System and method for determining energy requirements of a vehicle for a trip

Abstract

A system for determining an energy demand of a vehicle for a trip is described. The system may include a predictor mechanism that predicts the vehicle energy demand for a trip using an energy prediction algorithm. The system includes an updater mechanism that is configured to improve the energy prediction algorithm for the vehicle by determining, for each of a plurality of historical trips undertaken by the vehicle, an error between an actual vehicle energy usage for the historical trip and a predicted energy usage derived using the energy prediction algorithm for the historical trips. An aggregate error is calculated based on the errors for the plurality of historical trips. The updater is arranged to adjust the energy prediction algorithm to reduce the aggregate error.

Description

Technical Field

[0001] The present invention relates to a system and method for determining the energy requirements of a vehicle for a journey, and is particularly applicable to electric vehicles, although not exclusively. The system and method may be used as part of a vehicle dispatch system. Background Art

[0002] Existing vehicle range prediction systems work by extrapolating the vehicle's recent (e.g., within the past 15 minutes) consumption data to provide an indication of the vehicle's remaining range before refueling or charging. They do not take into account the full nature of the journey the vehicle is to make, so the range returned is often inaccurate. This is particularly problematic for vehicles with long refueling / charging intervals or in locations where refueling / charging stations are sparse. Summary of the invention

[0003] The present invention has been developed to improve the above-mentioned problems.

[0004] In a first aspect of the present invention, a system for determining an energy requirement of a vehicle for a trip is provided, the system being adapted to: obtain trip data relating to the trip and obtain vehicle data relating to the vehicle; the system comprising a predictor mechanism, the predictor mechanism using an energy prediction algorithm to predict the vehicle energy requirement for the trip based on the trip data and the vehicle data; the system comprising an updater mechanism, the updater mechanism being configured to adjust the energy prediction algorithm of the vehicle by: for each of a plurality of historical trips undertaken by the vehicle, determining an error between actual vehicle energy usage for the historical trip and predicted energy usage derived using an energy prediction algorithm for the historical trip based on the historical trip data and the historical vehicle data, each of the plurality of historical trips being associated with a corresponding error in a set of errors; determining a collective error from the set of errors; and adjusting the energy prediction algorithm to reduce the collective error.

[0005] In a second aspect of the present invention, a computer-implemented method for determining an energy requirement of a vehicle for a trip is provided, the method comprising: obtaining trip data and obtaining vehicle data related to the vehicle; using a predictor mechanism to predict the vehicle energy requirement for the trip based on the trip data and the vehicle data using an energy prediction algorithm; adjusting the vehicle's energy prediction algorithm by the following steps: for each of a plurality of historical trips undertaken by the vehicle, determining an error between actual vehicle energy usage for the historical trip and predicted energy usage derived from the historical trip data and the historical vehicle data using an energy prediction algorithm for the historical trip, each of the plurality of historical trips being associated with a corresponding error in a set of errors; determining a collective error from the set of errors; and adjusting the energy prediction algorithm to reduce the collective error.

[0006] According to another aspect of the present invention, there is provided a dispatch system implemented by one or more computers for selecting a vehicle from a plurality of vehicles to undertake a trip in response to receiving a trip request from a user of the system, the system comprising:

[0007] A receiver, configured to receive a travel request from the user;

[0008] means for determining, for each of the plurality of vehicles, an energy requirement of a vehicle that is a trip subject of the trip request;

[0009] a vehicle selector that uses the determined energy requirement to select a vehicle from the plurality of vehicles to undertake the trip; and

[0010] Means for sending a trip instruction including information related to a trip object of the trip request to a selected vehicle; and wherein the means for determining, for each of the plurality of vehicles, an energy requirement of a vehicle to undertake the trip object of the trip request comprises the above-mentioned system for determining an energy requirement of a vehicle for a trip. The selected vehicle may be an autonomous vehicle, the autonomous vehicle including a vehicle controller (e.g., implemented by a computer system) driving the vehicle, and wherein, in response to the controller receiving the trip instruction, the vehicle is caused to move away from its current location to undertake the trip.

[0011] By tuning the energy prediction algorithm to minimize the aggregate error of historical trips, the algorithm can more accurately predict the energy requirements of future trips undertaken by a specific vehicle and driver combination.

[0012] In this application, the term "driver" includes both the electronic systems used to drive the autonomous vehicle as well as more conventional humans.

[0013] The following features apply to all aspects of the invention above.

[0014] In the case where the energy storage device of the vehicle is a chemical energy storage device (such as, for example, a hydrogen battery cell or a gasoline tank), the on-board energy data may include an indication of the fuel level and / or pressure. In the case where the energy storage device includes one or more batteries, capacitors, etc., the on-board energy data may include an indication of the charge level.

[0015] The system may be arranged to retrieve vehicle data from the vehicle. The vehicle data may include: changing vehicle data, which is data that is expected to change over time, such as between trips or within a trip; and invariant vehicle data, which is expected to remain substantially static. While the vehicle is in operation, the changing vehicle data may be repeatedly retrieved from the vehicle, for example at least once every 60 seconds. This ensures that the predictor mechanism can use relatively accurate vehicle condition data to determine the prediction. This also means that the location of the vehicle can be known with reasonable accuracy so that the system can determine when the vehicle has completed a trip.

[0016] Examples of changing vehicle data include one or more of the following: vehicle location (e.g., provided by a global navigation system), energy storage system health (e.g., maximum storage capacity degradation over time), vehicle system temperature (e.g., temperature of the energy storage device), and tire pressure. This data may be obtained from the vehicle's computer.

[0017] The system may include a vehicle device (e.g., a computing device) arranged to receive vehicle data from a vehicle (e.g., via one or more sensors) and to transmit the received vehicle data, typically wirelessly, to be received by the predictor mechanism. In one arrangement, the vehicle device is connected to a computer of the vehicle, which is connected to the vehicle sensors. In a variant embodiment, the vehicle device includes a vehicle computer arranged to directly transmit the vehicle data. The vehicle data may be received by the prediction system including the predictor mechanism and the updater mechanism.

[0018] Invariant vehicle data may include one or more of: vehicle dimensions; curb mass; drag coefficient of the vehicle chassis; auxiliary loads (e.g. headlight consumption, internal heating system consumption) and type of tires installed. Since this information is not expected to change on a trip-by-trip basis, it is not typically obtained directly from the vehicle, but rather from other sources such as vehicle specifications published by the manufacturer and user input. Invariant vehicle data may be obtained when a vehicle is added to the system, or in the first instance of a vehicle of the same make being added to the system.

[0019] The system may be arranged to determine the energy requirements of a plurality of vehicles, in which case the system may obtain vehicle data from each vehicle for which a trip prediction is made. The system may be arranged to maintain a separate energy prediction algorithm for each vehicle. The system may be arranged to update each energy prediction algorithm based on historical trips performed by its associated vehicle. Thus, over time, each energy prediction algorithm will be adjusted to reflect differences in actual energy usage of its associated vehicle due to different characteristics, ageing and driving style of its driver.

[0020] Both the changing vehicle data and the invariant vehicle data may be used to predict the energy requirements of the vehicle for a trip.

[0021] The trip data may include, for example, the start and end points of the trip, geographic information (e.g., geographic coordinates) of waypoints, and / or other route information. It may also include further information, such as one or more of the following: capacity (e.g., number of passengers), route restrictions (e.g., to avoid tolls), and speed limits imposed on the vehicle (e.g., by the presence of an onboard speed limiter). The trip data may also include one or more of the following: local weather (i.e., weather applicable to the trip), such as wind speed, precipitation, temperature, air pressure, visibility, humidity, and cloud cover; topography of the route; route distance; traffic information (e.g., average vehicle speed along one or more segments of the route; stop and turn information—e.g., number of stops and turns; road surface, its type and / or condition; and sunset / sunrise times. The trip data may be retrieved, at least in part, from one or more third party providers (e.g., metering organizations and traffic information services). The trip data may be received, at least in part, via the Internet.

[0022] The system may be arranged to receive a request for a prediction of energy requirements for a trip from a trip requesting device (eg a user via a computer, a satellite navigation system or a vehicle dispatch system).The request may include the trip data or some of the trip data.

[0023] The system may be adapted to retrieve on-board energy data from the vehicle. The on-board energy data may be used to derive actual vehicle energy usage for historical trips.

[0024] Actual vehicle energy usage may be derived using on-board energy data retrieved from the vehicle at the start and end of historical trips. The system may determine that a trip has ended using geographic information retrieved from the vehicle and / or by the user indicating that the trip has ended.

[0025] The system may be adapted to use on-board energy data and a prediction of vehicle energy required for a trip from a predictor mechanism to provide an output indicating whether a trip is feasible with the current energy on-board the vehicle.

[0026] The system may be adapted to use on-board energy data and a prediction of vehicle energy required for a trip from a predictor mechanism to provide an indication of the remaining power that can be expected on-board the vehicle at the end of the trip.

[0027] The system may be arranged to store in the memory repository as historical data: trip data and / or vehicle data; or data derived from trip data and / or vehicle data used and / or derived by the predictor mechanism to determine a trip prediction. The updater mechanism may use the stored historical data to determine a prediction for the trip as a historical trip. The memory repository may also hold an energy prediction algorithm used by the predictor mechanism to predict the energy demand for the trip.

[0028] Predicting the energy demand of the vehicle for a trip may include determining: a plurality of factors for the trip, each factor derived from vehicle data and / or trip data; applying a weight to each factor; and combining the weighted factors to derive a total energy demand.

[0029] These factors may include one or more of: distance travelled, ascent, descent, one or more speeds along the route of travel; expected frequency of stops / starts; payload; location of refuelling stops; auxiliary loads; drag and impact of energy recovery systems.

[0030] The trip distance factor may be derived, for example, from trip data (e.g., start and end locations and map data). For example, the auxiliary load factor may be derived using invariant vehicle data (e.g., headlight and heater wattage and trip data such as outside temperature, visibility, and sunset / sunrise times).

[0031] The above factors relate to specific properties associated with the calculation of kinetic energy, potential energy, drag energy, etc. Alternatively or additionally, abstract factors may be used instead, each derived from a different combination of one or more of the vehicle data and the trip data, but having no direct relationship to real-world energy measurements.

[0032] The system can learn the specific individual contributions of each factor to the vehicle and driver's overall energy consumption for a trip in order to make better predictions.

[0033] In an implementation of this, the system may be arranged to update the energy prediction algorithm by adjusting the weights of the factors to minimize the aggregate error.

[0034] The predictor mechanism may be adapted to include a safety margin within the predicted vehicle energy demand for a trip, and the updater adjusts the size of the safety margin according to the size of the individual errors of the historical trips.Updating the prediction algorithm to minimize the aggregate error allows the safety margin to be reduced.

[0035] For each historical trip, the updater is arranged to use the energy prediction algorithm to derive the predicted energy usage by the following steps:

[0036] Retrieving and / or determining values ​​of a plurality of factors for the historical trips, the factors being derived from trip data associated with the historical trips and / or vehicle data associated with the historical trips; applying a weight to each factor; combining the weighted factors; and wherein the updater is further arranged to adjust the weight of one of the factors to reduce the aggregate error, the adjustment being applied to the corresponding factor for each historical trip.

[0037] The system may include a prediction system including the functionality of a predictor mechanism and an updater mechanism.The prediction mechanism may be connected to a network to receive vehicle information, trip information, and trip requests via the network.

[0038] The method may include predicting a vehicle energy demand for a trip, including: determining a plurality of factors for the trip, each factor derived from trip data and / or vehicle data; applying a weight to each factor; and applying the weighted factors in an energy prediction algorithm to derive a predicted vehicle energy demand for the trip.

[0039] The method may include, for each historical trip, using an energy prediction algorithm to derive a predicted energy usage by:

[0040] Retrieving and / or determining values ​​of a plurality of factors for historical trips, the factors being derived from trip data associated with the historical trips and / or vehicle data associated with the historical trips; applying a weight to each factor; and applying the weighted factors in an energy prediction algorithm to derive a predicted vehicle energy demand for the historical trips; and wherein an adjustment is applied to the weight of one of the factors to reduce an aggregate error, the adjustment being applied to the corresponding factor for each historical trip.

[0041] The functions of the predictor mechanism and updater mechanism may be implemented by appropriately programmed circuits comprising one or more processors and memory using techniques known to those skilled in the art.

[0042] In another aspect, the method may be embedded in a computer program product embodied in one or more non-transitory computer readable media, the computer program product comprising all the features enabling the implementation of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The present invention will now be described by way of examples with reference to the accompanying drawings.

[0044] Figure 1 is a schematic diagram of a system for determining the energy requirements of a vehicle for a trip;

[0045] Figure 2 is a schematic diagram of a process for deriving an energy demand forecast for a trip;

[0046] Figure 3 is a flow chart of a process for updating an algorithm used to derive an energy demand forecast for a trip;

[0047] Figure 4 is a schematic diagram of a variant system for determining the energy requirement of a vehicle for a trip;

[0048] Figure 5 is a flow chart of a process for filtering historical trips used to update the algorithm;

[0049] Figure 6 is a flow chart of a process for improving an estimate of the number of stops and starts in a trip by identifying a level of congestion associated with the trip;

[0050] Figure 7 is a flow chart of a process for improving the accuracy of road elevation data; and

[0051] Figure 8 is a schematic diagram of a vehicle dispatch system that uses Figure 4 The system determines the energy requirements of the vehicle for a trip in order to select a vehicle from the fleet to make the trip. DETAILED DESCRIPTION

[0052] Reference Figure 1 , a system 1 for determining the energy requirement of a vehicle for a journey is shown. The system 1 comprises a vehicle device 2 located in a vehicle 3 and a prediction system 4 .

[0053] The prediction system 4 includes a predictor mechanism 4A, a memory library 4B, and an updater mechanism 4C. The memory may include any one or more of the following: volatile memory, non-volatile memory, computer storage system, etc. The prediction system 4 is implemented by one or more appropriately programmed processors communicatively coupled to the memory and other computer components using techniques known to those skilled in the art. A typical combination of hardware and software for implementing the prediction system 4 may be a general purpose computer system with a computer program that, when loaded and executed, controls the computer system so that it performs the methods described herein.

[0054] The vehicle device 2 includes one or more processors communicatively coupled to a memory and other computer components, and is arranged to receive onboard energy data and other current vehicle data including geographic location data from a vehicle 3, and relay the current vehicle data to a prediction system 4 via a network 5 including a wireless receiver (not shown) via a wireless transmitter 2A (which may be a wireless transceiver).

[0055] In one arrangement, the vehicle device 2 is arranged to be connected to the vehicle's onboard computer and receive said vehicle data from the computer; however, in another arrangement, the computer itself may be arranged to send the vehicle information, so that the vehicle device may be omitted. The functions of the device may be performed by a mobile (cellular) telephone running an appropriate application; in this case and if the mobile phone is located on the vehicle, the vehicle's geographical location information may be derived from the mobile phone's own global navigation chip.

[0056] The system 1 is adapted to receive a request for a trip forecast from a trip request device 6, for example from a user (possibly in a vehicle) via a computer, a satellite navigation system or a vehicle trip dispatch system (such as might be used by a taxi or courier company). The request identifies the vehicle and driver combination that will perform the trip for which the forecast is requested, and includes trip information identifying the destination and any waypoints. Advantageously, the trip information also includes further information such as load capacity (e.g. number of passengers), route restrictions (e.g. to avoid tolls) and speed limits imposed on the vehicle.

[0057] The trip prediction request is received by the prediction system 4 via the network 5. In response, the predictor 4A implementing the energy usage algorithm stored in the memory bank 4B uses the current changing vehicle data received from the vehicle device 2 via the network 5 and the unchanging vehicle data related to the vehicle from the memory bank 4B and the trip data related to the trip from the external source 7 (e.g., one or more third-party data providers) to derive an energy demand prediction. A safety margin is applied to the calculated prediction to take into account the expected error in the predicted trip energy usage compared to the actual energy usage of the trip. The derived prediction (including the safety margin) is returned to the trip request device 6 via the network 5.

[0058] If the predicted energy demand (including safety margin) exceeds the on-board energy of vehicle 3 provided by vehicle device 2 (for example, the on-board energy data of an electric vehicle can be derived from the power data and the battery health status data), predictor 4A returns an indication to trip request device 6 that the trip cannot be achieved using the on-board energy.

[0059] Where a user submits a request, the request may be submitted via a computer (including a portable computer such as a smartphone) The request may be received from a satellite navigation system and may be used by the satellite navigation system to determine a suitable route for the journey, e.g., a route that best fits a given vehicle charge.

[0060] When the system is used in conjunction with a dispatch system 6, the dispatch system can use the system 1 to provide a forecast of a group of vehicles that are available to perform a trip that includes travel between at least two locations (e.g., in response to a customer's request for a taxi). The dispatch system can use the forecast to determine which vehicle in the group of vehicles is best suited to perform the trip given the different locations and charges of the vehicles.

[0061] The predictions are stored in the memory repository 4B together with the data used to create the predictions.

[0062] This process is repeated each time a trip prediction request is made for a vehicle-driver combination.

[0063] After the trip is completed, the system uses onboard energy data from the vehicle device 2 at the beginning and end of the trip to determine the actual energy use of the trip. The system 1 can determine that the vehicle has completed the trip by, for example, an indication from the user / dispatch system and / or geolocation information received from the vehicle device 2.

[0064] Reference Figure 2 To derive an energy usage forecast for a trip, the predictor 4A uses the vehicle data and the trip data to determine the various energy composition factors F1-F n These predictions are summed (or otherwise combined) to provide the predicted energy usage for the trip. n Associated with each factor. The associated weight is applied to (eg, multiplied by) each individual energy component factor F, such as (F1.W1, F2.W2....F n .W n ) and the weighted factors are summed:

[0065]

[0066] To provide a weighted predicted energy usage forecast (PE). For a new vehicle-driver combination, the weights may be initially set to have a neutral effect, i.e., so that the weighted forecast is the same as the non-weighted forecast. Alternatively, the weights may be selected to reflect the weight of another similar vehicle for which historical trip information has been recorded.

[0067] The system may determine the trip as including a route from a destination provided by the user / dispatch system to a nearby gas station. The system may also determine the trip as including a route from the vehicle's current location to a starting location provided by the user / dispatch system.

[0068] The prediction system 4 runs the algorithm updater 4C periodically or at some irregular time interval to improve the algorithm.

[0069] Reference Figure 3 , the updater 4C obtains information about a set of historical trips stored in the memory library 4B (10). For each historical trip, the updater retrieves (or derives) the value of each energy component factor F (11), applies a weight W to each factor F and sums the weighted factors to derive the predicted energy usage for the historical trip (12). The predicted energy is compared with the actual energy used for the trip stored in the memory library 4B to determine an error E for the trip (13). This produces a set of errors E1-E j , where j is the number of trips in the historical trip group. Error E1-E j are combined (eg, summed) to provide the aggregate error E t (14).

[0070] The updater then adjusts the weight W(15) by the factor F, making the same adjustment to the same factor for each historical trip, resulting in a new set of errors E1-E j If the adjustment produces a reduced set error E t , the algorithm stored in the memory bank 4B is updated with the new weight W (16). This process is iterated to update the weight of each factor, for example W1, then W2... to W n (17).

[0071] When updating weight W for factor F, the updater may anchor the value of weight W such that it becomes increasingly difficult for the weight value to move away from its anchor value as the difference between the weight value and the anchor value increases.

[0072] Anchoring is useful for providing prediction estimates for trips that have significantly different characteristics than trips that the vehicle typically undertakes. For example, a vehicle that rarely encounters elevation changes may, without anchoring, have a weight W for elevation that drifts to an unreasonable value, so that when the vehicle actually undertakes a trip that includes significant elevation changes, the energy prediction will have significant errors.

[0073] Repeat this process until the set error E t is within a desired tolerance or reaches a desired number of iterations (18). The refined algorithm is stored in the memory bank 4B. It will be appreciated that the order in which the factor weights are varied need not be particular.

[0074] By varying the weights, the algorithm is refined to account for the uniqueness of a particular vehicle and driver combination that affects energy consumption (such as changes in the driver's driving style and vehicle conditions), and therefore becomes more accurate in predicting energy requirements for future trips.

[0075] Those skilled in the art will appreciate that, in addition to the above-mentioned gradient descent method, there are many other possible methods of optimizing the algorithm to reduce the set error E t , including, for example, solving a set of simultaneous equations. It will be appreciated that the exact method used is not necessarily important.

[0076] As part of the refinement process, the updater 4C determines a safety margin to be applied to the calculated predicted PE. For the historical trips resulting from the refined algorithm, the size of the safety margin is proportional to the individual errors E i -E j The error E i -E j As the size of the individual errors decreases, the size of the safety margin used will also decrease. The size of the safety margin can be derived using the standard deviation of the individual errors, the size of the largest single error, or some other means.

[0077] In a variant embodiment, the system may be located as a whole in the vehicle on which the journey is being made and may, for example, form part of a vehicle satellite navigation system having the functionality of a prediction system executed by a computer of the vehicle. In such an arrangement, the vehicle device may be omitted. The user may provide a request for journey information via an interface of the vehicle's satellite navigation system.

[0078] The system may be adapted to determine the energy requirement for a trip without live vehicle data. In this variant arrangement, the system may use the latest vehicle data it has for the vehicle and / or values ​​provided by a user or a dispatch system.

[0079] Figure 4 A schematic diagram of a variant system for determining the energy requirement of a vehicle for a trip is shown, the variant system including features of a filter 4D and a pre-processor 4E. The functions of the filter and pre-processor will be described below with reference to Figure 5 , Figure 6 and Figure 7 Give a description.

[0080] As mentioned above Figure 3 As described, the ability of the updater to refine the algorithm depends in part on the quality of the historical trip information used in the refinement process. In particular, there are occasions where the trip data used to make historical predictions for historical trips is of poor quality, resulting in inaccurate historical predictions, such that including it in the historical trip set (10) would harm the refinement of the algorithm. In order to reduce the likelihood of using poor quality data, the prediction system may include a filter 4D for filtering the historical trip information used by the updater to refine the algorithm.

[0081] The process flow of filter 4D is Figure 5 . A set of historical trip data for historical trips is retrieved (20) from the memory bank 4B, the retrieved historical trip data comprising actual values ​​of certain characteristics of the trip, such as, for example, one or more of: the actual distance of the trip, the duration, the total elevation gain made during the trip and the total elevation loss made during the trip (21); and predicted values ​​of these characteristics used to make historical predictions (21). Typically, the actual values ​​will have been obtained from the onboard device 2; however, the values ​​used for the predictions have been obtained from an external source. A filter compares each actual value with the predicted value (22, 23, 24, 25) to determine the difference.

[0082] If the difference between the actual and predicted values ​​is within a threshold range (the threshold range may be different for each characteristic), the quality of the trip data used to make the prediction is considered to be good enough, and the prediction for that trip is included in the set of historical trips used to refine the algorithm (26). If the difference between the actual and predicted values ​​is greater than the threshold range, the trip data used to make the prediction is considered to be of poor quality, and the prediction for that historical trip is excluded from the set of historical trips used by the updater to refine the algorithm (27).

[0083] The prediction system may further comprise a pre-processing mechanism 4E for processing the trip data for energy prediction of the trip and for refining the algorithms to improve the quality of the trip data used for these purposes and thus, as a corollary, to improve the predictions and refinements. Figure 6 and Figure 7 Two examples of the process used by the pre-processing mechanism to improve the trip data are shown respectively.

[0084] A typical approach to predicting the number of stops and starts a vehicle makes during a trip is to determine from the trip data the number of intersections the trip includes, and each intersection includes a stop and a start. This makes the determination of the number of stops and starts often inaccurate, especially for trips through urban routes. Figure 6 The process shown can improve the prediction of the number of stop-start events for a trip using trip data.

[0085] Route information for a journey for which an energy forecast has been requested is obtained (30), the route is divided into a number of segments and for each segment, journey data is obtained (31) including a speed limit for the segment and a predicted average traffic speed, both values ​​typically obtained from an external traffic data source. A ratio of the average traffic speed to the speed limit is calculated (32), the ratio providing an indication of road congestion. The ratio is used to calculate a predicted number of additional stop-start events for the segment (33). This process is repeated for each segment of the journey (34) in order to predict the total number of additional stop-start events for the journey. The predicted number of stop-start events forms an input to a predictor 4A.

[0086] Figure 7 The process shown in is used to modify trip data representing the terrain of an area in order to provide more accurate values ​​for predicted vehicle elevation gain and elevation loss for the trip. The terrain information for a route will generally provide a fairly accurate indication of the grade and elevation changes of the road; however, significant differences may occur when the road passes through areas such as a tunnel through a mountain range or a bridge across a canyon. In such cases, the grade of the road may be relatively flat compared to what would be expected from the terrain data.

[0087] Figure 7The process shown includes obtaining route information for a trip for which a trip prediction request has been made (40). The route is divided into a plurality of segments, and for each segment, trip information related to the terrain of the segment is identified (41), the terrain information being used to determine one or more elevation changes within the segment. For each elevation change, a slope is calculated (42), the calculated slope is compared to a threshold slope value (43), which may be set, for example, to a known maximum slope for roads in the region, country, etc., or to select a limit of expected reasonable slopes. If the calculated slope is above the threshold, the calculated slope is ignored (44) and a default elevation change (e.g., zero) is added to the aggregate elevation gain or elevation loss sum of the trip (45). If the calculated slope is below the threshold, the elevation change associated with the slope is added to the aggregate elevation gain or elevation loss sum of the trip (45). This process is repeated for all segments of the trip (46).

[0088] Although limiting to the maximum allowed elevation may still not accurately represent road slope, it improves the accuracy of the prediction.

[0089] Figure 8 A method for use in conjunction with a vehicle dispatch system 10 (implemented by a computer system) is shown. Figure 4 system (but it can replace Figure 1 , i.e., without pre-processors or filters), the vehicle dispatch system, in response to receiving a trip request from a user 9 (e.g., by running a software application of an electronic device that communicates with the vehicle dispatch system via the Internet), obtains energy forecasts for the requested trip for multiple vehicles 3. The dispatch system 10 includes a receiver (not shown) for receiving the trip request from the user 9 and a vehicle selector 8, which selects a vehicle from the multiple vehicles for the trip using the energy forecast from each of the multiple vehicles.

[0090] The vehicle dispatch system 10 sends a trip invitation or instruction to the selected vehicle, for example, via the network 5 (but it can be sent directly). The instruction may include route information (and / or other trip information). The invitation or instruction is received by the dispatch terminal. In response, the dispatch terminal can display the invitation / instruction to the driver of the vehicle (if it is a non-autonomous driving vehicle) for the driver to accept or reject (if it is an invitation). The dispatch terminal can also display detailed information of the trip. If the vehicle is autonomous, the instruction is passed to the vehicle controller (or it can directly receive the instruction), which in response is used to automatically drive the vehicle to the starting point of the trip and undertake the trip.

[0091] exist Figure 8In FIG. 1 , the dispatch terminal and the vehicle controller are shown as part of the onboard device 2. In a variation, one or both of the dispatch terminal and the vehicle controller may be separated from the onboard device and from each other. The vehicle need not contain both the dispatch terminal and the controller.

[0092] Refer to above Figure 1 The prediction system 4, vehicle device 2, and trip request device 6 discussed above may each include various components. Some of these components that may be used by certain example implementations of a computer system are discussed below.

[0093] The prediction system 4 may include at least one processor / controller communicatively coupled to a persistent (non-volatile) memory 4B of the prediction system 4. A bus architecture communicatively couples the processor / controller of the prediction system 4 to the memory 4B and other components of the prediction system 4. The bus architecture facilitates communication between various system components in the prediction system 4.

[0094] The processor / controller of the prediction system 4 is communicatively coupled to one or more network interface devices of the prediction system 4. For example, but not limitation, the network interface device may include at least one wireless communication transceiver device (e.g., at least one wireless communication access point transceiver device). The one or more network interface devices of the prediction system 4 are communicatively coupled to one or more communication networks 5 to communicate with one or more of a trip request device, a vehicle device, and an external source, such as for third-party data. The network interface device may communicate with one or more communication networks, such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet).

[0095] The trip request device 6 may include at least one processor / controller of its own, which is itself communicatively coupled to memory, which may include main memory and persistent (non-volatile) memory. A bus architecture communicatively couples the processor / controller to the memory and other components of the trip request device 6. The bus architecture facilitates communication between the various system components in the trip request device 6.

[0096] The processor / controller of the trip request device 6 is communicatively connected to at least one user interface of the electronic device. The user interface includes a user input interface and a user output interface. Examples of elements of the user input interface may include a keyboard, a keypad, a mouse, a touch control pad, a touch pad, a touch screen, and a microphone for receiving audio signals. The received audio signal may be converted into an electronic digital representation of the audio signal and stored in a memory, and may optionally be used with a voice recognition software executed by the processor / controller to receive user input data and commands. Examples of elements of the user output interface may include a display, a speaker, one or more indicator lights, one or more transducers that produce audible indications, and a tactile signal generator. Some examples of displays are liquid crystal displays (LCDs), plasma displays, organic light emitting diode (OLED) displays, and the like. A touch screen display (also referred to as a touch input screen) is used as both an output interface device and an input interface device, and the input interface may receive direct user input in the form of "touch", "slide" or "tap".

[0097] The processor / controller of the trip request device 6 is communicatively coupled to one or more network interface devices. For example, but not limited to, the network interface device may include at least one wireless communication transceiver device (e.g., at least one wireless communication access point transceiver device). The one or more network interface devices are communicatively coupled to one or more communication networks 5. The network interface device may communicate with one or more communication networks 5 such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet).

[0098] The processor / controller of the trip request device may be communicatively coupled to a GPS receiver in one arrangement. That is, GPS data received by the GPS receiver may be used by the processor / controller of the trip request device 6 to determine the geographic location of the trip request device 6. The geographic location of the trip request device 6 may be considered to be the geographic location of the vehicle.

[0099] The vehicle device 2 may include at least one processor / controller communicatively coupled to a memory of the vehicle device 2, which may include main memory and persistent (non-volatile) memory. The bus architecture of the vehicle device 2 communicatively couples the processor / controller of the vehicle device 2 to the memory of the vehicle device 2 and other components of the vehicle device 2. The bus architecture facilitates communication between various system components in the vehicle device 2.

[0100] The processor / controller of the vehicle device 2 is communicatively coupled to one or more network interface devices of the vehicle device 2. For example, but not limited to, the network interface device may include at least one wireless communication transceiver device (e.g., at least one wireless communication access point transceiver device). The one or more network interface devices of the vehicle device 2 are communicatively coupled to one or more communication networks 5. The network interface devices of the vehicle device may communicate with one or more communication networks 5 such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet).

[0101] The processor / controller of the vehicle device 2 may be communicatively coupled to the GPS receiver of the vehicle device 2. That is, GPS data received by the GPS receiver of the vehicle device 2 may be used by the processor / controller of the vehicle device 2 to determine the geographic location of the vehicle device 2 and, therefore, the vehicle 3.

[0102] The processor / controller of the vehicle device 2 may be communicatively coupled to one or more sensor devices (or multiple sensors) of the vehicle 3, either directly or via the vehicle's computer. The sensors may include various types of sensor devices that provide sensor data to the processor / controller of the vehicle device 2. Examples include one or more of tire pressure sensors, onboard energy sensors (e.g., fuel level sensors or charge sensors), vehicle system temperature sensors, sensors for detecting the operating status of auxiliary electrical devices (e.g., vehicle interior heaters and headlights).

[0103] One or more of the predictor system, the vehicle device, and the trip request device may include a power system to provide power for operation thereof.

[0104] The processor / controller of trip request device 6 and the processor / controller of vehicle device 2 may communicate with the processor / controller of prediction system 4 via communication network 5 .

[0105] In an alternative arrangement, the functionality of the predictor mechanism, updater, vehicle device, and trip request device may be implemented by a single computer system located in the vehicle. In such an arrangement, the computer system may have a processor / controller coupled to persistent (non-volatile) memory, a bus architecture communicatively coupling the processor / controller to the memory and other components of the computer system, such as sensor devices, a GPS receiver, and a user interface device.

[0106] As will be appreciated by one of ordinary skill in the art, systems, methods, or computer program products may implement various embodiments of the claimed invention. Thus, aspects of the claimed invention may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, microcode, etc.), or an embodiment combining software and hardware aspects. Furthermore, aspects of the claimed invention may take the form of a computer program product embodied in one or more non-transitory computer-readable media having computer-readable program code embodied thereon.

Claims

1. A system for determining the energy requirement of a vehicle for a journey, the system being adapted to: obtaining trip data relating to the trip; and obtaining vehicle data associated with the vehicle; and The system includes a predictor mechanism that predicts vehicle energy demand for a trip based on the trip data and vehicle data using an energy prediction algorithm; The system includes an updater mechanism configured to adjust an energy prediction algorithm of the vehicle by: determining, for each of a plurality of historical trips undertaken by the vehicle, an error between an actual vehicle energy usage for the historical trip and a predicted energy usage for the historical trip derived using an energy prediction algorithm based on the historical trip data and the historical vehicle data, each historical trip in the plurality of historical trips being associated with a corresponding error in a set of errors; determining an aggregate error from the set of errors; and Adjust the energy prediction algorithm to reduce the ensemble error, Wherein the system comprises a first trip data comparator arranged to compare actual trip data for historical trips derived from vehicles undertaking the historical trips with predicted trip data used to create historical predictions of the trips, and to use the comparison to filter the historical trip data used by the updater mechanism to update the energy prediction algorithm.

2. The system according to claim 1, wherein: The predictor mechanism is configured to predict the vehicle energy demand for a trip by: determining a plurality of factors of the trip, each factor being derived from the trip data and / or vehicle data; Apply weights to each factor; and The weighted factors are applied in an energy prediction algorithm to derive a predicted vehicle energy demand for the trip.

3. A system according to claim 1 or 2, wherein for each historical trip, the updater mechanism is arranged to use the energy prediction algorithm to derive a predicted energy usage by: retrieving and / or determining values ​​of a plurality of factors for a historical trip, the factors being derived from trip data associated with the historical trip and / or vehicle data associated with the historical trip; Apply weights to each factor; and applying the weighted factors in an energy prediction algorithm to derive a predicted vehicle energy demand for historical trips; And wherein the updater mechanism is further arranged to adjust the weight of one of the factors to reduce the aggregate error, the adjustment being applied to the corresponding factor of each historical trip.

4. The system according to claim 3, wherein: The system is arranged to store in a memory repository: processed trip data and / or processed vehicle data; data derived from the trip data and / or vehicle data; and factors used by a predictor mechanism to predict vehicle energy demand for a trip.

5. The system of claim 1 adapted to retrieve on-board energy data from the vehicle.

6. The system according to claim 5, wherein: The system is adapted to use the on-board energy data to determine actual vehicle energy usage for historical trips undertaken by the vehicle.

7. The system according to claim 5 or 6, wherein: The system is adapted to use the on-board energy data and a prediction of vehicle energy required for the trip from the predictor mechanism to provide an output indicating whether the trip can be undertaken by the vehicle using current energy on-board the vehicle.

8. A system according to claim 5 or 6, wherein the system comprises a vehicle device arranged to be carried by the vehicle, the vehicle device being arranged to receive on-board energy data and / or vehicle data from the vehicle and to send the received on-board energy data and / or vehicle data to a predictor system comprising a predictor mechanism and an updater mechanism.

9. The system according to claim 1 or 2, wherein: The predictor mechanism is adapted to include a safety margin within the predicted vehicle energy demand for the trip, and wherein the updater mechanism adjusts the size of the safety margin based on the size of respective errors of the historical trips.

10. A system according to claim 1 or 2, comprising a second trip data comparator which compares, for a segment of a route of a trip, the predicted average traffic speed data for the segment with the speed limit data for the segment and uses the comparison to estimate the number of stop / start events for the segment.

11. A system according to claim 1 or 2, comprising a third trip data comparator which, for a segment of the route of the trip, compares the calculated slope of the road in the segment with a threshold gradient range, calculating the grade of the road using the elevation change within the segment obtained or derived from the terrain travel data for the segment; and If the calculated grade falls outside of the threshold gradient range, a different altitude change is assigned to the segment in order to determine a collective altitude gain or altitude loss sum for the trip.

12. A computer-implemented dispatch system for selecting a vehicle from a plurality of vehicles to undertake a trip in response to receiving a trip request from a user of the dispatch system, the dispatch system comprising: A receiver, configured to receive a travel request from the user; means for determining, for each of the plurality of vehicles, an energy requirement of a vehicle that is a trip subject of the trip request; a vehicle selector that uses the determined energy requirement to select a vehicle from the plurality of vehicles to undertake the trip; and means for sending a trip instruction including information related to a trip object of the trip request to the selected vehicle; and wherein The means for determining, for each of the plurality of vehicles, an energy requirement of a vehicle of a trip subject undertaking a trip request comprises the system of any one of claims 1 to 11.

13. The dispatching system according to claim 12, wherein: The trip instruction is received by a dispatch terminal of the selected vehicle, and the dispatch terminal displays information related to a trip object of the trip request on a display of the selected vehicle in response to receiving the trip instruction.

14. The dispatching system according to claim 12 or 13, wherein: The selected vehicle is an autonomous vehicle, which includes a vehicle controller that drives the autonomous vehicle, and wherein the vehicle controller uses the trip instruction to drive the autonomous vehicle away from its current location to undertake the trip.

Images