Vehicle speed or travel time estimation

Abstract

Systems and methods for estimating vehicle speed or travel time along a section of road by calculating a priori vehicle speed statistical distribution based on historical vehicle speed data along the road; receiving observed vehicle speed data, performing Bayesian fusion of the priori vehicle speed data and the observed vehicle speed data to determine a posteriori vehicle speed distribution; and determining a vehicle speed estimate or a travel time estimate based on the posterior vehicle speed profile.

Description

Technical Field

[0001] The present disclosure generally relates to methods and systems for traffic speed and / or travel time estimation. Background Art

[0002] The purpose of providing this background section is to generally present the context of the present disclosure. The contents of this background section are neither explicitly nor implicitly admitted to be prior art with respect to the present disclosure.

[0003] Traffic movement speeds, especially in urban environments, are subject to a large degree of natural variability due to weather conditions, events in the area, changes in traffic flow, road construction, accidents, and other unpredictable conditions or events. Conventional traffic speed estimation systems and methods rely on real-time traffic data or floating car data to estimate the speed of traffic or vehicles on a particular stretch of road (road segment). Such conventional systems rely on confidence thresholds, such as the number of samples of real-time or recent traffic speed data from moving vehicles, unique vehicle counts, etc.

[0004] However, conventional floating car data is susceptible to errors or inaccuracies arising from several reasons, including GPS errors, map matching inaccuracies, unavailability of floating car data in tunnels, etc. Since conventional systems and methods are subject to several sources of inaccuracy, these inaccuracies may accumulate and result in inaccurate estimates of traffic speed or travel time.

[0005] It would be desirable to address or ameliorate one or more disadvantages or limitations associated with conventional systems and methods for traffic speed or travel time estimation, or at least provide a useful alternative. Summary of the invention

[0006] The present disclosure provides a system for estimating vehicle speed along a section of road, the system comprising:

[0007] one or more processors (processors);

[0008] A memory accessible to a processor, the memory comprising program code executable by the processor to:

[0009] Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road;

[0010] receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road;

[0011] performing a Bayesian fusion of the prior distribution and the observed vehicle speed data to determine a posterior distribution by expressing the prior distribution as a conjugate of a likelihood function;

[0012] A vehicle speed estimate is determined based on the posterior distribution.

[0013] The present disclosure also provides a system for estimating travel time, the system comprising:

[0014] one or more processors;

[0015] A memory accessible to a processor, the memory comprising program code executable by the processor to:

[0016] receiving, from a computing device, a starting location, a destination location, and a proposed travel time;

[0017] Determine a driving route between a starting location and a destination location, the driving route including at least one road segment;

[0018] estimating the travel time by estimating the segment travel time for each segment and summing the segment travel times;

[0019] The following steps are used to estimate the road segment travel time:

[0020] Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road segment;

[0021] receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road;

[0022] performing a Bayesian fusion of the prior distribution and the observed vehicle speed data to determine a posterior distribution by expressing the prior distribution as a conjugate of a likelihood function;

[0023] determining an estimated vehicle speed estimate based on the posterior distribution;

[0024] A road segment travel time is determined based on the vehicle speed estimate and the road segment length.

[0025] The present disclosure also provides a method for estimating a vehicle speed along a section of road, the method comprising:

[0026] Based on historical vehicle speed data along the road, calculating a priori vehicle speed statistical distribution (prior distribution);

[0027] receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road;

[0028] performing a Bayesian fusion of the prior distribution and the observed vehicle speed data to determine a posterior distribution by expressing the prior distribution as a conjugate of a likelihood function;

[0029] A vehicle speed estimate is determined based on the posterior distribution.

[0030] The present disclosure also provides a method for estimating travel time, the method comprising:

[0031] receiving, from a computing device, a starting location, a destination location, and a proposed travel time;

[0032] Determining a driving route between a starting location and a destination location, the driving route including at least one road segment;

[0033] estimating the travel time by estimating the segment travel time for each segment and summing the segment travel times;

[0034] The travel time for each road segment is estimated as follows:

[0035] Based on the historical vehicle speed data along the road section, calculating the prior vehicle speed statistical distribution (prior distribution);

[0036] receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road segment;

[0037] performing a Bayesian fusion of the prior distribution and the observed vehicle speed data to determine a posterior distribution by expressing the prior distribution as a conjugate of a likelihood function;

[0038] determining a vehicle speed estimate based on a posterior vehicle speed distribution;

[0039] A road segment travel time is determined based on the vehicle speed estimate and the road segment length. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Exemplary embodiments of the present invention are illustrated by way of example in the accompanying drawings, in which like reference numerals refer to the same or similar elements, and in which:

[0041] Figure 1 A block diagram of a system for vehicle speed and / or travel time estimation and its associated components is shown;

[0042] Figure 2 A flow chart showing a method for vehicle speed estimation;

[0043] Figure 3 A flow chart showing a method for travel time estimation;

[0044] Figure 4A a graph showing rolling average speed of a vehicle as a function of cumulative count of vehicles;

[0045] Figure 4B Shown based on Figure 4A A graph of vehicle speed values ​​obtained from the data represented in;

[0046] Figure 4C Relative to Figure 4A a priori, observed, and a posteriori vehicle speed statistical distributions determined from data from graphs of; and

[0047] Figure 5 A flow chart showing a portion of a method of generating an a priori vehicle speed statistical distribution. DETAILED DESCRIPTION

[0048] Traffic speeds in urban environments are very dynamic, varying from area to area and influenced by prevailing conditions. A large number of variables influence how fast traffic moves in an area. These variables may be specific to a particular area or stretch of road within a city. The large number of variables and the degree of variation in factors that influence traffic speeds make the estimation of travel time a significant computational challenge. Accurate travel time estimates facilitate the efficient allocation of vehicle or fleet resources within urban areas. Accurate travel time estimates also improve the quality of information available to individuals who rely on vehicles or fleets deployed in urban areas for transportation.

[0049] The present disclosure provides a method and system for efficiently estimating travel time or traffic speed based on a combination of historical data about traffic movement patterns and data obtained from vehicles in real time or near real time. These embodiments advantageously provide more accurate travel time or travel speed estimates for vehicles in urban areas. For example, these embodiments allow better estimates to be provided to users of vehicle sharing services. Accurate travel time estimates also advantageously allow vehicles in a fleet of vehicles to be better allocated to potential users of vehicles in an urban environment. Some embodiments can be implemented to estimate the travel time of public transportation services (such as buses or shared vehicles, etc.). Advantageously, these embodiments also provide estimates of travel time or travel speed in response to ground traffic conditions (such as accidents, congestion, road closures, etc.). Due to unpredictable changes in traffic conditions in urban environments, these embodiments advantageously provide a more responsive system for estimating travel time or travel speed.

[0050] Some embodiments also perform travel time estimation by receiving a starting location and a destination location as input and identifying a route / path between the starting point and the destination. The identified route is segmented into different road segments. A travel time estimate can be determined for each road segment, and a total travel time estimate can be determined as the sum of the individual travel time estimates.

[0051] The estimation methods described herein perform Bayesian fusion or Bayesian inference to provide a more accurate estimate of travel time or travel speed. The travel time or vehicle speed is estimated using a combination of historical vehicle speed data and real-time or near real-time vehicle speed data collected from a fleet of vehicles. Some embodiments utilize the mathematical convenience of conjugate priors and closed-form solutions derived therefrom to obtain a posteriori estimates of road traffic conditions.

[0052] Figure 1 A block diagram of a system 100 for estimating vehicle speed or travel time along a stretch of road is shown. Associated components include one or more vehicles 140. Associated with each vehicle is a computing device 150 configured to communicate with the system 100 via a network 130. In some embodiments, the computing device 150 may be a smart phone or handheld computing device associated with the vehicle 140 to transmit information from the vehicle 140, the system 100, or to receive information from the system 100. The computing device may transmit and receive data wirelessly via the network 130 as the vehicle 140 navigates an area.

[0053] A history database 120 is also provided for storing historical vehicle speed or travel time information. Records within the history database 120 may include one or more of the following attributes: vehicle identifier, vehicle trip identifier, vehicle trip / travel time and date data, vehicle trip start location data, vehicle trip end location data, vehicle speed values, etc. Records within the history database 120 may be aggregated or categorized based on time of day, day of the week, and location values ​​associated with the record - in this context, the location value refers to the location of the vehicle when the data about the vehicle (e.g., speed and time of day) was obtained. This categorization allows for the determination of a priori vehicle speed statistical distributions that are specific to a particular time or time period on a particular day of the week. This prior vehicle speed statistical distribution provides a more focused starting point for estimating travel time or vehicle speed, thereby enabling more accurate estimates.

[0054] The vehicle speed values ​​stored in the historical database 120 are associated with a location or a section of road, and are thus used as an indication of the speed of traffic at a location or on the road section. The historical database 120 can be populated with data collected continuously or intermittently from a computing device 150 set in a fleet of vehicles traveling in a certain area - as used herein, a "fleet" can refer to a plurality of related vehicles (such as taxis in a taxi company's fleet), and / or a plurality of unrelated vehicles. The prior vehicle speed distribution obtained based on the records in the historical database 120 is used as a prior probability distribution for the Bayesian fusion operation performed by the vehicle speed estimation system 100.

[0055] The vehicle speed estimation system 100 includes one or more processors 102 configured to access a memory 104. The memory 104 stores program code 106 for implementing the method of travel time estimation. The vehicle speed estimation system 100 also includes a network interface 108 to allow communication with a historical database 120, or (multiple) computing devices 150 or other related components through a network 120.

[0056] The present disclosure contemplates computer system 100 taking any suitable physical or virtual form. System 100 may include one or more computer systems 100; may be single or distributed; across multiple locations; across multiple machines; across multiple data centers; or located in a cloud, which may include one or more cloud components in one or more networks. Where appropriate, one or more systems 100 may perform one or more steps of one or more methods described or illustrated herein without substantial spatial or temporal limitations. Where appropriate, one or more systems 100 may perform one or more steps of one or more methods described or illustrated herein at different times or in different locations.

[0057] The computing device 150 includes one or more processors 152 configured to access a memory 154 storing program code 156 to enable interaction between the computing device 150 and the system 100. The computing device 150 also includes a network interface 159 to allow communication with the historical database 120 or the system 100 or other related components through the network 120. The computing device includes a GPS sensor 157 for estimating the current position of the computing device 150 and thereby inferring the position of the vehicle 140. The inferred position of the vehicle 140 and the speed value based on the rate of change of the position can be transmitted to the system 100 to provide real-time or near real-time vehicle speed information. The speed and / or position information obtained from multiple computing devices 150 provides a basis for calculating the observed vehicle speed statistical distribution, which represents the real-time or near real-time traffic conditions. The observed vehicle speed statistical distribution is calculated relative to a section of road or a road segment to provide a targeted representation of the observed traffic conditions.

[0058] Figure 2A flow chart of a method 200 for vehicle speed estimation is shown. The method 200 is performed by the system 100. Specifically, the method 200 is performed by the processor 102 by calling the processing logic of the program code 106 in the memory 104. Step 210 includes calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data. The historical vehicle speed data is associated with a section of road or a road segment of interest for calculation and is retrieved from the historical database 120. The retrieved data may be associated with such vehicle speed: the vehicle speed is associated with a day of the week and a time of the day associated with the estimated time of the vehicle speed. The time of the day and / or the day of the week may be selected to include the current time of the day of the vehicle whose speed is being estimated, or the time of the day when the vehicle is expected to arrive at each corresponding road segment. The priori vehicle speed statistical distribution uses previously captured floating car data to model the prior knowledge of vehicle speed. By calculating the distribution based on historical data associated with a day of the week and a time of the day, the system 100 advantageously obtains a more representative prior distribution related to the day and time of the vehicle speed estimate calculated.

[0059] Step 210 includes a sub-step 212 for calculating an exponential moving average. Sub-step 212 includes calculating an average vehicle speed along the road segment of interest and passing the average vehicle speed through an exponentially smoothed moving average function to capture historical driving patterns. The calculated average value can be related to vehicle speed data in historical data for a predefined period of time.

[0060] Step 210 includes a sub-step 214 for calculating the variance of the average speed of the vehicle over a predefined period of time on the road segment of interest. The calculated variance is an indicator of the degree of variability of the average speed. A prior distribution with a low variance is considered to be a strong prior, whereas a high variance indicates a high degree of uncertainty and a low degree of confidence in the prior distribution.

[0061] Step 210 includes a sub-step 216 of calculating a variance mean. Specifically, sub-step 216 includes calculating the variance of the speeds of the individual vehicles on the road segment of interest, and then determining the average value of the calculated variance values ​​within a predefined period.

[0062] The predefined period used for the calculations at sub-steps 212, 214 and 216 may be, for example, 8 weeks, 10 weeks or 12 weeks. The predefined period may be biased towards more recent data - for example, the period may end at the current date or the previous day, or the most recently acquired N data may be used to make the a priori estimate, or both, using some combination or formula. The calculations at 212, 214 and 216 are used to define a priori vehicle speed statistical distributions, which are used as conjugate pairs and fused with the observed vehicle speed statistical distributions calculated at step 230. Figure 4CA prior distribution 410 calculated based on experimental data is shown.

[0063] At step 220, observed vehicle speed data related to real-time or near real-time vehicle traffic along the road segment of interest is received. For example, the real-time or near real-time vehicle speed data may include vehicle data received within a recent time period, such as the last 5 minutes, the last 15 minutes, or the last 30 minutes. A suitable time period may be selected based on the number of recent records. Based on the observed data, an average vehicle speed within the recent time period may be calculated.

[0064] A sample count or number of observations of vehicle speeds from a road segment of interest obtained within a recent period can be calculated based on the observed vehicle speed data. In some embodiments, the weight assigned to the observed vehicle data during fusion is proportional to the sample count value. The sample count value is used as a concept of confidence in the observed vehicle speed data. When a large number of samples are available, the fused estimate tends to over-weight the observed data (i.e., real-time data). When the number of sample counts is low, the fused estimate tends to over-weight the prior distribution - this is because a larger number of samples in each case leads to greater confidence in the data. The embodiment advantageously balances consideration of the prior data and the observed data to provide a more accurate estimate of vehicle speed or travel time. Figure 4C A graph 420 of observed data calculated based on received real-time, near real-time, or recent vehicle speed data is shown.

[0065] In step 230, a Bayesian fusion is performed on the prior data and the observed data to obtain the posterior vehicle speed distribution (posterior distribution). Given the likelihood function and the prior distribution, Bayes' theorem states that the posterior probability is proportional to the prior distribution multiplied by the likelihood term. However, the relationship of "proportional" is relatively vague to implement in a real-time system. The system 100 is configured to assume that the likelihood function and the prior distribution are in conjugate form. The system 100 determines a closed-form solution to the posterior distribution based on the variables / statistical properties of the prior and likelihood estimates. Under these constraints, the "proportional" relationship is converted to equivalence and can therefore be implemented in code.

[0066] Some conjugate pairs include Normal-Normal and Normal-Gamma. System 100 uses Normal-Normal approximation for both prior and likelihood distributions. System 100 performs Bayesian fusion based on the principle that for sufficiently large "n", the distribution of sample means tends to a normal distribution (central limit theorem).

[0067] The system 100 may implement methods such as variational inference and Markov chain Monte Carlo (MCMC) to infer prior / posterior distributions when receiving vehicle speed data in real time or near real time. Alternatively, the system 100 may treat these distributions as normal distributions using a rolling average of vehicle speed as the mean of the normal distribution.

[0068] Assuming that both the prior distribution and the likelihood (observed) distribution are normal-normal, the posterior hyperparameters can be expressed as:

[0069]

[0070] where μ0 (mean) and (variance) is related to the prior distribution, σ 2 is the variance of the observed vehicle speed, n is the number of observations, and x i is the i-th observation among n observations. Figure 4C A posterior distribution 430 calculated based on the prior distribution 410 and observed vehicle speed data 420 is shown.

[0071] The posterior distribution determined in step 230 can be used as an experimental prior distribution for further experimentation and estimation of vehicle speed or travel time for future time periods in an online learning environment. A survival time can be assigned to the posterior distribution determined in step 230, after which the posterior distribution is considered invalid, and an updated posterior distribution is calculated using updated historical and / or real-time or near real-time data through execution of method 200. By periodically updating the posterior distribution, system 100 takes into account recent vehicle speed data with relevant historical vehicle speed data to provide more accurate estimates of vehicle speed and travel time.

[0072] At step 240, the system 100 determines a vehicle speed estimate based on the calculated posterior distribution. The vehicle speed estimate may be the mean of the posterior distribution, or a range of estimated vehicle speeds defined based on the mean and standard deviation of the posterior distribution.

[0073] Figure 3A flow chart of a method 300 for travel time estimation is shown. Method 300 includes method 200. In step 310, the system 100 receives a starting location and a destination location from a requesting computing device 150. The starting location and the destination location can be received as part of a transportation hailing service request, which is described in PCT publication WO 2020046200 A1 entitled "E-hailing service". The starting location and the destination location can be received as part of a service request, which is described in PCT publication WO 2018208232 A1 entitled "Allocation of dynamically batched service providers and service requesters."

[0074] In step 320, the system 100 determines a travel route between the starting location and the destination location. The travel route includes at least one road segment. Longer routes can be divided into multiple road segments. Each road segment serves as a basis or standard for calculating a priori vehicle speed distribution and observed vehicle speed distribution, and then the two distributions are fused to determine the posterior distribution of each road segment. By dividing the route into one or more road segments, the system 100 advantageously simplifies the complex travel time estimation problem into a different and less complex travel time estimation problem associated with a specific road segment.

[0075] In some embodiments, the system 100 may determine a driving route based on an apparatus and method for processing route information described in PCT Publication WO 2020263176 A1 entitled “Processing route information”, the contents of which are incorporated herein by reference. In some embodiments, the system 100 may determine a driving route based on an apparatus and method for generating route navigation data described in PCT Publication WO 2022031222 A1 entitled “Processing apparatus and method for generating route navigation data”, the contents of which are incorporated herein by reference. In some instances, the system may identify one or more routes between a starting point and a destination, and estimate the travel time for each route. The system 100 may then recommend a route with the lowest total travel time (i.e., the sum of the time to travel each section of the road separately).

[0076] At step 330, the system 100 estimates a vehicle speed estimate for each road segment identified in step 320 by executing the method 200 for each road segment. Based on the estimated vehicle speed, the system 100 estimates a travel time or an estimated range of travel times for each road segment. At step 340, the system 100 estimates an estimate of a total travel time or a total travel time range based on the sum of the travel times estimated at step 330. The total travel time estimate may be transmitted by the system to the requesting computing device 150, which may then display the estimate for the user on its display 158.

[0077] Figure 4A A graph 401 showing the rolling average speed of vehicles associated with observed traffic conditions as a function of the cumulative count of vehicles reporting average speeds is shown. Graph 401 shows that when the number of participating vehicles is small, the a priori speed is between 14 and 16 km / h. Due to the occurrence of traffic congestion events in the area, the rolling average speed gradually decreases to between 2 and 4 km / h. Figure 4B Shown in Figure 4A The data source is a graph 402 of vehicle speeds obtained under the same experimental conditions. Figure 4B Two peaks are shown, one between 0 and 4 km / h (post-congestion peak) and the other between 15 and 20 km / h (pre-congestion peak).

[0078] Figure 4C Shows the formation of Figure 4A and Figure 4B A graph 403 shows a prior vehicle speed statistical distribution 410, an observed vehicle speed statistical distribution 420, and a posterior vehicle speed statistical distribution 430 related to the data on which the graph is based. The posterior distribution 430 serves as a basis for estimating vehicle speed and travel time by taking into account the prior vehicle speed data and the observed vehicle speed data.

[0079] Figure 5A flow chart of a portion of a method 500 for generating a priori vehicle speed statistical distributions is shown. The driver application 510 is a smartphone application installed on the computing device 150 that triggers periodic transmission of data received from the GPS 157 to the system 100 or the historical database 120. The system 100 performs steps 1 to 6 based on data received from the GPS 157 of multiple computing devices 150. In step 1, the system 100 aggregates speed data based on roads, road segments, or road categories. Road categories can be related to road types or categories in urban areas, rural areas, paved roads, unpaved roads, etc., so as to roughly reflect the traffic patterns or speeds on a certain category of roads. Step 2 includes retrieving historical data for a given hour of the day and a day of the week over an n-week period. In step 3, the system 100 sorts the data points based on the timestamp and date of the data point generation.

[0080] At step 4.1, the system 100 evaluates whether a minimum number of samples associated with the road segment are available to generate a statistically meaningful prior distribution. If a sufficient number of samples are not available, then at step 4.2, data points from the relevant road category are used to generate the prior distribution. Data from the relevant road category may be used in place of the samples, or may be added to the samples so that there are a sufficient number of data points (or records) to generate a meaningful prior distribution. However, if a sufficient number of samples are available, then at step 5, as described in reference Figure 2 As described in step 210, various statistical indicators representing the prior distribution are calculated. In step 6, the outputs of steps 5.1, 5.2 and 5.3 are stored / written in the memory 104 of the system 100. "Sufficient number" can be a predetermined number.

[0081] Reference in this specification to any prior publication (or information derived therefrom) or any known matter is not and should not be taken as an acknowledgment or permission or any form of suggestion that the prior publication (or information derived therefrom) or known matter forms part of the common general knowledge in the field relevant to the present specification.

[0082] In this specification and the claims that follow, unless the context requires otherwise, the word "comprise" and variations such as "comprises" and "comprising", will be understood to imply the inclusion of stated integers or steps or groups of integers or steps but not the exclusion of any other integers or steps or groups of integers or steps.

[0083] The scope of the present disclosure includes all changes, substitutions, variations, alterations and modifications to the exemplary embodiments described or shown herein that will be understood by those of ordinary skill in the art. The scope of the present disclosure is not limited to the exemplary embodiments described or shown herein. Moreover, although the present disclosure describes and illustrates each embodiment herein as including specific parts, elements, features, functions, operations or steps, it should be understood by those of ordinary skill in the art that any of these embodiments may include any combination or arrangement of any parts, elements, features, functions, operations or steps described or shown anywhere herein. Although the present disclosure describes or illustrates specific embodiments as providing specific advantages, specific embodiments may not provide these advantages, or provide some or all of these advantages.

Claims

1. A system for estimating vehicle speed along a length of road, the system comprising: one or more processors (processors); a memory accessible to the processor, the memory comprising program code executable by the processor to: Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road; receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road; performing a Bayesian fusion of the prior distribution and the observed vehicle speed data by expressing the prior distribution as a conjugate of a likelihood function to determine a posterior distribution; A vehicle speed estimate is determined based on the posterior distribution.

2. The system according to claim 1, wherein: The fusion of the prior distribution and the observed vehicle speed data includes: exponentially smoothing the prior vehicle statistical distribution to incorporate the observed vehicle speed data.

3. The system according to claim 1, wherein: The prior distribution is calculated based on historical vehicle data for the time of day and day of the week associated with the estimated time of vehicle speed.

4. The system according to claim 1, wherein: The processor performs Bayesian fusion of the prior distribution and the observed vehicle speed data by calculating hyperparameters, the hyperparameters comprising at least one of the following: The variance of the historical vehicle speed data; an exponential moving average of the historical vehicle speed data; an average value of standard deviations of the historical vehicle speed data; Mean of real-time or near real-time vehicle speed data; as well as a sample count of the real-time or near real-time vehicle speed data; Wherein the calculated hyperparameters define a closed-form solution as a normal-normal conjugate pair of the prior distribution and the posterior distribution to perform the Bayesian fusion.

5. The system according to claim 4, wherein: The weight assigned to the observed vehicle speed data during the fusion is proportional to the sample count of the observed vehicle speed received.

6. The system according to claim 1, wherein: The processor infers the distribution of the observed vehicle speed data using a Markov Chain Monte Carlo (MCMC) method or a variational inference method.

7. The system according to claim 1, wherein: The fusion of the prior vehicle seed data and the observed vehicle seed data is performed by modeling the prior vehicle speed distribution and the posterior vehicle speed distribution as normal distributions.

8. The system according to any one of claims 1 to 7, wherein: The Bayesian fusion is performed in real time or near real time upon receipt of the observed vehicle speed data.

9. A system for estimating travel time, the system comprising: one or more processors (processors); a memory accessible to the processor, the memory comprising program code executable by the processor to: receiving, from a computing device, a starting location, a destination location, and a proposed travel time; Determining a driving route between the starting location and the destination location, the driving route comprising at least one road segment; estimating the travel time by estimating the segment travel time for each segment and summing the segment travel times; The travel time of the road section is estimated in the following manner: Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road segment; receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road segment; performing a Bayesian fusion of the prior distribution and the observed velocity data by expressing the prior distribution as a conjugate of a likelihood function to determine a posterior distribution; determining an estimated vehicle speed estimate based on the posterior distribution; The road segment travel time is determined based on the vehicle speed estimate and the road segment length.

10. The system according to claim 9, wherein: The processor is also configured to transmit the estimated travel time to the computing device.

11. A computing device comprising: one or more processors (processors); a memory accessible to the processor, the memory comprising program code executable by the processor to: transmitting the starting location, the destination location, and the proposed travel time to the system according to claim 10; An estimated travel time is received from the system of claim 10.

12. A method for estimating vehicle speed along a length of road, the method comprising: Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road; receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road; performing a Bayesian fusion of the prior vehicle speed data and the observed vehicle speed data to determine a posterior distribution by expressing the prior distribution as a conjugate of a likelihood function; A vehicle speed estimate is determined based on the posterior distribution.

13. A method for estimating travel time, the method comprising: receiving, from a computing device, a starting location, a destination location, and a proposed travel time; determining a driving route between the starting location and the destination location, the driving route comprising at least one road segment; estimating a travel time estimate by estimating a segment travel time for each segment and adding the segment travel times; The travel time for each road segment is estimated as follows: Calculating a priori vehicle speed statistical distribution (prior distribution) based on historical vehicle speed data along the road segment; receiving observed vehicle speed data associated with real-time or near real-time vehicle traffic along the road segment; performing a Bayesian fusion of the prior distribution and the observed vehicle speed data by expressing the prior distribution as a conjugate of a likelihood function to determine a posterior distribution; determining a vehicle speed estimate based on a posterior vehicle speed distribution; The road segment travel time is determined based on the vehicle speed estimate and the road segment length.

14. One or more non-transitory computer-readable storage media storing instructions which, when executed by one or more processors, cause the one or more processors to perform the method according to any one of claims 12 and 13.

Images