112 depicting a predetermined portion of a road 116 for travel by the host vehicle 14. If road congestion is identified in the predetermined portion of the road 116, the road map 112 will depict one or more congestion symbols 118. When congestion clears before the congested travel portion, a travel delay symbol 120 and one or more TT symbols 122 can be presented. A recommendation window 124 can be opened on the road map 112 to further remind the user or operator 14 of the host vehicle that a lane change is recommended to maximize traffic flow through the congested area. For example, the recommendation window 124 can visually display a slow traffic warning bar 126. Additionally, separate recommended directional arrows 128 may be presented on the presentation of the road 116, such as on each of the first lane 76, the second lane 78, and the fourth lane 82 to visually present a recommended vehicle travel path 80 into the third lane representing a recommended travel lane to minimize travel disruptions in congested areas.

[0052] refer to Figure 4 as well as Figure 3 Component A of the system 10 for incorporating lane-level road congestion to enhance vehicle navigation defines a lane-level congestion notification, which may also include a lane change suggestion, for identifying an upcoming exit for the host vehicle 14 . Figure 4 The predetermined portion of the road 116 is presented with the third lane 80 and the fourth lane 82 in which the host vehicle 14 is currently traveling superimposed on the predetermined portion of the road 116. The lane change direction arrow 132 is superimposed to visually indicate to the user or operator a recommended lane change from the third lane 80 to the fourth lane 82 to allow the vehicle to exit the road 116.

[0053] See also Figure 5 and Figure 2 , the component B of the system for vehicle navigation 10 that combines lane-level road congestion to enhance the definition of lane-level congestion-aware navigation can be presented on an exemplary road segment map 134, which has a similar Figure 274 . Once the lane-level congestion state is known, the AV routing engine can develop a lane selection strategy that is presented as a first suggested route 136 that strictly minimizes TT. The user or operator can choose to follow the first suggested route 136 because the route selection has minimal or no identified congested road lane segments. Therefore, following the first suggested route 136 can minimize the time of vehicle stop-and-go and provide substantially continuous movement in the available driving lanes.

[0054] refer to Figure 6 and Figure 5 , further application of component B generates a second suggested route 138, which is also presented on the exemplary road segment map 134. Figure 5 The first suggested route 136 discussed here focuses on the possibility that TT may result in an uncomfortable ride experience such as excessive lane changes. To mitigate this possibility, according to some aspects, lane-level TT or estimated arrival time and passenger comfort are weighed as a trade-off to be balanced by minimizing the number of vehicle lane changes and maximizing passenger comfort through the trade-off, and combining travel time or estimated arrival time with customer comfort through a utility function. In this aspect of component B, following the second suggested route 138 requires fewer lane changes than following the first suggested route 136 while also achieving a minimum travel time. Therefore, following the second suggested route 138 extends the lane-level TT while improving passenger comfort.

[0055] Equation 1 or Equation 2 below provide exemplary formulas that may be used to implement the calculations required for component B. For example, on lane #i in road segment j, the utility function U may be applied i,j To represent the travel time value, instead of using the travel time t i,j , and use it as a metric for the navigation route. Here, U i,j is the utility function that characterizes the metric in segment j and lane I, taking into account the ETA time t through the mathematical formula in Formula 1 or Formula 2 i,j and comfort index L i,j .

[0056] Formula 1:

[0057] U i,j =f(t i,j ,L i,j )=k1t i,j +k2L i,j

[0058] Formula 2:

[0059] U i,j =f(t i,j ,L i,j )=k1ln(ti,j )+k2ln(L i,j )

[0060] For the above formulas 1 and 2, use (1) lane / segment + (2) utility function U ij As a metric, use the segmented ETA (t ij ) or TT as a metric for route decision making, which is inserted into a shortest path algorithm, such as Dijkstra's algorithm or Bellmen-Ford's algorithm to calculate the route / lane selection strategy. Here, the travel time estimates of the route segments are aggregated into the estimated time or arrival time of the journey. It is noted that the algorithms identified herein, including Dijkstra's algorithm and Bellmen-Ford's algorithm, are provided only as examples. A variety of other algorithms may also be used within the scope of the present disclosure.

[0061] refer to Figure 7 And see again Figure 3 , component B of the system for incorporating lane-level road congestion to enhance vehicle navigation 10 defines lane-level congestion-aware navigation, which is presented on a combined road and lane map 140 having a map portion 142 and a lane presentation portion 144. Initially, a host vehicle 14 traveling on a predetermined first road 146 should exit from an exit 148 to change the driving path to a second road 150 to achieve an initial target driving path. If the host vehicle 14 attempts to change lanes but cannot reach a currently severely congested target or congested lane 152 in time to use the exit 148, such as due to congestion in the current driving lane 154 ahead of the host vehicle 14, the host vehicle 14 in this example misses the exit 148 due to the intermediate driving lane 156 and the congested target lane 152, which adds 5.1 miles of driving distance to reach the original driving destination 158.

[0062] Component B defining lane-level congestion-aware navigation allows the host vehicle 14 to be aware of backup vehicles or congestion ahead of the host vehicle 14 while still traveling on the first road 146. Using this data, the host vehicle 14 can adjust vehicle speed and / or change to the target lane 152 in time to exit the exit 148.

[0063] Component C of the system for incorporating lane-level road congestion to enhance vehicle navigation 10 defines lane-level load balancing to mitigate situations where a non-congested lane identified by the host vehicle 14 and suggested for lane change suddenly becomes congested. The load balancing guidance provided by component C is also used to ensure that lane change suggestions recommended by the system do not adversely affect lane traffic flow. For example, after all vehicle traffic has been directed to an uncongested lane route, the previously uncongested lane route may suddenly become congested because other vehicle operators see an opportunity to move into the previously uncongested lane route. Two lane route assignment strategies are provided.

[0064] The load balancing method of the present disclosure requires the top K selections of lane routing options based on the shortest path algorithm outlined above with reference to component B. Disconnected lane routing is preferred over connected lane routing. Vehicle traffic assignments are randomly assigned among the K legal options for lane routing. Vehicle routing assignments are assigned based on the available remaining capacity of each lane routing.

[0065] Two lane routing strategies using the present load balancing method may be applied. In the first strategy or equality strategy, the top K choices of lane routing options are identified by an algorithm such as Dijkstra's algorithm or Bellman-Ford algorithm using the utility function metric outlined above with respect to component B. Each lane routing option is assigned an equal weight. The assignment of a single vehicle to a particular lane routing option is random and assigned a probability, where the probability P i =1 / K.

[0066] In the second strategy or proportional lane routing strategy, the top K choices of lane routing options are identified using a utility function metric using an algorithm such as Dijkstra's algorithm or Bellman-Ford algorithm. For each lane routing option i (i∈k), the design capacity C is determined i and its current traffic volume U i Different weights are assigned to different lane routing options depending on the amount of available or remaining traffic. The probability of assigning each vehicle to a specific lane routing option i is random using Formula 3 below:

[0067] Formula 3:

[0068]

[0069] When a deceleration is identified ahead of the host vehicle 14, a component D1 of the system is provided that incorporates lane-level road congestion to enhance vehicle navigation 10, which defines a lane-level dynamic following distance. When a deceleration is identified, the host vehicle 14 dynamically modifies the following distance between the host vehicle 14 and the lead vehicle (e.g., a second vehicle 16 located ahead of the host vehicle 14) to prepare for the upcoming deceleration of the second vehicle 16. The following distance is increased to allow the host vehicle 14 to plan to initiate a longer deceleration profile when the host vehicle sensors detect that the second vehicle 16 begins to decelerate.

[0070] refer to Figure 8, a system flow chart 160 identifies the steps taken by the operation of the above-described component D1. When a forward deceleration is identified, the host vehicle 14 will dynamically modify the following distance from the lead vehicle or the second vehicle 16 to prepare for the upcoming deceleration. The increase in following distance allows the host vehicle 14 to plan a longer deceleration curve, which will be initiated when the vehicle sensors of the host vehicle 14 detect that the lead vehicle or the second vehicle 16 begins to decelerate. First, a traffic query 162 is performed to identify whether the cloud-based data indicates that there is slower traffic ahead of the host vehicle. If the answer to the traffic query 162 is "no" 164, the program ends at a first end step 166. If the answer to the traffic query 162 is "yes" 168, a vehicle query 170 is performed, which identifies whether a lead vehicle is present. If the response to the vehicle query 170 is "yes" 172, an increase command 174 is issued to increase the following distance between the host vehicle 14 and the second vehicle 16 (e.g., the lead vehicle). At a time after the increase command 174 is issued, an update query 176 is made to determine if the lead vehicle is decelerating. If the response to the update query 176 is yes 178, a deceleration command 180 is issued, which commands the host vehicle 14 to decelerate with the lead vehicle while continuing to maintain the following distance. After the deceleration command 180, the program ends at a second end step 182.

[0071] If the response to the vehicle query 170 is "no" 164, a curve deceleration command 186 is issued, which increases the deceleration profile of the host vehicle 14 based on the expected or planned deceleration of the traffic speed reported in front of the host vehicle 14. After the curve deceleration command 186, a sensor query 188 is made to identify whether the sensors of the host vehicle 14 detect other actors or vehicles in front of the host vehicle 14. If the response to the sensor query 188 is "no" 190, a continue command 192 is issued, which commands the host vehicle 14 to continue to decelerate at a higher deceleration profile until the host vehicle 14 exceeds the detected deceleration. If the response to the sensor query 188 is "yes" 196, a start deceleration command 198 is issued, which initiates deceleration of the host vehicle 14 when a slower moving actor or vehicle in front of the host vehicle 14 is first actually detected. After the initiate deceleration command 198 is issued, a continue command 200 is issued which commands the host vehicle 14 to continue following the slower moving actor at the calculated new distance until the host vehicle 14 exceeds the detected deceleration and the process ends at a third step 202 .

[0072] If the response to the update query 176 is “no” 204 ​​, a continue following command 206 is issued which commands the host vehicle 14 to continue following the decelerating actor vehicle at the newly calculated distance until the host vehicle 14 passes the decelerating vehicle and the process ends at a fourth end step 208 .

[0073] refer to Fig. 9 , system flow chart 210 identifies the steps taken for the operation of the above-mentioned component D2. Component D2 of the system 10 for enhancing vehicle navigation in conjunction with lane-level road congestion involves lane-level dynamic following distance. For lane-level speed limits using lane friction, when lane-level traffic data indicates that a vehicle in an adjacent lane is moving slower than the main vehicle 14 by a predetermined threshold within a predetermined distance, the presence of slower adjacent traffic is confirmed by the vehicle sensors of the main vehicle 14, and the maximum speed of the autonomous driving system, such as a navigation module, is limited to a configurable speed difference. An exemplary method for configuring the maximum speed difference is to reference a severity level based on a collision speed given on an international scale. Organization for Standardization (ISO) 26262 provides functional safety of road vehicles and aligns calibration with the safety integrity level of the autonomous driving system.

[0074] First, a program start 212 is performed and a differential speed query 214 is performed to check whether the differential speed with the adjacent traffic threshold is met. After the differential speed query 214, a neighboring actor query 216 is performed to identify whether the actor vehicle is adjacent to the host vehicle 14. If the response to the neighboring actor query 216 is "no" 218, the program ends at a first program end step 220. If the response to the neighboring actor query 216 is "yes" 222, a speed determination 224 is performed to identify whether the neighboring actor vehicle is moving at the host vehicle speed or the host vehicle set speed. If the response to the speed determination 224 is "no" 226, the program ends at a second end step 228. If the response to the speed determination 224 is "yes" 230, a host vehicle lane query 232 is performed to identify whether the host vehicle 14 is in an open lane. If the response to the host vehicle lane query 232 is "no" 234, a modify speed profile command 236 is generated and the program ends at a third end step 238. If the response to the host vehicle lane query 232 is "yes" 240, a deceleration command 242 is generated to temporarily reduce the speed of the host vehicle 14 to a predetermined difference in the speed of the neighboring actor. Following the speed reduction command 242, a reaction command 244 is generated that allows the host vehicle 14 to react to any reduction or decrease in the number of neighboring active vehicles, and the process ends at a fourth end step 246.

[0075] refer to Fig.10, a system block diagram 248 provides some parts of a system for incorporating lane-level road congestion to enhance vehicle navigation 10 in a cloud computing network 250 and in a vehicle 252 such as the host vehicle 14. Within the cloud computing network 250, a data ingestion module 254 collects system sensor data including third-party traffic data 256. A first output of the data ingestion module 254 is directed to a lane congestion model 258. The lane congestion model 258 may also receive data from a base map 260. A second output of the data ingestion module 254 is directed to a data aggregation module 262, and the output of the data aggregation module 262 is directed to a lane congestion map 264, which identifies various road lanes with different levels of congestion, including non-congested lanes, busy or congested traffic lanes, and one or more congested lanes where traffic may be stopped or slowed to a minimum speed to force vehicles into fewer lanes, such as in a construction zone. The output of the lane congestion map 264 is directed to a lane routing engine 266, and the output of the lane congestion model 258 is also directed to the lane routing engine 266. The lane routing engine 266 identifies possible routes for the host vehicle 14 and route changes to apply to improve vehicle TT.

[0076] Further references Fig.10 , data is exchanged between the cloud computing network 250 and the vehicle 252, such as via wireless data transmission. The components of the vehicle 252 and the information and data exchange with the cloud computing network 250 are identified thereafter. For example, the output of the lane routing engine 266, which is used to define lane navigation route data 268, is forwarded to the navigation module 270. The navigation request 272 is forwarded from the navigation module 270 to the lane routing engine 266 and can be returned as the lane navigation route data 268. The navigation module 270 also receives user input 274 provided by a vehicle user 276 or operator.

[0077] The vehicle 252 also includes a telematics module 278 that can communicate with the navigation module 270 via a global navigation satellite system (GNSS) and using a vehicle-to-network (V2N) communication protocol. The telematics module 278 is responsible, in part, for performing data collection tasks. For example, the telematics module 278 can receive GPS data, collision event data, emergency braking data, and stability data, all of which together define vehicle data 280 from a crowdsourced fleet 282. The output of the navigation module 270 can also define guided navigation route information 284 and free navigation alerts 286.

[0078] According to some aspects, the method of the present disclosure provides lane-level congestion identification using large amounts of vehicle data and advanced algorithms. The method of the present disclosure applies lane-level congestion plus comfort measurements to assist vehicle decision-making, including making AV navigation decisions. The method of the present disclosure performs a trade-off analysis between minimum travel time, which may include multiple vehicle lane changes to minimize the impact of one or more road congestion locations in front of the host vehicle, and customer comfort, which also provides an enhancement to the human-machine interface (HMI) by reducing the method of the present disclosure. According to some aspects, lane routing options are provided to different vehicles based on load balancing between different available options. According to some aspects, headway dynamics and speed recommendation information are provided based on lane-level traffic dynamics and traffic dynamics of adjacent lanes.

[0079] The disclosed system 10 for enhancing vehicle navigation in conjunction with lane-level road congestion also provides a method for performing lane-level distribution, clustering, and analysis for non-congested lane profiles. The method for performing lane-level congestion prediction may apply a Markov chain process to identify vehicle arrivals and provide system steady-state estimates. The method also includes performing a trade-off analysis between an estimated arrival time of the host vehicle and determining a lane change count operation during travel to minimize lane changes, thereby ensuring operator or user comfort.

[0080] The disclosed system 10 for enhancing vehicle navigation by incorporating lane-level road congestion provides a number of advantages. These include providing lane-level congestion and comfort measures to assist in navigation decisions. This includes selecting a driving lane that best matches customer preferences and avoids traffic congestion. This also includes input to a behavior planning module to ensure that the autonomous vehicle successfully navigates to a vehicle exit ramp.

Claims

1. A method for enhancing vehicle navigation by incorporating lane-level road congestion, the method comprising: Collect vehicle data from fleets; identifying a lane-level distribution of a plurality of vehicles operating on a road including a plurality of lanes, the plurality of vehicles including a host vehicle; comparing the lane-level distribution to a non-congested lane of the road; as well as It is identified whether lane-level congestion occurs in at least one lane of the plurality of lanes.

2. The method according to claim 1, further comprising: identifying at least one vehicle ahead of the host vehicle and moving slower than the host vehicle, defined as a lead vehicle, the lead vehicle defining a vehicular traffic deceleration; as well as A deceleration profile of the host vehicle is augmented based on the planned deceleration of the lead vehicle.

3. The method according to claim 2, further comprising: operating a sensor of the host vehicle to detect the lead vehicle; as well as Upon initial detection of the lead vehicle, deceleration of the host vehicle is initiated.

4. The method according to claim 3, further comprising: modifying a following distance between the host vehicle and the lead vehicle; as well as The lead vehicle continues to be followed at the following distance until the host vehicle reaches a deceleration of the vehicular traffic ahead of the host vehicle.

5. The method according to claim 1, further comprising: performing lane-level load balancing to alleviate a situation in which the non-congested lane identified by the host vehicle becomes congested; as well as A plurality of lane routing options are created for a plurality of vehicles on the road, the plurality of vehicles including the host vehicle.

6. The method according to claim 5, further comprising: applying an equal lane routing strategy to create the plurality of lane routing options, wherein a best selection of the plurality of lane routing options is algorithmically identified; applying a utility function metric wherein equal weight is assigned to each of the plurality of lane route options; randomizing the plurality of lane route options and assigning a probability to each of the plurality of lane route options; as well as A vehicle among the plurality of vehicles is assigned to one of the plurality of lane routing options.

7. The method according to claim 5, further comprising: applying a proportional lane routing strategy to create the plurality of lane routing options, wherein the plurality of lane routing options are identified by an algorithm; applying a utility function metric including identifying a design capacity and current traffic usage for each of the plurality of lane route options; assigning different weights to different ones of the plurality of lane route options based on available traffic volume; as well as A probability is randomly assigned to a vehicle among the plurality of vehicles and an option among the plurality of lane route options.

8. The method according to claim 1, further comprising: identifying a decelerating vehicle defined as a lead vehicle ahead of the host vehicle; as well as The host vehicle is decelerated while maintaining a following distance from the lead vehicle.

9. The method of claim 1, further comprising performing a trade-off analysis between minimum travel time and customer comfort, wherein: The shortest travel time has multiple lane changes to be completed to minimize the impact of one or more road congestion locations in front of the host vehicle, and the customer comfort is maximized by minimizing the number of lane changes and weighing them using a utility function that combines travel time or estimated arrival time with the customer comfort.

10. The method of claim 1, further comprising applying a chaining process to identify a system steady-state estimate.

Images