A method and apparatus for optimizing a trip

Abstract

The application provides a trip optimization method and device, and a computer readable storage medium. The trip optimization method comprises the following steps: obtaining trip planning data and user data of a user; determining a destination and an estimated arrival time of the user according to the trip planning data; predicting an estimated residence time of the user at the destination according to the user data; determining a trip abnormal point according to the estimated arrival time and the estimated residence time; and performing trip optimization processing according to the trip abnormal point. Through the execution of these steps, the trip optimization method can predict the estimated residence time of the user at each interest point according to the actual situation of the user, and optimize the trip of the user in combination with the actual situation of each interest point, thereby helping the user to avoid the trip abnormal point, and improving the trip experience of the user.

Description

Technical Field

[0001] This invention relates to travel data processing technology, and more particularly to a travel optimization method, a travel optimization device, and a computer-readable storage medium. Background Technology

[0002] Journey planning technology manages multiple points of interest (POIs) provided by a user based on time, thus assisting with travel, tourism, and schedule planning. Currently, this technology is widely used in various applications such as navigation, travel, and scheduling.

[0003] However, existing travel planning technologies generally provide travel planning services based on users' actively inputted points of interest and time, which places high demands on users' time management abilities and cannot anticipate various situations that may occur at points of interest. As a result, users are prone to getting stuck in traffic jams, missing meal times, encountering park closures, and experiencing travel conflicts, which seriously affects the user's travel experience.

[0004] In order to overcome the above-mentioned defects in the existing technology, there is an urgent need in the field for a trip data processing technology to optimize the user's trip by combining the actual situation of the user and each point of interest, thereby improving the user's trip experience. Summary of the Invention

[0005] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed descriptions that follow.

[0006] To overcome the aforementioned deficiencies in the prior art, the present invention provides a travel optimization method, a travel optimization device, and a computer-readable storage medium.

[0007] Specifically, the trip optimization method provided by the first aspect of the present invention includes the following steps: acquiring user trip planning data and user data; determining the user's destination and expected arrival time based on the trip planning data; predicting the user's expected stay time at the destination based on the user data; identifying trip anomalies based on the expected arrival time and the expected stay time; and performing trip optimization processing based on the trip anomalies. By performing these steps, the trip optimization method can predict the user's expected stay time at each point of interest based on the user's actual situation, and optimize the user's trip by combining the actual situation of each point of interest, thereby helping the user avoid anomalies in the trip and improving the user's trip experience.

[0008] The trip optimization apparatus provided according to a second aspect of the present invention includes a memory and a processor. The processor is connected to the memory and configured to implement the trip optimization method provided in the first aspect of the present invention. By implementing the trip optimization method, the trip optimization apparatus can predict the user's expected dwell time at each point of interest based on the user's actual situation, and optimize the user's trip by combining the actual situation of each point of interest, thereby helping the user avoid abnormal points in the trip and improving the user's trip experience.

[0009] The computer-readable storage medium provided according to a third aspect of the present invention stores computer instructions thereon. When the computer instructions are executed by a processor, they implement the trip optimization method provided in the first aspect of the present invention. By implementing this trip optimization method, the computer-readable storage medium can predict the user's expected dwell time at each point of interest based on the user's actual situation, and optimize the user's trip by combining the actual situation of each point of interest, thereby helping the user avoid abnormal points in the trip and improving the user's trip experience. Attached Figure Description

[0010] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0011] Figure 1 A schematic diagram illustrating the working principle of a travel optimization device provided according to some embodiments of the present invention is shown.

[0012] Figure 2 A flowchart illustrating a route optimization method provided according to some embodiments of the present invention is shown.

[0013] Figure 3 A schematic diagram of a trip optimization unit provided according to some embodiments of the present invention is shown.

[0014] Figures 4A to 4D A schematic diagram of a trip optimization prompt interface provided according to some embodiments of the present invention is shown. Detailed Implementation

[0015] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Although the description of the present invention is presented in conjunction with preferred embodiments, this does not mean that the features of the invention are limited to these embodiments. On the contrary, the purpose of describing the invention in conjunction with embodiments is to cover other options or modifications that may be derived based on the claims of the present invention. To provide a thorough understanding of the invention, many specific details will be included in the following description. The invention may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of the invention, some specific details will be omitted in the description.

[0016] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0017] Furthermore, the terms "upper," "lower," "left," "right," "top," "bottom," "horizontal," and "vertical" used in the following description should be understood as the orientations shown in the relevant paragraphs and accompanying drawings. These relative terms are for illustrative purposes only and do not imply that the described apparatus must be manufactured or operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0018] It is understood that although terms such as "first," "second," and "third" may be used herein to describe various components, regions, layers, and / or parts, these components, regions, layers, and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers, and / or parts. Therefore, the first components, regions, layers, and / or parts discussed below may be referred to as second components, regions, layers, and / or parts without departing from some embodiments of the present invention.

[0019] As mentioned above, existing travel planning technologies generally provide travel planning services based on users' actively inputted points of interest and time. This places high demands on users' time management abilities and cannot anticipate various situations that may occur at points of interest. Consequently, users are prone to encountering congestion, missing meal times, park closures, and travel conflicts, which seriously affect their travel experience.

[0020] To overcome the aforementioned deficiencies in the prior art, the present invention provides a trip optimization method, a trip optimization device, and a computer-readable storage medium, which can predict the user's expected dwell time at each point of interest based on the user's actual situation, and optimize the user's trip by combining the actual situation of each point of interest, thereby helping the user avoid abnormal points in the trip and improving the user's trip experience.

[0021] In some non-limiting embodiments, the trip optimization method provided in the first aspect of the present invention can be implemented by the trip optimization device provided in the second aspect of the present invention. This trip optimization device can be configured on a cloud server of a trip planning application and / or a client running the trip planning application, in the form of hardware devices and / or software programs. Here, the trip planning application includes, but is not limited to, various applications, applets, and web apps such as map navigation, ride-hailing, travel, schedules, reminders, and smart assistants. The client includes, but is not limited to, various electronic devices such as in-vehicle systems, tablets, PDAs, laptops, personal computers (PCs), smartphones, smartwatches, smart bracelets, and smart glasses.

[0022] Furthermore, the trip optimization apparatus includes a memory and a processor. The memory includes, but is not limited to, the computer-readable storage medium described in the third aspect of the present invention, on which computer instructions are stored. The processor is connected to the memory and configured to execute the computer instructions stored in the memory to implement the trip optimization method described in the first aspect of the present invention.

[0023] The working principle of the above-mentioned trip optimization device will be described below with reference to some embodiments of trip optimization methods. Those skilled in the art will understand that these embodiments of trip optimization methods are merely non-limiting implementations provided by the present invention, intended to clearly demonstrate the main concepts of the invention and provide specific solutions convenient for public implementation, rather than limiting all functions or all operating methods of the trip optimization device. Similarly, the trip optimization device is also only one non-limiting implementation provided by the present invention, and does not limit the types of trip planning applications described above, or the entities executing the steps in these trip optimization methods.

[0024] Please refer to the reference. Figure 1 and Figure 2 . Figure 1 A schematic diagram illustrating the working principle of a travel optimization device provided according to some embodiments of the present invention is shown. Figure 2 A flowchart illustrating a route optimization method provided according to some embodiments of the present invention is shown.

[0025] like Figure 1 As shown, in some embodiments of the present invention, the trip optimization device may be configured with a trip planning unit, a stay time estimation unit, and a trip optimization unit. The trip planning unit is used to realize data interaction between the trip optimization device and clients such as mobile phones and in-vehicle systems, and to provide trip planning services to users based on acquired trip planning data, user data, and / or trip optimization data. The stay time estimation unit is equipped with a pre-trained artificial intelligence (AI) model, which can predict the user's estimated stay time at the destination based on acquired user data. The trip optimization unit can identify unreasonable anomalies in the user's trip and perform trip optimization processing based on the identified anomalies, thereby helping users avoid anomalies in the trip and improving the user's trip experience.

[0026] Specifically, such as Figure 2 As shown, during the implementation of the trip optimization method, in response to users' trip-related operations such as navigation, ride-hailing, travel, calendar, reminders, and smart assistants using in-vehicle systems, mobile phones, and other clients, the pre-installed tracking programs in each application can obtain the user's provided trip planning data and user data based on the user's authorized permissions, and upload them to the trip optimization device. In addition, the trip optimization device can also obtain user data from the user's cloud account as a data basis for predicting the user's expected stay time.

[0027] In some embodiments, the aforementioned trip planning data includes, but is not limited to, navigation information indicating the user's departure location, departure time, destination location, and / or estimated arrival time; taxi information; flight information; train ticket information; long-distance bus ticket information; attraction ticket information; entertainment ticket information; accommodation booking information; and / or restaurant reservation information. The aforementioned user data can be divided into basic user data and user behavior data. Basic user data includes, but is not limited to, the user's age, gender, and / or job title. User behavior data includes, but is not limited to, the user's historical travel data and / or historical length of stay data.

[0028] For example, for in-vehicle infotainment systems, the trip optimization device can obtain the destination location of a relevant trip through navigation locations, related service order records, and schedules, and calculate the time the vehicle spends near each destination. Furthermore, based on user-authorized permissions, the trip optimization device can also obtain basic user data such as gender, age, and occupation of each occupant from the in-vehicle infotainment system, and obtain business data of the user's transactions at the destination through orders and schedules.

[0029] For example, for smartphones, trip optimization devices can obtain the destination location of a trip through GPS positioning, navigation locations, related software order records, calendar records, and related merchant call records, and calculate the time the user spends near each destination. Furthermore, based on user-provided authorization, trip optimization devices can also obtain basic user data such as gender, age, and occupation from the smartphone.

[0030] After acquiring the user's trip planning data and user data, the trip optimization device can first filter the trip planning data regarding destination location and estimated arrival time to directly determine the user's destination and corresponding estimated arrival time. Furthermore, for trip planning data that does not include the estimated arrival time, the trip optimization device can preferably calculate the user's estimated arrival time at the corresponding destination based on information such as departure location, departure time, current location, current time, remaining mileage, and / or average speed.

[0031] In addition, such as Figure 1 and Figure 2 As shown, after acquiring the user's itinerary planning data and user data, the itinerary optimization device can also input the acquired user data into the stay time prediction unit, which will then predict the user's expected stay time at the destination based on the acquired user data.

[0032] Specifically, the dwell time prediction unit is equipped with a pre-trained clustering algorithm model and a neural network model. The clustering algorithm model includes, but is not limited to, DBSCAN clustering, K-Means, mean-shift clustering, and expectation-maximum clustering based on Gaussian mixture models. The neural network model includes, but is not limited to, a Long Short-Term Memory (LSTM) neural network model.

[0033] Taking the dwell time prediction unit composed of the DBSCAN clustering algorithm model and the LSTM neural network model as an example, during the training process of the DBSCAN clustering algorithm model, the trainer can first obtain user data samples from multiple user samples. Each user data sample includes basic user data such as the user's job title, age, and / or gender, as well as user behavior data such as historical travel data and / or historical dwell time data. Further, the historical travel data includes, but is not limited to, the user's historical departure location, historical departure time, historical destination location, historical arrival time, and / or historical travel transactions. Then, the trainer can sequentially input the user data samples from each user sample into the DBSCAN clustering algorithm model to be trained, and iteratively update the weight parameters of the DBSCAN clustering algorithm model based on the output error, in order to train the DBSCAN clustering algorithm model to output corresponding user category labels based on the input user data.

[0034] Because the density-based DBSCAN clustering algorithm can determine user categories based on the density of sample distribution, user samples within the same category identified by DBSCAN are closely linked and often have similar or even identical dwell times. Furthermore, users with different job positions may exhibit different behavioral patterns at the same point of interest. For example, company HR personnel might visit a bank for corporate business, while other employees primarily handle personal transactions. Therefore, users with different job positions may have different dwell times at the same point of interest. Thus, by incorporating user job positions into the DBSCAN clustering algorithm's input parameters, users can be more accurately classified, leading to a more precise determination of the dwell time for each category at each point of interest.

[0035] Furthermore, compared to K-Means, mean-shift clustering, and Gaussian mixture model-based expectation-maximum clustering, the density-based DBSCAN clustering algorithm does not require pre-setting the K value to adaptively determine the clustering results. Moreover, because DBSCAN can effectively identify outliers, and even misidentification of outliers has little impact on the final clustering result, the residence time prediction unit using DBSCAN is less sensitive to noise. Additionally, unlike K-Means and other clustering algorithms where initial values ​​significantly influence the results, DBSCAN's clustering results are unbiased and unaffected by initial values. Furthermore, since DBSCAN discovers clusters by continuously connecting high-density neighboring points, it only requires defining the neighborhood size and density threshold to discover clusters of different shapes and sizes. Therefore, the residence time prediction unit using DBSCAN can discover clusters of arbitrary shapes and more accurately determine the user category to which each user belongs.

[0036] Furthermore, after completing the training process of the DBSCAN clustering algorithm model described above, the trainer can obtain user data from multiple user samples. Historical travel data and historical stay time data are used as the trip planning data samples for the corresponding user samples, while age, gender, job title, business type, and weather data are used as the user data samples for each user sample. Then, the trainer can input the user data samples from each user sample into the trained DBSCAN clustering algorithm model to determine the user category for each user sample and perform bucketing operations on each user sample based on historical stay time.

[0037] In some preferred embodiments, the above-described bucketing step can be performed based on the principle of equal-frequency bucketing to prevent excessive differences in data volume between buckets. Please refer to Tables 1 and 2 for details. Table 1 records user data for multiple user samples without bucketing, according to some embodiments of the present invention. Table 2 records user data for multiple user samples after bucketing, according to some embodiments of the present invention.

[0038] Table 1

[0039] Time of arrival at destination Duration of stay (minutes) Handling business User Categories weather ... 2021-09-15 9:03:57 32 0 0 0 ... 2021-09-15 9:08:02 14 1 1 0 2021-09-15 9:24:45 67 1 3 0 2021-09-15 9:31:21 45 1 2 0 2021-09-15 10:08:18 22 2 0 0 2021-09-15 10:42:07 78 1 1 2

[0040] As shown in Table 1, during the process of binning user samples, the trainer can bin each user sample equally based on the historical dwell time, and determine the boundary of each bin's sample data based on the maximum and minimum historical dwell time of each bin, so that each bin has the same number of samples. Specifically, for the embodiment shown in Table 1, the trainer can divide each user sample into 10 buckets based on historical dwell time. The historical dwell time of the first bucket of sample data is between 0 and 15 minutes, the historical dwell time of the second bucket of sample data is between 16 and 30 minutes, the historical dwell time of the third bucket of sample data is between 31 and 45 minutes, the historical dwell time of the fourth bucket of sample data is between 46 and 60 minutes, the historical dwell time of the fifth bucket of sample data is between 61 and 90 minutes, the historical dwell time of the sixth bucket of sample data is between 91 and 120 minutes, the historical dwell time of the seventh bucket of sample data is between 121 and 150 minutes, the historical dwell time of the eighth bucket of sample data is between 151 and 180 minutes, the historical dwell time of the ninth bucket of sample data is between 181 and 240 minutes, and the historical dwell time of the tenth bucket of sample data is more than 240 minutes, thereby obtaining the user data of multiple user samples after bucketing as shown in Table 2.

[0041] Table 2

[0042] Time of arrival at destination Y X1 X2 X3 ... 2021-09-15 9:03:57 3 0 0 0 ... 2021-09-15 9:08:02 1 1 1 0 2021-09-15 9:24:45 5 1 3 0 2021-09-159:31:21 4 1 2 0 2021-09-15 10:08:18 2 2 0 0 2021-09-15 10:42:07 5 1 1 2

[0043] After completing the binning operation for each user sample, the trainer can also determine the activation function of the LSTM neural network model, the discard rate of each network node, the error calculation method, the iterative update method of the weight parameters, as well as the number of training cycles (epochs) and the batch size.

[0044] Taking the Keras library in Python as an example, the trainer can set the default tanh in Keras as the activation function of the LSTM neural network model, determine the activation function of the fully-connected artificial neural network that receives the LSTM output based on the default linear in Keras, set the default discard rate of each network node to 0.2 to prevent overfitting, determine the error calculation method as mean squared error, use the RMSprop algorithm to determine the iterative update method of the weight parameters, and determine the number of training loops and the window size of the model.

[0045] After determining and adjusting the structure of the LSTM neural network model, the trainer can sequentially input the trip planning data samples and user data samples of each user sample into the LSTM neural network model to be trained, and iteratively update the weight parameters of the LSTM neural network model according to the output error of the LSTM neural network model, so as to train the LSTM neural network model to output the corresponding expected dwell time label according to the input trip planning data and user data.

[0046] Those skilled in the art will understand that the above description of a trainer is merely a non-limiting one, and includes, but is not limited to, a technician performing the above training steps, and / or a processor executing computer instructions conforming to the above training steps.

[0047] Subsequently, in predicting the user's expected length of stay at the destination, the length of stay prediction unit first inputs basic user data such as age, gender, and / or job title provided by the trip planning unit, as well as user behavior data such as historical travel data and / or historical length of stay data, into a pre-trained DBSCAN clustering algorithm model. The DBSCAN clustering algorithm model then determines the user's category based on the input user data. Next, the length of stay prediction unit inputs the user category label output by the DBSCAN clustering algorithm model, along with the user's expected arrival time at the corresponding destination, into a pre-trained LSTM neural network model. This LSTM neural network model then predicts the user's expected length of stay at the corresponding destination based on the input user category and expected arrival time.

[0048] Furthermore, in some embodiments of the present invention, the trip optimization device can also obtain weather data near the destination from third-party platforms such as meteorological monitoring agencies, and input this weather data along with user data such as basic user data, user behavior data, and business data into a pre-trained LSTM neural network model. The LSTM neural network model then predicts the user's expected stay time at the destination based on the input user category, estimated arrival time, business data, and weather data. By further incorporating the computational dimensions of weather data and / or business data, the present invention can further combine the influencing factors of weather and / or business on stay time, thereby more accurately predicting the user's expected stay time at the destination.

[0049] Furthermore, in some embodiments of the present invention, the trip optimization device can preferably obtain information about the user's companions from the client. This companion information indicates at least one companion of the user, which can be actively input by the user or automatically determined by the processor based on each user's navigation information, taxi information, flight information, train ticket information, long-distance bus ticket information, attraction ticket information, entertainment ticket information, accommodation reservation information, and / or restaurant reservation information. In predicting the expected stay time, the trip optimization device can obtain user data of each companion based on the user's companion information, and through a pre-trained clustering algorithm model, determine the user category of each companion based on the various user data. Then, through a pre-trained neural network model, predict the expected stay time of each companion at the destination based on their user category. Afterwards, the trip optimization device can predict the user's expected stay time at the destination based on the maximum value among the user's and each companion's expected stay times. Thus, the present invention can further incorporate the mutual influence of companions, thereby more accurately predicting the user's expected stay time at the destination.

[0050] like Figure 1 and Figure 2 As shown, after determining the user's expected stay time at the corresponding destination, the trip optimization device can send the user's destination, expected arrival time, and expected stay time to the trip optimization unit. The trip optimization unit then identifies unreasonable anomalies in the trip based on the user's expected arrival time and expected stay time, and performs trip optimization processing based on the identified anomalies.

[0051] Please refer to the details. Figure 3 and Figures 4A to 4D . Figure 3 A schematic diagram of a trip optimization unit provided according to some embodiments of the present invention is shown. Figures 4A to 4D A schematic diagram of a trip optimization prompt interface provided according to some embodiments of the present invention is shown.

[0052] like Figure 3 As shown, in some embodiments of the present invention, the trip optimization unit can first predict the user's expected departure time from the destination based on the expected arrival time t1 and the expected stay time Δt, i.e., t2 = t1 + Δt. Then, the trip optimization unit can determine whether there are conflict trip anomalies, peak trip anomalies, and / or stay trip anomalies related to the destination based on the expected arrival time t1 and / or the expected departure time t2, and perform trip optimization processing based on the identified trip anomalies.

[0053] For example, the trip optimization unit can compare the estimated arrival time t1 (e.g., 14:00) and estimated departure time t2 (e.g., 15:30) of a destination (e.g., the Ancestral Temple) with the user's existing trip planning data to determine if there are any conflicting trip anomalies related to the corresponding destination. In response to the existence of planned trips within the estimated arrival time t1 and estimated departure time t2 (e.g., having afternoon tea) in the trip planning data, the trip optimization unit can determine that there is a conflicting trip anomaly related to the Ancestral Temple. At this time, the trip optimization device can... Figure 4A As shown, the client provides users with itinerary optimization prompts, suggesting that users postpone or cancel the planned afternoon tea trip, or advance or cancel the planned visit to the Ancestral Temple, in order to help users eliminate the conflicting itinerary anomalies.

[0054] For example, the itinerary optimization unit can compare the estimated arrival time t1 (e.g., 10:00) and estimated departure time t2 (e.g., 14:00) of a destination (e.g., Fairy Mountain) with the user's default itinerary data to determine if there are any conflicting itinerary anomalies for the corresponding destination. This default itinerary data includes, but is not limited to, lunch at 12:00, dinner at 18:00, and accommodation at 22:00. Users can manually add, adjust, and / or delete default itineraries, or the processor can automatically adjust the default itinerary time based on the user's historical itinerary data. In response to the existence of a default itinerary (e.g., lunch) between the estimated arrival time t1 and estimated departure time t2 in the default itinerary data, the itinerary optimization unit can determine that there is a conflicting itinerary anomaly for the Fairy Mountain itinerary. At this time, the itinerary optimization device can... Figure 4B As shown, the app provides users with trip optimization tips via the client and recommends restaurants and points of interest near Fairy Mountain based on the destination's location, helping users complete their default trip of having lunch at 12:00.

[0055] For example, the itinerary optimization unit can also obtain a large number of users' arrival start times and departure end times from third-party platforms such as the destination's (e.g., the Ancestral Temple's) management server and navigation application server, and count the number of people at the destination during each time period to determine the peak tourist season (e.g., 14:00-15:00). Then, the itinerary optimization unit can compare the user's expected arrival time t1 (e.g., 14:00) with this peak tourist season. In response to the comparison result where the peak tourist season coincides with the expected arrival time t1, the itinerary optimization unit can determine that there is an anomaly in the peak tourist season for the Ancestral Temple visit. At this time, the itinerary optimization device can... Figure 4CThe system provides users with trip optimization tips via the client, suggesting them to advance, postpone, or cancel their visit to the Ancestral Temple to help them eliminate peak travel disruptions.

[0056] For example, in applications such as driving navigation and ride-hailing, the trip optimization unit can obtain the traffic flow at the destination (e.g., the Ancestral Temple) at the expected departure time t2 (e.g., 15:30) from third-party platforms such as traffic management platforms and navigation application servers, and compare this traffic flow with a preset traffic flow threshold. If the comparison result indicates that the traffic flow exceeds the preset threshold, the trip optimization unit can determine that the user's expected departure time t2 coincides with the peak traffic period of the destination, indicating an anomaly in the peak travel time for the Ancestral Temple visit. In this case, the trip optimization device can... Figure 4D As shown, the trip optimization device provides users with trip optimization tips via the client, prompting them to change their mode of transportation to and / or from their destination to eliminate peak travel disruptions. Alternatively, the trip optimization device can also... Figure 4D The system provides users with trip optimization tips via the client, recommending locations such as cafes, teahouses, and water bars near the Ancestral Temple based on the destination's location, to help users complete their plans for afternoon tea at 3:15 PM.

[0057] Furthermore, such as Figure 3 As shown, the trip optimization unit can also preferably obtain the historical dwell time of multiple users at a destination (e.g., a popular spring pancake shop) and compare it with a preset dwell time threshold (e.g., 10 minutes). In response to the judgment result that the historical dwell time of multiple users at the destination is less than the preset dwell time threshold, the trip optimization unit can infer that the destination may be closed and determine that there is an abnormal dwell time point in the trip to that destination.

[0058] In addition, the itinerary optimization unit can crawl information such as business hours, holiday notices, and renovation / maintenance announcements from the destination's servers, official websites, news websites, and other third-party platforms. It can then parse this information to filter out details indicating destination closures. In response to the parsing results indicating a destination closure, the itinerary optimization unit can also determine if there are any abnormal points in the itinerary related to that destination.

[0059] Subsequently, in response to the determination that there is an abnormal point in the itinerary regarding the destination, the itinerary optimization device can provide the user with itinerary optimization prompts via the client as described above, prompting the user to cancel the itinerary regarding the destination, so as to help the user eliminate the abnormal point in the itinerary.

[0060] Based on the above description, the present invention can predict the user's expected stay time at each point of interest according to the user's actual situation, and optimize the user's itinerary by combining the actual situation of each point of interest, thereby helping the user avoid abnormal points in the itinerary and improving the user's travel experience.

[0061] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0062] Those skilled in the art will understand that information, signals, and data can be represented using any of a variety of different techniques and arts. For example, the data, instructions, commands, information, signals, bits, symbols, and chips described throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, light fields or optical particles, or any combination thereof.

[0063] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0064] Although the travel optimization device described in the above embodiments can be implemented through a combination of software and hardware, it is understood that the travel optimization device can also be implemented independently in software or hardware. For hardware implementation, the travel optimization device can be implemented using one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic devices for performing the above functions, or a selected combination of the above devices. For software implementation, the travel optimization device can be implemented using independent software modules such as procedures and functions running on a general-purpose chip, each module performing one or more functions and operations described herein.

[0065] The various illustrative logic modules and circuits described in conjunction with the embodiments disclosed herein may be implemented or performed using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but in alternatives, it may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0066] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A route optimization method, Its features are, Includes the following steps: Obtain user travel planning data and user data; Based on the trip planning data, the user's destination and estimated arrival time are determined; Based on the user data, predict the user's expected length of stay at the destination; Based on the estimated arrival time and the estimated stay time, identify the points of trip anomaly; as well as Based on the aforementioned trip anomalies, trip optimization processing is performed; The step of predicting the user's expected length of stay at the destination based on the user data includes: Using a pre-trained clustering algorithm model, the user category to which the user belongs is determined based on various user data; and The expected length of stay of a user at the destination is predicted by a pre-trained neural network model based on the user category and the expected arrival time. The neural network model is an LSTM neural network model. The step of determining itinerary anomalies based on the estimated arrival time and the estimated stay time includes: Based on the estimated arrival time and the estimated stay time, predict the user's estimated departure time from the destination; and The estimated arrival time and estimated departure time are compared with the user's trip planning data and / or default trip data to determine whether there are any trip anomalies regarding the destination.

2. The trip optimization method as described in claim 1, wherein, The trip planning data includes the user's departure location, departure time, destination location, and / or estimated arrival time, and / or The user data includes basic user data and user behavior data. The basic user data includes the user's age, gender, and / or job title. The user behavior data includes the user's historical travel data and / or historical stay time data.

3. The trip optimization method as described in claim 1, wherein, The trip planning data also includes weather data for the destination, and the user data also includes the user's business data at the destination. The step of predicting the user's expected length of stay at the destination using a pre-trained LSTM neural network model, based on the user category and the expected arrival time, includes: Using a pre-trained LSTM neural network model, the expected length of stay of the user at the destination is predicted based on the user category, the expected arrival time, the business data, and the weather data.

4. The trip optimization method as described in claim 3, wherein, The steps for training the LSTM neural network model include: Obtain trip planning data samples and user data samples from multiple user samples; Based on historical dwell time, the user samples are divided into buckets. Determine the activation function of the LSTM neural network model, the dropout rate of each network node, the error calculation method, the iterative update method of the weight parameters, and the number of training loops and window size; and The trip planning data samples and user data samples of each user sample are sequentially input into the LSTM neural network model to be trained, and the weight parameters of the LSTM neural network model are iteratively updated according to the output error of the LSTM neural network model, so as to train the LSTM neural network model to output the corresponding expected stay time label according to the input trip planning data and user data.

5. The trip optimization method as described in claim 4, wherein, The step of binning each user sample based on historical dwell time includes: The user samples are divided into equal-frequency buckets based on the historical dwell time, and the boundaries of the sample data in each bucket are determined based on the maximum and minimum values ​​of the historical dwell time in each bucket, so that each bucket has the same number of samples.

6. The trip optimization method as described in claim 1, wherein, The clustering algorithm model is the DBSCAN clustering algorithm model, and the steps for training the clustering algorithm model include: Obtain user data samples from multiple user samples; and The user data samples of each user sample are sequentially input into the DBSCAN clustering algorithm model to be trained, and the weight parameters of the DBSCAN clustering algorithm model are iteratively updated according to the output error of the DBSCAN clustering algorithm model, so as to train the DBSCAN clustering algorithm model to output the corresponding user category label according to the input user data.

7. The route optimization method as described in claim 1, wherein, The user data includes information about fellow travelers, which indicates at least one fellow traveler. The step of predicting the user's expected length of stay at the destination based on the user data further includes: Based on the aforementioned companion information, obtain the user data of each of the aforementioned companions; Based on the clustering algorithm model, the user category to which each of the aforementioned peers belongs is determined according to the various user data of each peer. Using the neural network model, the expected length of stay for each of the accompanying persons is predicted at the destination based on their user category; and The expected length of stay of the user at the destination is predicted based on the maximum of the expected length of stay of the user and each of the accompanying persons at the destination.

8. The trip optimization method as described in claim 1, wherein, The step of determining whether there are any outliers in the itinerary to the destination includes: In response to the existence of a planned itinerary in the itinerary planning data located between the estimated arrival time and the estimated departure time, a conflicting itinerary anomaly is determined regarding the destination; and / or In response to the existence of a default itinerary in the default itinerary data that lies between the estimated arrival time and the estimated departure time, a conflicting itinerary anomaly is determined regarding the destination.

9. The trip optimization method as described in claim 8, wherein, The step of optimizing the trip based on the trip anomalies includes: In response to a conflict with the planned itinerary regarding the destination, the user is prompted to advance, postpone, or cancel the planned itinerary; or the user is prompted to advance, postpone, or cancel the itinerary regarding the destination; and / or In response to a conflicting trip point with the default trip for the destination, the system recommends points of interest (POIs) to the user that can perform the corresponding functions based on the location of the destination.

10. The trip optimization method as described in claim 1, wherein, The step of determining itinerary anomalies based on the estimated arrival time and the estimated stay time further includes: Determine the peak time period for the destination; The estimated arrival time of the user at the destination is compared with the peak time period of the destination; and In response to the overlap between the peak time period and the expected arrival time, it is determined that there is a peak travel anomaly in the itinerary to the destination.

11. The trip optimization method as described in claim 10, wherein, The step of determining itinerary anomalies based on the estimated arrival time and the estimated stay time further includes: Obtain the traffic flow at the destination at the expected departure time; and In response to the traffic flow exceeding a preset traffic flow threshold, it is determined that there is a peak travel anomaly point in the trip to the destination.

12. The route optimization method as described in claim 11, wherein, The step of optimizing the trip based on the trip anomalies includes: In response to an anomaly in the travel schedule regarding the estimated arrival time at the destination, the user is prompted to reschedule their trip to the destination in advance. In response to an anomaly in the trip to the destination regarding the estimated departure time, the user is prompted to postpone the trip to the destination. In response to the presence of peak travel anomalies for both the estimated arrival time and the estimated departure time for the trip to the destination, the user is prompted to cancel the trip to the destination; and / or In response to an anomaly in the travel schedule regarding the destination, with an expected arrival time and / or expected departure time, the user is prompted to change their mode of transportation to and / or from the destination.

13. The trip optimization method as described in claim 1, wherein, The step of determining itinerary anomalies based on the estimated arrival time and the estimated stay time further includes: Obtain the historical stay time of multiple users at the destination; and In response to the judgment result that the historical stay time of the multiple users at the destination is less than a preset stay time threshold, it is determined that there is an abnormal stay time point in the itinerary of the destination.

14. The trip optimization method as described in claim 1, wherein, After determining the user's destination based on the trip planning data, the trip optimization method further includes the following steps: Obtain and parse the announcement information of the destination; and In response to the parsing result that the destination is closed, it is determined that there is a stop-trip anomaly in the itinerary related to the destination.

15. The route optimization method as described in claim 13 or 14, wherein, The step of optimizing the trip based on the trip anomalies includes: In response to the presence of an abnormal stop point in the itinerary for the destination, the user is prompted to cancel the itinerary for the destination.

16. A travel optimization device, characterized in that, include: Memory; as well as A processor, connected to the memory, and configured to implement the trip optimization method as described in any one of claims 1 to 15.

17. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, the trip optimization method as described in any one of claims 1 to 15 is implemented.

Images