Vehicle power-on time prediction method, vehicle management method, and electronic device

Abstract

This application discloses a vehicle power-on time prediction method, a vehicle management method, and an electronic device. The vehicle power-on time prediction method includes: determining the historical travel probability, historical travel time, and corresponding historical travel time probability of the target vehicle on a target historical travel segment starting from the current location, based on the target vehicle's current location and historical usage information including the target vehicle's travel information on multiple historical travel segments; determining the predicted user's vehicle usage segment and the corresponding predicted user's vehicle usage time based on the target vehicle's power-off time, historical travel probability, historical travel time, and historical travel time probability; and obtaining the predicted power-on time of the target vehicle based on the predicted user's vehicle usage time. Thus, by comprehensively considering the historical travel time probability and the historical travel probability, the most likely predicted user's vehicle usage time is determined, resulting in a more accurate predicted vehicle power-on time that better meets user needs.

Description

Technical Field

[0001] This application relates to the field of vehicle management technology, and in particular to a method for predicting vehicle power-on time, a vehicle management method, and an electronic device. Background Technology

[0002] For vehicles that are in a parked, power-off state, the vehicle's power-on time can be predicted by forecasting the user's next usage time. Automatically powering on and starting the vehicle at the predicted power-on time allows the vehicle to better meet the user's needs.

[0003] Currently, the main method for predicting a user's next car usage time is to statistically analyze the frequency of the user's historical car usage times and use the most frequent usage times as the predicted usage times. However, since the actual car usage time is influenced by many factors, this method, which only considers historical usage times, results in low accuracy. Therefore, predicting the vehicle's power-on time based on this method is not very accurate. Starting the vehicle based on this predicted power-on time may result in the vehicle's power-on time not matching the user's actual travel time, thus impacting the user experience. Summary of the Invention

[0004] This application provides a vehicle power-on time prediction method, a vehicle management method, and an electronic device to solve the problems of low accuracy in predicting vehicle power-on time in the prior art, which affects user experience.

[0005] The process involves determining the probability of a target vehicle's journey, its historical journey time, and its historical journey time probability on a target historical route based on the current location and historical usage information. Then, based on the power-off time, historical journey probability, historical journey time, and historical journey time probability, the predicted user's most likely route and the corresponding predicted user's journey time are determined, resulting in the predicted power-on time for that route. Thus, the system predicts the vehicle's next journey route and the corresponding power-on time, ensuring that the predicted route and power-on time are the user's most desired routes and times. Furthermore, the travel time prediction is based on the specific route; different predicted routes will result in different predicted power-on times, leading to more accurate power-on time predictions and a better user experience.

[0006] To address the aforementioned technical problems, in a first aspect, this application discloses a method for predicting vehicle power-on time. The method includes: determining the power-off time, current location, and historical vehicle usage information of a target vehicle, where the historical usage information includes the target vehicle's driving information on multiple historical driving segments; determining the historical travel probability and historical travel time of the target vehicle on a target historical driving segment based on the current location and historical usage information, and determining the historical travel time probability corresponding to the historical travel time, where the target historical driving segment is the historical driving segment whose starting point is the current location among the historical driving segments included in the historical usage information; determining the predicted user driving segment and the predicted user driving time corresponding to the predicted user driving segment based on the power-off time, historical travel probability, historical travel time, and historical travel time probability; and obtaining the predicted power-on time of the target vehicle based on the predicted user driving time.

[0007] The vehicle power-on time prediction method provided in this application determines the historical travel probability and historical travel time of the target vehicle on a target historical travel segment starting from the current location based on the current location and historical vehicle usage information. It also determines the historical travel time probability corresponding to the historical travel time. This allows for the determination of the most likely user travel segment and the most likely user travel time for that segment based on the power-off time, historical travel probability, historical travel time, and historical travel time probability, thus obtaining the vehicle power-on time. In this way, the method predicts the next travel segment of the target vehicle and the travel time from the current location to that segment based on historical vehicle usage information. This ensures that the predicted travel segment and power-on time are the travel segments and travel times that the user needs most. Furthermore, the travel time prediction is based on the travel segment; different predicted travel segments will result in different predicted power-on times, making the power-on time prediction more accurate and improving the user experience.

[0008] According to another specific implementation of this application, the method for predicting vehicle power-on time disclosed in this application determines the predicted user's vehicle usage segment and the predicted user's vehicle usage time corresponding to the predicted user's vehicle usage segment based on the power-off time, historical travel probability, historical travel time, and historical travel time probability. This includes: using the historical travel time that is later than the power-off time and has the highest historical travel time probability as the target historical travel time; determining the predicted travel probability corresponding to the target historical travel segment based on the historical travel probability of the target historical travel segment and the historical travel time probability corresponding to the target historical travel time; using the target historical travel segment with the highest predicted travel probability as the predicted user's vehicle usage segment, and using the target historical travel time corresponding to the predicted user's vehicle usage segment as the predicted user's vehicle usage time.

[0009] By employing the above technical solution, the target historical travel time with the highest probability of being later than the power-off time among the historical travel times corresponding to the target historical travel segment is determined. Then, based on the historical travel probability of this target historical travel segment and the historical travel time probability corresponding to the target historical travel time, the predicted travel probability for the target historical travel segment is obtained. The target historical travel segment with the highest predicted travel probability is taken as the predicted user's vehicle usage segment, and the target historical travel time corresponding to the predicted user's vehicle usage segment is taken as the predicted user's vehicle usage time. In this way, by comprehensively considering the historical travel time probability and the historical travel probability, the predicted user's vehicle usage segment most likely to be traveled on and the predicted user's vehicle usage time most likely to be traveled on this predicted user's vehicle usage segment are determined, thus obtaining the vehicle power-on time. This makes the vehicle power-on time prediction more accurate and better meets user needs.

[0010] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time. The driving information of the target vehicle in multiple historical driving segments is obtained in the following way: determining the initial historical driving position and the corresponding initial historical driving time of the target vehicle; performing clustering processing on the initial historical driving position and the initial historical driving time based on a clustering algorithm to obtain the historical driving position and historical driving time of the target vehicle; and obtaining the driving information of the target vehicle in multiple historical driving segments based on the historical driving position and the corresponding historical driving time.

[0011] By employing the above technical solution, clustering algorithms are used to cluster the initial historical vehicle usage locations and times to obtain the target vehicle's historical usage locations and times. Then, based on the historical usage locations and corresponding times, the driving information of the target vehicle in multiple historical road segments is obtained. This makes the historical usage information more standardized, enabling better and more accurate prediction of vehicle power-on times.

[0012] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time, which determines the historical travel probability of a target vehicle on a target historical driving segment based on the current location and historical vehicle usage information, including: determining the vehicle location transfer probability matrix of the target vehicle based on the historical vehicle usage information; and determining the historical travel probability of the target vehicle on the target historical driving segment from the vehicle location transfer probability matrix based on the current location.

[0013] By employing the above technical solution, a vehicle location transfer probability matrix is ​​determined based on historical vehicle usage information. Then, when predicting vehicle power-on time, the historical travel probability of the target vehicle on the target historical travel segment can be obtained by querying the vehicle location transfer probability matrix. This makes vehicle power-on time prediction much faster.

[0014] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time, which determines the vehicle location transfer probability matrix of the target vehicle based on historical vehicle usage information, including: determining the historical usage frequency corresponding to the historical driving segment based on the historical driving information; determining the historical travel probability of the target vehicle in the historical driving segment based on the historical usage frequency of the target vehicle in the historical driving segment; and obtaining the vehicle location transfer probability matrix based on the historical travel probability of the target vehicle in the historical driving segment.

[0015] By employing the above technical solution, the historical usage frequency corresponding to all historical driving segments is determined based on historical vehicle usage information. This yields the historical travel probability for each historical driving segment, generating a vehicle location transfer probability matrix. Consequently, when predicting vehicle power-on time, the historical travel probability of the target vehicle on the target historical driving segment can be obtained by querying the vehicle location transfer probability matrix. This makes vehicle power-on time prediction much faster.

[0016] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time. The historical vehicle usage location includes power-on location and power-off location. The method determines the historical travel probability of the target vehicle in the historical driving segment based on the historical usage frequency of the target vehicle in the historical driving segment. This includes: obtaining the historical travel probability of the target vehicle in the historical driving segment by the ratio of the historical usage frequency of the target vehicle in the historical driving segment corresponding to the first power-on location as the starting point and the first power-off location as the ending point to the historical usage frequency of the target vehicle in all historical driving segments corresponding to the first power-on location as the starting point and all power-off locations as the ending points.

[0017] Using the above technical solution, the historical travel probability of the target vehicle on historical travel segments is obtained by comparing the historical usage frequency of the historical travel segment corresponding to the first power-on position as the starting point and the first power-off position as the ending point with the historical usage frequency of all historical travel segments corresponding to the first power-on position as the starting point and all power-off positions as the ending points. This allows us to obtain the historical travel probability of each historical travel segment from the vehicle's historical usage information, facilitating subsequent prediction of user travel segments and user travel time.

[0018] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time. The historical vehicle usage time includes power-on time and power-off time. Based on the historical vehicle usage location and the corresponding historical vehicle usage time, the driving information of multiple historical driving segments of the target vehicle is obtained, including: if it is determined that the time interval between the power-off time of the target vehicle traveling from the first power-on location to the first power-off location and the power-on time of traveling from the first power-off location to the second power-off location is less than a preset first time interval threshold, then the driving information of the historical driving segment of the target vehicle is determined as follows: power-on location is the first power-on location, power-off location is the second power-off location, power-on time is the power-on time of the target vehicle at the first power-on location, and power-off time is the power-off time of the target vehicle at the second power-off location.

[0019] By adopting the above technical solution, the historical driving segments are integrated and processed based on the time interval between the power-off time of the target vehicle traveling from the first power-on position to the first power-off position and the power-on time of traveling from the first power-off position to the second power-off position, so as to make the prediction of user driving segments and user driving time more accurate.

[0020] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time, which determines the number of historical vehicle uses corresponding to a historical driving segment based on historical vehicle use information, including: if it is determined that the time interval between the power-on time and the power-off time of the target vehicle traveling from the first power-on position to the first power-off position in the historical driving segment is less than a preset second time interval threshold, and the average speed of the target vehicle is less than a preset speed threshold, then the number of historical vehicle uses in the historical driving segment is determined to be zero.

[0021] By adopting the above technical solution, the historical usage frequency of the target vehicle in the historical driving segment is determined based on the time interval between the power-on time and the power-off time when the target vehicle travels from the first power-on position to the first power-off position in the historical driving segment, as well as the average speed of the target vehicle. This eliminates the interference of invalid data, making the determination of the historical usage frequency more accurate, and thus making the prediction of user driving segment and user driving time more accurate.

[0022] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time, which determines the vehicle position transition probability matrix of the target vehicle based on historical vehicle usage information, including: obtaining the vehicle position transition probability matrix based on the historical vehicle usage information when it is determined that the historical vehicle usage information satisfies the corresponding conditions of the Markov chain model.

[0023] By adopting the above technical solution, when it is determined that the historical vehicle usage information meets the corresponding conditions of the Markov chain model, the vehicle location transition probability matrix is ​​obtained based on the historical vehicle usage information. This makes it possible to obtain the vehicle location transition probability matrix based on the historical vehicle usage information that meets the corresponding conditions of the Markov chain model, which facilitates subsequent historical travel probability queries.

[0024] According to another specific implementation of this application, the method for predicting vehicle power-on time disclosed in this implementation determines the historical travel time of the target vehicle on the target historical driving segment based on the current location and historical vehicle usage information, and determines the historical travel time probability corresponding to the historical travel time. The method includes: determining the historical travel time and historical travel time probability of the target vehicle on the historical driving segment based on the historical vehicle usage information; and determining the historical travel time and historical travel time probability corresponding to the historical travel time of the target vehicle on the target historical driving segment from the historical travel time and historical travel time probability of the historical driving segment based on the current location of the target vehicle.

[0025] By adopting the above technical solution, the historical travel time and historical travel time probability of the target vehicle in all historical driving segments are first determined. Then, when determining the historical travel time and historical travel time probability, the historical travel time and historical travel time probability of the corresponding target historical driving segment can be queried, which facilitates the query of the historical travel time and historical travel time probability of the target historical driving segment and speeds up the prediction efficiency of vehicle power-on time.

[0026] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time. Historical vehicle usage time includes power-on time. Based on historical vehicle usage information, the method determines the historical travel time and historical travel time probability of the target vehicle on historical driving segments. This includes: determining all historical travel times of the target vehicle on historical driving segments based on the power-on time, and the historical usage counts corresponding to all historical travel times; determining the historical usage counts of the target vehicle at historical travel times; and determining the historical travel time probability of the target vehicle on historical driving segments at historical travel times based on the ratio of the historical usage counts of the target vehicle at historical travel times to the historical usage counts corresponding to all historical travel times.

[0027] By adopting the above technical solution, the historical travel time probability corresponding to each historical travel time in each historical travel segment is determined. Based on the historical travel time probability, the predicted travel probability of the target historical travel segment can be determined. Furthermore, based on the predicted travel probability, the predicted user's vehicle usage segment and predicted user's vehicle usage time can be obtained, making the determination of the predicted user's vehicle usage segment and predicted user's vehicle usage time more accurate.

[0028] According to another specific implementation of this application, the implementation of this application discloses a method for predicting vehicle power-on time, which determines the historical travel time probability corresponding to historical travel time, including: determining the historical travel time probability corresponding to each historical travel time of the target vehicle on historical driving segments based on a Gaussian mixture model.

[0029] By adopting the above technical solution, the probability of historical travel time corresponding to each historical travel time of the target vehicle on the historical driving segment is determined based on the Gaussian mixture model, making the calculation of historical travel time probability faster and simpler.

[0030] According to another specific implementation of this application, the implementation of this application discloses a vehicle power-on time prediction method, which determines the historical travel probability and historical travel time of the target vehicle in the target historical driving segment based on the current location and historical vehicle usage information, and determines the historical travel time probability corresponding to the historical travel time, including: when it is determined from the current location and historical vehicle usage information that the current location is within a preset range of one of the power-on locations of the historical vehicle usage location, determining the historical travel probability and historical travel time of the target vehicle in the target historical driving segment based on the historical vehicle usage information, and determining the historical travel time probability corresponding to the historical travel time.

[0031] By adopting the above technical solution, vehicle power-on time prediction can be performed only when the current location is within a preset range of one of the power-on locations included in the historical vehicle usage information, making the vehicle power-on time predicted based on historical vehicle usage information more accurate.

[0032] According to another specific implementation of this application, the vehicle power-on time prediction method disclosed in this application uses the target historical driving segment with the highest predicted travel probability as the predicted user driving segment, including: when it is determined that the highest predicted travel probability is greater than a preset travel probability threshold, the target historical driving segment with the highest predicted travel probability is used as the predicted user driving segment.

[0033] By adopting the above technical solution, the target historical travel segment with the highest predicted travel probability is used as the predicted user's travel segment. In this way, the target travel segment, historical travel probability, and historical travel time probability are fully considered, ensuring that the predicted user's travel segment is the most likely route for the user's next trip, thus meeting user needs and improving user experience.

[0034] Secondly, the implementation of this application also discloses a vehicle management method, which includes: determining the predicted power-on time of a target vehicle in a parked and power-off state, wherein the predicted power-on time is obtained based on the vehicle power-on time prediction method provided by any implementation of the first aspect; and powering on the target vehicle at the predicted power-on time.

[0035] By adopting the above technical solution, the vehicle can be powered on in advance through accurate power-on time prediction, thus improving the overall user experience.

[0036] According to another specific implementation of this application, the vehicle management method disclosed in this implementation further includes: determining the predicted destination of a target vehicle in a parked and powered-off state, wherein the predicted destination is determined based on the predicted user vehicle segment obtained from the vehicle power-on time prediction method provided in any implementation of the first aspect; and determining the driving route of the target vehicle based on the predicted destination.

[0037] By adopting the above technical solution, the driving route of the target vehicle can be determined based on the predicted destination, which enables the planning of the driving route for the user's next trip and provides the user with a better driving experience.

[0038] Thirdly, this application provides an electronic device, including: a processor and a memory communicatively connected to the processor; the memory stores a computer program; the processor executes the computer program stored in the memory to enable the electronic device to implement the vehicle power-on time prediction method provided by any implementation of the first aspect, and / or implement the vehicle management method provided by any implementation of the second aspect.

[0039] Fourthly, the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, is used to implement the vehicle power-on time prediction method provided by any implementation of the first aspect, and / or to implement the vehicle management method provided by any implementation of the second aspect.

[0040] Fifthly, an implementation of this application provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle power-on time prediction method provided by any implementation of the first aspect, and / or implements the vehicle management method provided by any implementation of the second aspect.

[0041] It is understood that the beneficial effects of the third to fifth aspects mentioned above can also be found in the relevant descriptions in the first and / or second aspects mentioned above, and will not be repeated here. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating a method for predicting vehicle power-on time provided in an embodiment of this application;

[0043] Figure 2 This is a schematic diagram of the process for obtaining driving information of historical driving segments provided in an embodiment of this application;

[0044] Figure 3 This is a schematic diagram illustrating the principle of obtaining the vehicle position transition probability matrix provided in an embodiment of this application;

[0045] Figure 4 This is a flowchart illustrating the process of determining historical travel probabilities provided in an embodiment of this application;

[0046] Figure 5 This is a flowchart illustrating the process of determining the vehicle location transition probability matrix provided in an embodiment of this application;

[0047] Figure 6 This is a flowchart illustrating the process of determining historical travel time and the probability of historical travel time provided in an embodiment of this application;

[0048] Figure 7 This is a schematic diagram of the process for determining the predicted user's vehicle usage route and the predicted user's vehicle usage time, provided in an embodiment of this application.

[0049] Figure 8 This is a schematic flowchart of a vehicle management method provided in an embodiment of this application;

[0050] Figure 9 This is a flowchart illustrating another vehicle management method provided in an embodiment of this application;

[0051] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0052] Specific implementation method

[0053] Currently, for vehicles in a parked, off-power state, the vehicle's power-on time can be predicted by forecasting the user's next usage time. Automatically powering on and starting the vehicle at the predicted time allows it to better meet user needs. Therefore, predicting vehicle power-on time is crucial for enabling personalized services and improving user experience.

[0054] However, current vehicle prediction methods mostly focus on predicting the vehicle's destination, that is, predicting the user's destination based on the user's power-on time and the vehicle's location at the time of power-on. This method does not predict the power-on time.

[0055] Furthermore, even if a solution exists to predict vehicle power-on time, it typically relies on statistical analysis of the frequency of historical user usage, using the most frequent usage times as the predicted time. However, actual user usage time is influenced by numerous factors, such as varying travel destinations, which in turn affect the predicted power-on time. This method, which only considers historical usage times, results in low accuracy. Therefore, the predicted power-on time obtained using this method is not highly accurate. Powering on and starting the vehicle based on this predicted time may not reflect the user's actual travel time, negatively impacting the user experience.

[0056] Based on this, this application provides a vehicle power-on time prediction method and a vehicle management method. Based on the current location and historical vehicle usage information, it determines the probability of a target vehicle's trip on a target historical driving segment, its historical trip time, and its historical trip time probability. Then, based on the power-off time, historical trip probability, historical trip time, and historical trip time probability, it determines the predicted user driving segment that the user is most likely to need to travel on and the corresponding predicted user driving time for that segment, thus obtaining the predicted power-on time for traveling that segment. In this way, the next travel segment and the vehicle's power-on time for that segment are predicted. The predicted travel segment and power-on time are the most needed travel segment and travel time for the user. Furthermore, the travel time prediction can be based on the travel segment; if the predicted travel segment is different, the predicted power-on time for that travel segment will be different, making the power-on time prediction more accurate and improving the user experience.

[0057] Next, the vehicle power-on time prediction method provided in this application will be described in detail.

[0058] In the implementation method of this application, such as Figure 1 As shown, the vehicle power-on time prediction method specifically includes the following steps:

[0059] S100 determines the power-off time, current location, and historical vehicle usage information of the target vehicle. The historical vehicle usage information includes the driving information of the target vehicle in multiple historical driving segments.

[0060] S200 determines the historical travel probability and historical travel time of the target vehicle on the target historical driving segment based on the current location and historical vehicle usage information, and determines the historical travel time probability corresponding to the historical travel time.

[0061] Among them, the target historical driving segment is the historical driving segment whose starting point is the current location, which is included in the historical driving information.

[0062] S300 determines the predicted user's vehicle usage segment and the corresponding predicted user usage time based on the power-off time, historical travel probability, historical travel time, and historical travel time probability.

[0063] S400 obtains the predicted power-on time of the target vehicle based on the predicted user vehicle usage time.

[0064] The vehicle power-on time prediction method provided in this application determines the historical travel probability and historical travel time of the target vehicle on a target historical travel segment with the starting point (i.e., the originating location) as the current location based on historical vehicle usage information, and determines the historical travel time probability corresponding to the historical travel time. This allows for the determination of the most likely user travel segment and the most likely user travel time for that segment based on the power-off time, historical travel probability, historical travel time, and historical travel time probability, thus obtaining the vehicle power-on time. In this way, the method predicts the next travel segment and the travel time from the current location to that segment based on historical vehicle usage information, ensuring that the predicted travel segment and power-on time are the travel segments and travel times more needed by the user. Furthermore, the travel time prediction is based on the travel segment; different predicted travel segments will result in different predicted power-on times, making the power-on time prediction more accurate and improving the user experience.

[0065] In one implementation of this application, the vehicle power-on time prediction method provided by this application can be applied to the cloud, and the vehicle power-on time prediction method is performed by the cloud.

[0066] Specifically, in step S100, determining the power-off time, current location, and historical vehicle usage information of the target vehicle includes obtaining the vehicle identification number (VIN), power-off time T, and GPS signal L at the time of power-off (as an example of the current location) of the target vehicle after detecting that the vehicle is powered off, and determining the historical vehicle usage information of the target vehicle through the VIN, which includes the vehicle's driving information in multiple historical driving segments.

[0067] The vehicle's driving information across multiple historical road segments includes, for example, the target vehicle's power-on location A, power-off location B, and power-on time T on road segment AB. AB1 Power-off time T AB1 The target vehicle's power-on position A, power-off position B, and power-on time T on the AB driving segment. AB2 Power-off time T AB2 ′,……,Target vehicle's power-on position A, power-off position B, and power-on time T on the AB driving segment ABn Power-off time T ABnThe target vehicle's power-on position B, power-off position A, and power-on time T on the BA driving section. BA1 Power-off time T BA1 ′,……,Target vehicle's power-on position B, power-off position A, and power-on time T in the BA driving section. BAn Power-off time T BAn The target vehicle's power-on position A, power-off position C, and power-on time T on the AC driving segment. AC1 Power-off time T AC1 ′,……,Target vehicle's power-on position A, power-off position C, and power-on time T in the AC driving segment. ACn Power-off time T ACn Of course, the driving route is determined based on the target vehicle's historical location and historical time of use; this is just an example.

[0068] In one implementation of this application, such as Figure 2 As shown, the target vehicle's driving information on multiple historical road segments was obtained in the following way:

[0069] S110, determine the initial historical vehicle location and corresponding initial historical vehicle usage time of the target vehicle.

[0070] For example, such as Figure 3 As shown, the process obtains the processed driving cycle record for a certain period of time (e.g., three months or six months of history) of the target, and determines that the driving cycle record includes the power-on GPS (as an example of the power-on location), power-off GPS (as an example of the power-off location), power-on time, and power-off time.

[0071] It should be noted that the GPS that is powered off is the same GPS that will be powered on for the next trip.

[0072] S120, based on the clustering algorithm, performs clustering processing on the initial historical vehicle usage location and initial historical vehicle usage time to obtain the historical vehicle usage location and historical vehicle usage time of the target vehicle.

[0073] In the implementation method of this application, the historical vehicle usage location includes the power-on location and the power-off location, and the historical vehicle usage time includes the power-on time and the power-off time.

[0074] For example, clustering algorithms are used to cluster the powered-on GPS devices, resulting in multiple powered-on GPS devices labeled as the same category. For instance, points A and B are the clustered powered-on locations, while points C, D, E, and F are the powered-on GPS devices of outliers, each representing a different powered-on location (as an example of historical vehicle usage locations).

[0075] Furthermore, in this application, a clustering algorithm is used to cluster the power-on time and power-off time to obtain power-on time and power-off time of multiple categories labeled as the same category (as an example of historical vehicle usage time).

[0076] S130 obtains the target vehicle's driving information on multiple historical driving segments based on historical vehicle location and corresponding historical vehicle time.

[0077] For example, based on the clustered historical power consumption locations (e.g., power-on and power-off locations (i.e., power-on and power-off GPS latitude and longitude data)), GPS mapping is performed one by one to obtain multiple driving cycle records (e.g., AB, BC, CA, ..., ED, DA, AF, FB) containing corresponding location categories (as an example of historical driving segments). This leads to the driving information of the target vehicle in each historical driving segment, including the target vehicle's power-on location (e.g., A, B, C, ..., E, D, A, F) for that historical driving segment, the power-on time corresponding to the power-on location, the power-off location (e.g., B, C, A, ..., D, A, F, B) corresponding to the power-off location, and the power-off time corresponding to the power-off location.

[0078] In another implementation of this application, the driving information of multiple historical driving segments of the target vehicle is obtained based on the historical vehicle location and the corresponding historical vehicle time, including: if it is determined that the time interval between the power-off time of the target vehicle from the first power-on location to the first power-off location and the power-on time from the first power-off location to the second power-off location is less than a preset first time interval threshold, then the driving information of the historical driving segment of the target vehicle is determined as follows: power-on location is the first power-on location, power-off location is the second power-off location, power-on time is the power-on time of the target vehicle at the first power-on location, and power-off time is the power-off time of the target vehicle at the second power-off location.

[0079] For example, continuous driving cycle data (i.e. driving cycle record) of a vehicle over a period of time is acquired and corresponding data processing is performed. If the time interval between the power-off time of the target vehicle traveling from the first power-on position to the first power-off position and the power-on time of traveling from the first power-off position to the second power-off position is less than a preset first time interval threshold (e.g., 15 minutes), then the two segments of driving cycle data are determined to be a historical driving segment.

[0080] In other words, if the downtime of two adjacent driving cycles is less than 15 minutes, these two driving cycles are merged into one driving cycle, resulting in a historical driving segment with the power-on position as the first power-on position, the power-off position as the second power-off position, the power-on time as the power-on time of the target vehicle at the first power-on position, and the power-off time as the power-off time of the target vehicle at the second power-off position. Of course, the preset first time interval threshold can be set according to actual needs; this is only an example.

[0081] Furthermore, in step S200, determining the historical travel probability and historical usage time of the target vehicle on the target historical driving segment based on the current location and historical usage information, and determining the historical travel time probability corresponding to the historical travel time, includes: when the current location is determined to be within a preset range of one of the power-on locations of the historical usage locations based on the current location and historical usage information, determining the historical travel probability and historical usage time of the target vehicle on the target historical driving segment based on the historical usage information, and determining the historical travel time probability corresponding to the historical travel time.

[0082] For example, the distance between the GPS signal L (i.e., the current location) of the target vehicle when it is powered off and different location categories in the historical vehicle usage locations is calculated. The minimum distance value is obtained (assuming that the distance between L and point A is the smallest). If the distance is within a preset range (e.g., 1 meter, 5 meters, etc.) (i.e., less than or equal to the preset distance threshold), it is determined that the current location belongs to the historical vehicle usage location of location category C, and the subsequent steps can continue. If it is not within the preset range (i.e., greater than the preset distance threshold), it is considered that the current location is a location that does not exist in the historical vehicle usage locations, and therefore there is no matching location category. In this case, the prediction of the user's driving route segment is not performed, and the vehicle power-on time can be predicted directly based on the historical vehicle usage time. Of course, the preset distance threshold (i.e., the preset range) can be set according to actual needs; this is only an example.

[0083] Furthermore, in one implementation of this application, such as Figure 4 As shown, determining the historical travel probability of a target vehicle on a target historical route segment based on the current location and historical vehicle usage information includes the following steps:

[0084] S210, determine the vehicle location transfer probability matrix of the target vehicle based on historical vehicle usage information.

[0085] For example, in the implementation of this application, self-learning can be performed based on historical vehicle usage information to obtain a vehicle position transfer probability matrix for each historical driving segment that includes different location categories (i.e., from different power-on locations to different power-off locations).

[0086] Furthermore, such as Figure 5 As shown, in one implementation of this application, determining the vehicle location transition probability matrix of the target vehicle based on historical vehicle usage information includes:

[0087] S211, determine the number of times a vehicle was used for a historical route based on historical vehicle usage information.

[0088] For example, the historical vehicle usage counts for each historical driving segment are obtained based on the historical location information obtained after clustering. For instance, the historical vehicle usage counts for AA are 0, AB are 50, AC are 10, AD are 30, AE are 0, AF are 10, and so on, thus determining the historical vehicle usage counts for all power-on locations to all power-off locations.

[0089] In another implementation of this application, determining the number of historical vehicle uses corresponding to the historical driving segment based on historical vehicle use information includes: if it is determined that the time interval between the power-on time and the power-off time of the target vehicle traveling from the first power-on position to the first power-off position in the historical driving segment is less than a preset second time interval threshold, and the average speed of the target vehicle is less than a preset speed threshold, then the number of historical vehicle uses in the historical driving segment is determined to be zero.

[0090] For example, in order to ensure the accuracy of the probability after self-learning, the implementation of this application will also perform data processing on the driving cycle data. For example, if the time interval between the power-on time and the power-off time from the first power-on position to the first power-off position is less than a preset second time interval threshold (e.g., 15 minutes), and the average speed of the target vehicle is less than a preset speed threshold (e.g., 3 km / h), then it is determined that the number of times the vehicle was used in this historical driving segment is zero.

[0091] That is, if the driving duration of a certain driving cycle (the duration of the first power-off time minus the first power-on time) is less than 15 minutes and the average driving speed of the vehicle is less than 3 km / h, then the driving cycle is deleted and not included in the calculation. Of course, the preset second time interval threshold and the preset vehicle speed threshold can be set according to actual needs; this is only an example.

[0092] S212, determine the historical travel probability of the target vehicle on the historical travel segment based on the number of times the target vehicle has used the vehicle on the historical travel segment.

[0093] For example, in one implementation of this application, determining the historical travel probability of a target vehicle on a historical driving segment based on the number of times the target vehicle has used the vehicle on the historical driving segment includes: obtaining the historical travel probability of the target vehicle on the historical driving segment by comparing the number of times the target vehicle has used the vehicle on the historical driving segment corresponding to the first power-on position as the starting point and the first power-off position as the ending point with the number of times the vehicle has used the vehicle on all the historical driving segments corresponding to the first power-on position as the starting point and all power-off positions as the ending points with the vehicle on the historical driving segment.

[0094] For example, to calculate the historical travel probability of AB (as an example of a historical driving route), we obtain the driving cycle records of the powered-on GPS category (i.e., powered-on location) = A. For example, taking A (as an example of the first powered-on location) as the starting point of the journey and all powered-off locations as the end points of the journey, there are a total of n historical driving times for the historical driving routes. In the example above, there are 100 such records.

[0095] Then obtain the driving cycle records for the power-on GPS category = A and the power-off GPS category (i.e., power-off location) = B (as an example of the first power-off location). For example, the historical driving number of the historical driving segment with A as the starting point and B as the ending point is m records, which is 50 records in the example above.

[0096] The historical travel probability of the historical travel segment with the first power-on position as the starting point and the first power-off position as the ending point is m / n, which is 50 / 100 = 0.5.

[0097] Thus, the total historical usage counts from all first power-on locations to all first power-off locations are calculated (e.g., the total historical usage counts of AA, B, C, D, E, F; the total historical usage counts of BA, B, C, D, E, F; the total historical usage counts of CA, B, C, D, E, F; ..., FA, B, C, D, E, F). Then, the historical usage counts from the first power-on location to each first power-off location are calculated (e.g., the historical usage counts of AA, AB, ..., AF, BA, ..., BF, ..., FA, ..., FF). Finally, the historical travel probabilities from the first power-on location to each first power-off location are obtained (e.g., the historical travel probability of AA relative to AA, B, C, D, E, F; ..., the historical travel probability of BF relative to BA, B, C, D, E, F; ..., the historical travel probability of FF relative to FA, B, C, D, E, F).

[0098] S213, obtain the vehicle location transfer probability matrix based on the historical travel probability of the target vehicle on the historical driving route.

[0099] For example, the historical travel probabilities from each first power-on location to each first power-off location are summarized as follows: Figure 3 The table shown yields the GPS state transition matrix (as an example of the vehicle location transition probability matrix). Thus, by pre-learning to obtain the vehicle location transition probability matrix, when predicting vehicle power-on time, the current location is determined, and the historical travel probability of the target historical driving segment corresponding to the power-on location as the current location and the power-off location as the corresponding power-off location is retrieved from the vehicle location transition probability matrix, thereby accelerating the efficiency of vehicle power-on prediction.

[0100] For example, Figure 3 As shown, the historical travel probabilities are as follows: AA = 0, AB = 0.5, AC = 0.1, AD = 0.3, AE = 0, AF = 0.1, BA = 0.6, BB = 0, BC = 0.2, BD = 0, BE = 0.1, BF = 0.1, CA = 0.8, CB = 0, CC = 0, CD = 0.1, CE = 0, CF = 0.1, and D = 0. The vehicle location transfer probability matrix is ​​as follows: A has a historical travel probability of 0.7, DB has a historical travel probability of 0.1, DC has a historical travel probability of 0.1, DD has a historical travel probability of 0.1, DE has a historical travel probability of 0, DF has a historical travel probability of 0, EA has a historical travel probability of 0, EB has a historical travel probability of 0, EC has a historical travel probability of 0, ED has a historical travel probability of 0.9, EE has a historical travel probability of 0.1, EF has a historical travel probability of 0, FA has a historical travel probability of 0.1, FB has a historical travel probability of 0.6, FC has a historical travel probability of 0.3, FD has a historical travel probability of 0, FE has a historical travel probability of 0, and FF has a historical travel probability of 0.

[0101] Furthermore, in another implementation of this application, the state transition of the vehicle's GPS is actually calculated based on a Markov chain model. Therefore, in this application, determining the vehicle position transition probability matrix of the target vehicle based on historical vehicle usage information includes: obtaining the vehicle position transition probability matrix based on the historical vehicle usage information, provided that the historical vehicle usage information satisfies the corresponding conditions of the Markov chain model.

[0102] It should be noted that Markov chains can be defined by transition matrices and transition graphs. Therefore, if the historical vehicle usage information satisfies the calculation principle of the transition matrix of the Markov chain model (as an example of the corresponding conditions), the historical travel probability of the target vehicle's historical driving segments can be calculated based on the calculation principle of the Markov chain model and the historical vehicle usage information, thereby obtaining the vehicle location transition probability matrix.

[0103] S220, determine the historical travel probability of the target vehicle in the target historical travel segment from the vehicle location transfer probability matrix based on the current location.

[0104] For example, based on the current location, the historical travel probability of the target historical driving segment corresponding to the current location (power-on location) and the power-off location (power-off location) can be queried from the vehicle location transfer probability matrix.

[0105] For example, if the current location is determined to be A, then the historical travel probabilities of AA, AB, AC, AD, AE, and AF can be queried from the vehicle location transfer probability matrix.

[0106] Furthermore, in one implementation of this application, in order to ensure the self-learning accuracy of the vehicle position transfer probability matrix, the travel mode (i.e., historical travel probability) of the target vehicle is only trained when the driving cycle data (i.e., historical driving segments) of the target vehicle is greater than or equal to a preset threshold.

[0107] Thus, in this application, the vehicle location transfer probability matrix is ​​obtained in advance. When it is necessary to predict the vehicle power-on time, the vehicle location transfer probability matrix corresponding to the vehicle VIN number can be directly obtained, and then the historical travel probability of the target historical driving segment can be queried from the vehicle location transfer probability matrix.

[0108] Furthermore, in the implementation of this application, after calculating the historical travel probability of each historical travel segment, it is also necessary to calculate the historical travel time and historical travel time probability of each historical travel segment.

[0109] In one implementation of this application, such as Figure 4 As shown, the historical travel time of the target vehicle on the target historical driving segment is determined based on the current location and historical vehicle usage information, and the historical travel time probability corresponding to the historical travel time is determined, including the following steps.

[0110] S230 determines the historical travel time and probability of the target vehicle on the historical driving route based on historical vehicle usage information.

[0111] In one implementation of this application, such as Figure 6 As shown, determining the historical travel time and probability of the target vehicle on historical driving routes based on historical vehicle usage information includes the following steps.

[0112] S231, determine all historical travel times of the target vehicle on historical driving segments based on the power-on time, as well as the number of historical vehicle uses corresponding to all historical travel times.

[0113] For example, taking the calculation of historical travel time and historical travel time probability of AB (as an example of historical driving segment) as an example, we obtain all power-on times (as an example of historical travel time) of the driving cycle record set I with power-on GPS category (i.e. power-on location) = A and power-off GPS category (i.e. power-off location) = B (as an example of the first power-off location), and the historical number of vehicle use based on each power-on time, i.e., the aforementioned m times (e.g., 50 times).

[0114] S232, determine the number of times the target vehicle was used in the historical travel time.

[0115] For example, suppose that the historical travel times for the 50 historical travel segments corresponding to AB are 8:00, 12:00, 13:00 and 15:00, and that the number of historical trips at 8:00 is 12, at 12:00 is 15, at 13:00 is 20 and at 15:00 is 3.

[0116] S233, determine the probability of the target vehicle's historical travel time on the historical route segment at the historical travel time based on the ratio of the target vehicle's historical usage frequency at historical travel times to the historical usage frequency corresponding to all historical travel times.

[0117] For example, the probability of a target vehicle's historical travel time on a given historical travel segment is determined by the ratio of the number of historical trips at each historical travel time to the total number of historical trips on that historical travel segment. For instance, if a target vehicle used the vehicle 12 times at 8:00 AM on historical travel segment AB, the probability of its historical travel time on segment AB at 8:00 AM is 0.24; if it used the vehicle 15 times at 12:00 PM, the probability is 0.3; if it used the vehicle 20 times at 1:00 PM, the probability is 0.4; and if it used the vehicle 3 times at 3:00 PM, the probability is 0.06.

[0118] Thus, the historical travel time of the target vehicle in each historical travel segment (e.g., AA, AB, ..., AF, BA, ..., BF, ..., FA, ..., FF) is calculated using the above calculation method. Then, based on the historical number of trips at each historical travel time in each historical travel segment (e.g., the historical number of trips at historical travel time 1 of AB) and the total number of historical trips in the corresponding historical travel segment (e.g., the historical number of trips of AB), the probability of the historical travel time of the target vehicle in the corresponding historical travel segment at each historical travel time is obtained.

[0119] S240, based on the current location of the target vehicle, determines the historical travel time probability of the target vehicle in the target historical travel segment from the historical travel time and historical travel time probability of the historical travel segment.

[0120] For example, based on the current location, the historical travel times and corresponding historical travel time probabilities of the target historical driving segments are determined from the historical travel times of the historical driving segments, with the current location being the power-on location and the power-off location being the power-off location.

[0121] For example, if the current location is determined to be A, then query the vehicle location transfer probability matrix to find all historical travel times and historical travel time probabilities for each historical travel time of AA, AB, AC, AD, AE, and AF.

[0122] In another implementation of this application, before calculating the probability of historical travel times, as mentioned above, the initial historical travel times of each historical travel segment set I are clustered using a clustering algorithm to obtain the processed historical travel times. That is, multiple historical travel times within a preset time interval are clustered into the same historical travel time.

[0123] For example, if the power-on time (i.e., the historical travel time) is determined to be 12:55, 13:05, 13:10, and 13:12, then after clustering, the historical travel time is obtained as 13:00. In this way, all initial historical travel times are clustered to obtain multiple categories K of historical travel times.

[0124] It should be noted that the clustering algorithm in the implementation of this application can be the k-medoids algorithm, the Clara algorithm, or other algorithms. Those skilled in the art can choose according to the data distribution or application scenario.

[0125] In another implementation of this application, the probability of historical travel time can also be calculated based on a Gaussian mixture model. Therefore, the vehicle power-on time prediction method provided by this application determines the probability of historical travel time corresponding to historical travel time, including: determining the probability of historical travel time corresponding to each historical travel time of the target vehicle on the historical driving segment based on a Gaussian mixture model.

[0126] For example, for each historical travel time of multiple categories K corresponding to a historical travel segment, K sub-models of a Gaussian mixture model (i.e., Gaussian mixture model) are established. Each sub-model corresponds to a historical travel time. Then, the historical number of trips, historical travel time and K of the historical travel segment are input. Based on the historical travel times in set I, the parameters of the Gaussian mixture model are solved using the EM algorithm (Expectation Maximization Algorithm). A Gaussian mixture model suitable for the historical travel segment is obtained. Then, based on the Gaussian mixture model, the historical travel time probability of each historical travel time of the historical travel segment is obtained.

[0127] Furthermore, in another implementation of this application, to ensure the accuracy of the historical travel time probability, self-learning of historical travel time and historical travel time probability is only performed if the number of historical trips on the historical travel segment is greater than or equal to a preset threshold. If the condition is not met, the predicted user's car usage time is directly determined based on the historical travel time that occurs most frequently. Of course, the preset threshold can be set based on actual needs, and is not specifically limited here.

[0128] Thus, in this application, the historical travel time probability of each historical driving segment is obtained in advance, and when it is necessary to predict the vehicle power-on time, the historical travel time probability of the target historical driving segment can be directly queried.

[0129] Furthermore, in one implementation of this application, such as Figure 7 As shown, in step S300, based on the power-off time, historical travel probability, historical travel time, and historical travel time probability, the predicted user's vehicle usage segment is determined, as well as the predicted user's vehicle usage time corresponding to the predicted user's vehicle usage segment is determined, including:

[0130] S310, the historical travel time that is later than the power-off time and has the highest probability among historical travel times is taken as the target historical travel time.

[0131] S320, determine the predicted travel probability corresponding to the target historical travel segment based on the historical travel probability of the target historical travel segment and the historical travel time probability corresponding to the target historical travel time.

[0132] For example, based on the vehicle location transition probability matrix, the transition probability (i.e., the historical travel probability) of traveling from point A (as an example of the current location) to another power-off location i along the historical route can be obtained as P. AiHere, i represents different location categories (i.e., different power-off locations), and Ai (e.g., AA, AB, AC, AD, AE, AF, as examples of target historical travel segments) is obtained based on a Gaussian mixture model (i.e., the historical travel time probability corresponding to the historical travel time). Based on the power-off time T, the time t with the highest historical travel time probability for each target historical travel segment, where t>T, is obtained. i (As an example of the target historical travel time) and the probability P corresponding to that target historical travel segment. ci_time Then at t i The predicted travel probability r of the target historical travel segment from point A to point i i =P ci ×P ci_time .

[0133] For example, taking the target historical driving segment AB as an example, and the power-off time T as 12:00, as mentioned earlier, the 10 historical travel times for the target historical driving segment AB are 8:00, 12:00, 13:00, and 15:00. Furthermore, the probability of the target vehicle traveling on the historical driving segment AB at 8:00 is 0.24, at 12:00 is 0.3, at 13:00 is 0.4, and at 15:00 is 0.06.

[0134] The historical travel times with a travel time t>T are determined to be 13:00 and 15:00. The historical travel time with the highest probability between 13:00 and 15:00 has a probability of 0.4 corresponding to 13:00. Therefore, the target historical travel time t for the target historical travel segment AB is determined. i If the time is 13:00 and the historical travel probability of AB is 0.5, then the predicted travel probability of going from point A to point B after powering on at 13:00 is 0.5 * 0.4 = 0.2.

[0135] Furthermore, calculate the target historical travel time with the highest probability that the historical travel time is later than (i.e., greater than) the power-off time among all target historical travel segments corresponding to the starting point (i.e., the power-on position) and other travel ends (i.e., the power-off position).

[0136] For example, suppose that in AA, the historical travel time is later than the power-off time, and the target historical travel time with the highest probability of historical travel is 2 PM. The historical travel probability of 2 PM is 0.2, and the historical travel probability of AA is 0. Then the predicted travel probability of going to point A after powering on at 2 PM is 0.2 * 0 = 0. Suppose that in AC, the historical travel time is later than the power-off time, and the target historical travel time with the highest probability of historical travel is 5 PM. The historical travel probability of 5 PM is 0.5, and the historical travel probability of AC is 0.1. Then the predicted travel probability of going to point C after powering on at 5 PM is 0.5 * 0.1 = 0.05. Suppose that in AD, the historical travel time is later than the power-off time, and the target historical travel time with the highest probability of historical travel is 5 PM. The historical travel probability of 5 PM is 0.2, and the historical travel probability of AC is 0.1. If the probability is 0.8 and the historical travel probability of AD is 0.3, then the predicted travel probability of going from point A to point D after powering on at 5 PM is 0.8 * 0.3 = 0.24. If the historical travel time of AE is later than the power-off time, and the target historical travel time with the highest probability is 1 PM, with a historical travel time probability of 0.3 and a historical travel probability of AE of 0, then the predicted travel probability of going from point A to point B after powering on at 2 PM is 0.3 * 0 = 0. If the historical travel time of AF is later than the power-off time, and the target historical travel time with the highest probability is 4 PM, with a historical travel time probability of 0.4 and a historical travel probability of AF of 0.1, then the predicted travel probability of going from point A to point F after powering on at 4 PM is 0.1 * 0.4 = 0.04.

[0137] S330 uses the target historical driving segment with the highest predicted travel probability as the predicted user's vehicle usage segment, and uses the target historical travel time corresponding to the predicted user's vehicle usage segment as the predicted user's vehicle usage time.

[0138] For example, the target historical travel segment with the highest predicted travel probability is identified as AD (as an example of a predicted user's vehicle usage segment), and the target historical travel time corresponding to AD is 5 PM (as an example of a predicted user's vehicle usage time). Thus, based on the user's historical vehicle usage information, the predicted user's vehicle usage segment with the highest probability of travel (i.e., the highest predicted travel probability) is identified as AD, and the predicted user's vehicle usage time with the highest probability of travel to this predicted user's vehicle usage segment (i.e., the highest historical travel time probability) is identified as 5 PM. This allows the vehicle's power-on time to be determined based on the predicted user's vehicle usage segment, making the vehicle's power-on time prediction more accurate.

[0139] Furthermore, in one implementation of this application, the target historical driving segment with the highest predicted travel probability is used as the predicted user's vehicle segment, which includes: when it is determined that the highest predicted travel probability is greater than a preset travel probability threshold, the target historical driving segment with the highest predicted travel probability is used as the predicted user's vehicle segment.

[0140] For example, suppose point A t F The probability of powering on at time (e.g., 5 PM) and heading to location D is highest. If this value is greater than or equal to a preset threshold (e.g., 0.2), the target historical driving segment corresponding to AD is determined as the predicted user's driving segment. If it is less than this threshold, the predicted user's driving segment is not determined.

[0141] In step S400, if the predicted user's vehicle usage segment (e.g., AD) and the predicted user's vehicle usage time (e.g., 5 PM) are determined, then the predicted next power-on time for the vehicle is 5 PM (as an example of predicted power-on time).

[0142] In another implementation of this application, the predicted power-on time of the target vehicle can be obtained by predicting the user's vehicle usage time, or by determining a preset time period before the predicted user's vehicle usage time as the predicted power-on time of the target vehicle.

[0143] The preset time period can be 3 minutes, 5 minutes, 8 minutes, 10 minutes, 15 minutes, etc.

[0144] For example, if the predicted user's vehicle usage time is 5 PM, then the predicted power-on times are determined to be 4:57 PM, 4:55 PM, 4:52 PM, 4:50 PM, and 4:45 PM. In this way, the vehicle can be started in advance based on the predicted user's usage time to complete the corresponding vehicle preparation operations. As a result, when the user uses the vehicle at the predicted usage time, they can directly operate the vehicle without manually powering it on.

[0145] The vehicle power-on time prediction method provided in this application can help car owners plan their vehicle power-on time and improve user experience, whether for new energy vehicles or fuel vehicles. Furthermore, for new energy vehicles, accurate power-on time prediction can help car owners optimize their charging plans, avoid charging during peak hours, save energy costs, and optimize energy management. The vehicle power-on time prediction method provided in this application is a user travel model-based method. It utilizes historical user travel data (i.e., historical vehicle usage information) and employs techniques such as Gaussian mixture models, Markov chain models, and clustering algorithms to learn user travel patterns (i.e., predict user travel routes). Then, based on the current vehicle power-off GPS (i.e., current location) and power-off time, it predicts the next vehicle power-on time.

[0146] Specifically, the first step is to learn users' travel patterns.

[0147] The process begins with data preparation: acquiring historical vehicle usage information, such as continuous driving cycle data over a period of time (i.e., driving information for historical routes). Each driving cycle includes the vehicle's power-on time and the corresponding GPS information (i.e., power-on location) and power-off time and the corresponding GPS information (i.e., power-off location). Further, data processing is performed. As mentioned earlier, if the downtime of two adjacent driving cycles is less than 15 minutes, these two driving cycles are merged into one driving cycle; if the driving time of a certain driving cycle is less than 5 minutes and the average driving speed is less than 3 km / h, that driving cycle is deleted, and the number of historical vehicle uses is not recorded. Furthermore, when the number of historical routes for the vehicle is greater than or equal to a preset threshold, the vehicle's travel patterns are trained.

[0148] Next, model learning is performed: driving from the GPS at power-on (i.e., power-on location) to the GPS at power-off (i.e., power-off location) is called a GPS transfer. Based on driving cycle data over a historical period, the transfer probabilities between different GPS locations are statistically analyzed, and a transfer probability matrix (i.e., vehicle location transfer probability matrix) is constructed. Further, for each pair of GPS transfer relationships, its power-on time model is learned. Taking GPS1→GPS2 as an example: all driving cycle records satisfying "Power-on GPS = GPS1 and Power-off GPS = GPS2" are obtained from the aforementioned driving cycle dataset, and the power-on times (i.e., historical travel times) of these records are obtained, forming a power-on time set on_time_list. A density clustering algorithm (i.e., the aforementioned EM algorithm) is used to cluster on_time_list to obtain the number of categories K of the time set (i.e., the number of historical travel times). Inputting initial parameters (e.g., historical travel times) and K, a Gaussian mixture model is learned based on the time set on_time_list to obtain the historical travel time probability.

[0149] Next, the vehicle power-on time is predicted: obtain the GPS data (i.e., the current location) and power-off time of the vehicle's most recent power-off. Through the GPS transfer probability matrix (i.e., the vehicle location transfer probability matrix), the possible destination of the vehicle's next trip (i.e., the predicted user's driving route) and probability (i.e., the predicted travel probability) can be obtained. Combined with the vehicle power-off time, the time point with the highest probability of the next trip (i.e., the predicted user's driving time) is calculated, which is the time point of the next power-on (i.e., the predicted power-on time).

[0150] In other words, the vehicle power-on time prediction method provided in this application first obtains the vehicle's historical travel data (i.e., historical vehicle usage information), including: vehicle power-on time and the corresponding GPS information at power-on (i.e., power-on location), vehicle power-off time and the corresponding GPS information at power-off (i.e., power-off location). Then, it learns the user's travel patterns based on the historical travel data, including: learning the transition probability matrix between different GPS locations (i.e., the vehicle location transition probability matrix) and learning a power-on time model. Finally, it makes predictions based on the learned user travel patterns, including: predicting the vehicle's next travel destination based on the GPS at the time of power-off (i.e., predicting the user's travel route), and calculating the most probable next power-on time based on the power-off time (i.e., predicting the user's vehicle usage time).

[0151] Furthermore, this application also provides a vehicle management method, see [link to relevant documentation]. Figure 8 Specifically, it includes the following steps.

[0152] S10, determine the predicted power-on time of the target vehicle that is in the parked and power-off state.

[0153] For example, after obtaining the predicted power-on time using the aforementioned vehicle power-on time prediction method, the predicted power-on time of the target vehicle in the parked and power-off state is obtained.

[0154] S20, powering up the target vehicle at the predicted power-up time.

[0155] For example, the target vehicle can be powered on and ready for operation before the predicted power-on time. In this way, accurate power-on time prediction can help car owners perform pre-operational actions such as turning on the air conditioning and adjusting the seats, improving the overall user experience.

[0156] Furthermore, in the implementation of this application, after obtaining the predicted user's driving route and predicted user's driving time, a notification message can be generated based on the predicted user's driving route and predicted user's driving time (for example, it is predicted that you will depart for location D at 17:00, and the vehicle will start automatically at 17:00; please confirm whether you need to start it), and sent to the user terminal to confirm whether the user should power on the vehicle at the predicted user driving time. The user's response information based on the notification information is obtained. If the user confirms that the vehicle should be powered on at the predicted user driving time, the target vehicle is controlled to power on at the predicted vehicle power-on time.

[0157] Furthermore, in one implementation of this application, see [link to relevant documentation]. Figure 9 The vehicle management method also includes the following steps.

[0158] S30 determines the predicted destination of the target vehicle that is in a parked, power-off state.

[0159] For example, after obtaining the predicted user's driving route using the aforementioned vehicle power-on time prediction method, the destination of the predicted user's driving route is taken as the predicted destination.

[0160] S40 determines the route of the target vehicle based on the predicted destination.

[0161] For example, if the response information confirms that the vehicle will be powered on during the predicted user's vehicle usage time, navigation information will be generated by planning the target vehicle's route based on the power-off location and the predicted destination.

[0162] It should be noted that steps S10 and S30 can be executed simultaneously, or S10 can be executed first and then S30, or S30 can be executed first and then S10. Steps S20 and S30 can be executed simultaneously, or S20 can be executed first and then S30, or S30 can be executed first and then S20.

[0163] The vehicle power-on time prediction method and vehicle management method provided in this application can also be applied to electronic devices such as vehicles and remote terminals.

[0164] Please see Figure 10 , Figure 10 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this application. Figure 10 As shown, the electronic device may include: transceiver 121, processor 122, and memory 123.

[0165] Processor 122 executes computer execution instructions stored in memory, causing processor 122 to execute part of the technical solutions of the vehicle power-on time prediction method and / or vehicle management method in the above embodiments. Processor 122 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital data processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0166] The memory 123 is connected to the processor 122 via the system bus and completes communication between them. The memory 123 is used to store computer program instructions.

[0167] For example, and not as a limitation, memory 123 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 123 may include removable or non-removable (or fixed) media. Where appropriate, memory 123 may be internal or external to the integrated gateway device. In a particular embodiment, memory 123 is non-volatile solid-state memory. In a particular embodiment, memory 123 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0168] Transceiver 121 can be used to obtain the task to be run and its configuration information.

[0169] The system bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The system bus can be divided into address bus, data bus, control bus, etc. For ease of representation, only one thick line is used in the diagram, but this does not indicate that there is only one bus or one type of bus. Transceivers are used to enable communication between database access devices and other computers (e.g., clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) and may also include non-volatile memory.

[0170] This application also provides a chip for executing instructions, which is used to execute the technical solutions of the vehicle power-on time prediction method and / or vehicle management method in the above embodiments.

[0171] This application also provides a computer-readable storage medium storing computer instructions. When the computer instructions are executed on the processor of an electronic device, the processor of the electronic device executes the technical solutions of the vehicle power-on time prediction method and / or vehicle management method described in the above embodiments.

[0172] In some possible implementations, various aspects of the methods provided in this application may also be implemented as a program product, which includes program code. When the program product is run on the processor of an electronic device, the program code is used to cause the processor of the electronic device to perform the steps in the methods of the various exemplary implementations of this application described above. For example, the electronic device may perform the vehicle power-on time prediction method and / or vehicle management method described in the embodiments of this application.

[0173] The program product may take the form of any combination of one or more readable media. A readable medium may be a readable data medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0174] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When the at least one processor executes the computer program, it can implement the technical solutions of the vehicle power-on time prediction method and / or vehicle management method in the above embodiments.

[0175] This application is described with reference to flowchart illustrations and / or block diagrams of the methods, apparatus, and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable information processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable information processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0176] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable information processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0177] These computer program instructions may also be loaded onto a computer or other programmable information processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0178] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0179] It should be noted that, in addition to the specific embodiments described above, those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Although the description of this application is presented in conjunction with preferred embodiments, this does not mean that the features of this invention are limited to this implementation. On the contrary, the purpose of describing the invention in conjunction with the implementation is to cover other options or modifications that may be derived based on the claims of this application. To provide a thorough understanding of this application, many specific details are included in the above description, and this application may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of this application, some specific details will be omitted in the description. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0180] It should be noted that in this specification, similar reference numerals and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0181] Although this application has been illustrated and described with reference to certain preferred embodiments, those skilled in the art should understand that the above description is a further detailed explanation of the application in conjunction with specific implementations, and should not be construed as limiting the specific implementation of the application to these descriptions. Those skilled in the art can make various changes in form and detail, including some simple deductions or substitutions, without departing from the spirit and scope of this application.

Claims

1. A method for predicting vehicle power-on time, characterized in that, The method includes: Determine the power-off time, current location, and historical vehicle usage information of the target vehicle. The historical vehicle usage information includes the driving information of the target vehicle in multiple historical driving segments. The driving information is obtained based on the historical vehicle usage location and the corresponding historical vehicle usage time. The historical vehicle usage location includes the power-on location and the power-off location. The number of historical vehicle uses corresponding to the historical driving segment is determined based on the historical vehicle use information. The historical travel probability of the target vehicle on the historical travel segment is obtained by comparing the historical number of times the target vehicle travels on the historical travel segment corresponding to the first power-on position among the power-on positions as the starting point and the first power-off position among the power-off positions as the ending point with the historical number of times the target vehicle travels on all the historical travel segments corresponding to the first power-on position as the starting point and all power-off positions as the ending points. The vehicle location transfer probability matrix is ​​obtained based on the historical travel probability of the target vehicle in the historical travel segment. Based on the current location, the historical travel probability of the target vehicle in the target historical travel segment is determined from the vehicle location transfer probability matrix. The target historical travel segment is the historical travel segment in which the starting point of the journey is the current location. Based on the current location and the historical vehicle usage information, determine the historical travel time of the target vehicle on the target historical travel segment, and determine the historical travel time probability corresponding to the historical travel time; Based on the power-off time, the historical travel probability, the historical travel time, and the historical travel time probability, the predicted user's vehicle usage segment is determined, and the predicted user's vehicle usage time corresponding to the predicted user's vehicle usage segment is determined. The predicted power-on time of the target vehicle is obtained based on the predicted user vehicle usage time.

2. The vehicle power-on time prediction method according to claim 1, characterized in that, Based on the power-off time, the historical travel probability, the historical travel time, and the historical travel time probability, the predicted user's vehicle usage segment is determined, and the predicted user's vehicle usage time corresponding to the predicted user's vehicle usage segment is determined, including: The historical travel time that is later than the power-off time and has the highest probability among the historical travel times is taken as the target historical travel time. Based on the historical travel probability of the target historical travel segment and the historical travel time probability corresponding to the target historical travel time, the predicted travel probability corresponding to the target historical travel segment is determined. The target historical travel segment with the highest predicted travel probability is taken as the predicted user's vehicle usage segment, and the target historical travel time corresponding to the predicted user's vehicle usage segment is taken as the predicted user's vehicle usage time.

3. The vehicle power-on time prediction method according to claim 2, characterized in that, The historical vehicle usage locations and corresponding historical vehicle usage times are obtained in the following manner: Determine the initial historical vehicle usage location and corresponding initial historical vehicle usage time of the target vehicle; Clustering algorithms are used to cluster the initial historical vehicle usage locations and the initial historical vehicle usage times to obtain the historical vehicle usage locations and the historical vehicle usage times of the target vehicle.

4. The vehicle power-on time prediction method according to claim 3, characterized in that, The historical vehicle usage time includes power-on time and power-off time. Based on the historical vehicle usage location and the corresponding historical usage time, the driving information of multiple historical driving segments of the target vehicle is obtained, including: If it is determined that the time interval between the power-off time of the target vehicle traveling from the first power-on position to the first power-off position and the power-on time of traveling from the first power-off position to the second power-off position is less than a preset first time interval threshold, then the driving information of the target vehicle's historical driving segment is determined as follows: power-on position is the first power-on position, power-off position is the second power-off position, power-on time is the power-on time of the target vehicle at the first power-on position, and power-off time is the power-off time of the target vehicle at the second power-off position.

5. The vehicle power-on time prediction method according to claim 4, characterized in that, Based on the historical vehicle usage information, the number of historical vehicle uses corresponding to the historical driving segments is determined, including: If it is determined that the time interval between the power-on time and the power-off time of the target vehicle traveling from the first power-on position to the first power-off position in the historical driving segment is less than a preset second time interval threshold, and the average speed of the target vehicle is less than a preset speed threshold, then the historical vehicle usage count in the historical driving segment is determined to be zero.

6. The vehicle power-on time prediction method according to claim 5, characterized in that, The vehicle location transfer probability matrix is ​​obtained based on the historical travel probability of the target vehicle on the historical travel segment, including: If the historical vehicle usage information satisfies the corresponding conditions of the Markov chain model, the vehicle location transition probability matrix is ​​obtained based on the historical travel probability of the target vehicle on the historical driving segment.

7. The vehicle power-on time prediction method according to claim 6, characterized in that, Based on the current location and the historical vehicle usage information, the historical travel time of the target vehicle on the target historical travel segment is determined, and the historical travel time probability corresponding to the historical travel time is determined, including: Based on the historical vehicle usage information, determine the historical travel time and historical travel time probability of the target vehicle on the historical driving segment; Based on the current location of the target vehicle, the historical travel time of the target vehicle in the target historical travel segment and the historical travel time probability corresponding to the historical travel time are determined from the historical travel time and historical travel time probability of the historical travel segment.

8. The vehicle power-on time prediction method according to claim 7, characterized in that, Determining the historical travel time and historical travel time probability of the target vehicle on the historical driving segment based on the historical vehicle usage information includes: Based on the power-on time, determine all historical travel times of the target vehicle on the historical travel route, and the number of historical vehicle uses corresponding to all historical travel times; Determine the number of times the target vehicle was used in the historical travel period; The probability of the target vehicle's historical travel time on the historical travel segment at the historical travel time is determined by the ratio of the number of times the target vehicle used the vehicle at the historical travel time to the number of times the vehicle used the vehicle at all historical travel times.

9. The vehicle power-on time prediction method according to claim 8, characterized in that, Determining the historical travel time probability corresponding to the historical travel time includes: The probability of the historical travel time corresponding to each historical travel time of the target vehicle on the historical travel segment is determined based on the Gaussian mixture model.

10. The vehicle power-on time prediction method according to any one of claims 3-9, characterized in that, Based on the current location and the historical vehicle usage information, the historical travel time of the target vehicle on the target historical travel segment is determined, and the historical travel time probability corresponding to the historical travel time is determined, including: If, based on the current location and the historical vehicle usage information, it is determined that the current location is within a preset range of one of the power-on locations of the historical vehicle usage locations, the historical travel time of the target vehicle on the target historical driving segment is determined based on the historical vehicle usage information, and the historical travel time probability corresponding to the historical travel time is determined.

11. The vehicle power-on time prediction method according to claim 2, characterized in that, The target historical driving segment with the highest predicted travel probability is used as the predicted user's vehicle usage segment, including: If the predicted travel probability is greater than a preset travel probability threshold, the target historical travel segment with the highest predicted travel probability is taken as the predicted user travel segment.

12. A vehicle management method, characterized in that, The method includes: Determine the predicted power-on time of the target vehicle in the parked and power-off state, wherein the predicted power-on time is obtained based on the vehicle power-on time prediction method according to any one of claims 1-11; The target vehicle is powered on at the predicted power-on time.

13. The vehicle management method according to claim 12, characterized in that, The method further includes: The predicted destination of the target vehicle in a parked, power-off state is determined, and the predicted destination is determined based on the predicted user route obtained in the vehicle power-on time prediction method according to any one of claims 1-11. The route of the target vehicle is determined based on the predicted destination.

14. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer programs; The processor executes a computer program stored in the memory to enable the electronic device to implement the vehicle power-on time prediction method as described in any one of claims 1-11, and / or the vehicle management method as described in claim 12 or 13.

Images