An outdoor activity trip dynamic generation method and system based on multi-source heterogeneous data

Abstract

The application provides a kind of outdoor activity trip dynamic generation method and system based on multi-source heterogeneous data, it is related to big dipper positioning technology field, the application is by real-time monitoring to activity object, collect position and trajectory, line performance characteristics, activity environment and path constraint and trip execution feedback and other multi-source data;Multi-source heterogeneous trip data is trained using convolutional neural network, constructs trip feature extraction model, and constructs AI dynamic adjustment model in combination with trip execution feedback;Respectively calculate trip continuous adaptation coefficient, trip environment risk coupling coefficient and trip execution deviation feedback coefficient, and evaluate, when the evaluation result does not satisfy the requirement, respectively trigger shortening trip, path replacement, time window adjustment and trip intensity self-adaptive correction and other strategies, dynamically re-plan and continuously optimize trip.The application can realize the comprehensive promotion of trip safety, continuity and user experience in complex and changeable outdoor environment.

Description

Technical Field

[0001] This invention relates to the field of BeiDou positioning technology, specifically to a method and system for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data. Background Technology

[0002] With the increasing popularity of outdoor activities such as hiking, mountaineering, cycling, and camping, users are demanding higher levels of safety, adaptability, and personalization in their trip planning. Existing outdoor trip planning technologies largely rely on static route recommendations or navigation methods based on single location data. These typically use satellite positioning to obtain the location information of the participants and generate a fixed itinerary plan before the trip begins, lacking the ability to continuously perceive and respond to dynamic changes during the activity. On the one hand, during outdoor activities, the participants' abilities, behavioral states, and individual preferences change over time, while weather conditions, terrain slope, and route accessibility in the activity environment also exhibit significant uncertainty. Existing BeiDou-based trip planning schemes primarily focus on location and trajectory recording, making it difficult to incorporate these changing factors into the trip decision-making process in a timely manner. This can easily lead to a mismatch between the trip's intensity and actual capabilities, unreasonable route selection, or increased potential safety risks.

[0003] On the other hand, existing technologies do not make sufficient use of deviation feedback during the execution of the trip. Although real-time trajectory can be obtained and path deviation can be detected through BeiDou positioning, most solutions only record the event or issue simple alarms afterward. They lack a mechanism to adaptively correct and continuously optimize the trip model based on the execution deviation, making it difficult to form an effective closed-loop adjustment of the trip.

[0004] In addition, although some existing solutions have introduced BeiDou positioning and environmental information collection methods, they mostly remain at the level of data display or rule judgment. They lack sufficient exploration of the inherent correlation between multi-source heterogeneous travel data and have failed to use intelligent models to uniformly model travel preferences, path selection characteristics and environmental adaptability, which restricts the intelligent and dynamic adaptability of outdoor travel planning in complex environments. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data, comprising the following steps: Step 1: By monitoring the participants in hiking, mountaineering and camping outdoor activities in real time, collect data on the location and trajectory of the participants, their behavior and movement characteristics, the environment and path constraints, and the feedback data on the execution of the trip. Step 2: Train a multi-source heterogeneous travel data using a convolutional neural network to build a travel feature extraction model; characterize travel preferences, path selection, and environmental adaptation based on intermediate layer feature vectors, and enhance these characteristics by integrating travel execution feedback to build an AI dynamic adjustment model; Step 3: Calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing risky paths, and calling AI to dynamically adjust the model to replan the trip will be adopted. Step 4: Calculate the coupling coefficient of the trip environment risk and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If not, implement strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. Step 5: Calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If not, provide an adaptive correction strategy for trip intensity, path selection, and environmental matching.

[0007] Further, step one includes: S11. Real-time monitoring of the spatial location and trajectory status of participants during hiking, mountaineering, and camping outdoor activities. By configuring satellite positioning equipment on the participants' terminals and using BeiDou satellite navigation and positioning technology, the real-time location coordinates, positioning update timestamps, and continuous trajectory point sequences of the participants are obtained; and a dataset of the participants' location and trajectory is established. S12. Real-time monitoring of the behavior and movement characteristics of participants during outdoor activities. Through the accelerometer and angular velocity sensor built into the terminal device, the movement speed, direction of movement, and duration of stay of the participants are continuously collected. Combined with historical activity records in the cloud, the path selection behavior and historical trajectory characteristics of the participants in similar outdoor activities are obtained. A dataset of activity behavior and movement characteristics is established. S13. Monitor the environmental conditions and path conditions of the outdoor activity area in real time. Obtain weather information in the activity area through the external environment data interface, and obtain terrain slope information, path accessibility information and environmental risk warning information corresponding to the activity area through the path status interface; establish an activity environment and path constraint dataset. S14. Monitor the actual execution process of outdoor activities in real time. By comparing and analyzing the real-time trajectory of the activity subject obtained based on Beidou satellite positioning technology with the current planned route, obtain trajectory offset data; synchronously collect the deviation data between the actual execution time and the expected duration of the trip, as well as the user interaction operation information generated by the activity subject during the trip execution; and establish a trip execution feedback dataset.

[0008] Furthermore, step two includes: S21. Using convolutional neural networks, construct an initial convolutional neural network model for outdoor activity itineraries. Use the activity object location and trajectory dataset, activity behavior and movement feature dataset, activity environment and path constraint dataset, and itinerary execution feedback dataset as model inputs to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as the itinerary feature extraction model. S22. Based on the trip feature extraction model, extract the intermediate layer output as the trip feature vector, analyze the travel preferences, path selection characteristics and environmental adaptation characteristics of the activity object; and enhance the trip feature vector by combining real-time trajectory offset data, trip execution time deviation data and user interaction operation information. S23. Based on the enhanced trip feature vector, train and test the trip feature extraction model. Use the trained trip feature extraction model as the AI ​​dynamic adjustment model to dynamically generate and predict outdoor activity trips, and output trip planning results that match the current activity status and environmental constraints.

[0009] Furthermore, step three includes: S31. Based on the real-time location coordinates and positioning update timestamps of the active objects in the active object location and trajectory dataset, and combined with the analysis results of trajectory time series features by the travel feature extraction model, the speed change rate calculation method of adjacent sampling points is used to quantify the speed fluctuation of the active object during its movement and obtain the movement speed stability parameter. S32. Based on the continuous trajectory point sequence in the active object location and trajectory dataset, and combined with the statistical analysis results of the path direction change characteristics by the travel feature extraction model, the direction offset angle distribution statistical method is used to obtain the travel direction consistency parameter. S33. Based on the real-time trajectory offset data in the trip execution feedback dataset, and combined with the fusion analysis results of the path execution deviation features by the trip feature extraction model, the trajectory-path spatial deviation calculation method is used to obtain the path offset parameter.

[0010] Furthermore, step three also includes: S34. After dimensionless processing of the obtained travel speed stability parameters, travel direction consistency parameters, and path offset parameters, the travel continuity adaptation coefficient is calculated. S35. By setting a preset travel continuity adaptation threshold and comparing the travel continuity adaptation coefficient with the travel continuity adaptation threshold, the first evaluation result is obtained, including: When the trip continuity adaptation coefficient is greater than or equal to the trip continuity adaptation threshold, it indicates that the current trip continuity adaptation is qualified and there is no risk of trip interruption. Continuous monitoring is required. When the trip continuity adaptation coefficient is less than the trip continuity adaptation threshold, it indicates that the current trip continuity adaptation is unqualified, and there is a risk of unstable travel and directional deviation, which may lead to trip interruption or decreased experience. This triggers the first warning instruction and generates the first strategy: shorten the distance of subsequent trip segments to reduce the burden of continuous travel; replace trip segments with path deviation risk and select alternative paths; after completing the trip segment adjustment, regenerate subsequent trip planning segments based on the AI ​​dynamic adjustment model to match them with the current activity status and environmental constraints; recalculate until the trip continuity adaptation coefficient is greater than or equal to the trip continuity adaptation threshold.

[0011] Furthermore, step four includes: S41. Based on the weather status information and environmental risk warning information in the activity environment and path constraint dataset, and combined with the analysis results of environmental adaptability characteristics by the itinerary feature extraction model, the weighted risk scoring method is used to quantify the environmental risk intensity of the current activity area and obtain the environmental risk intensity parameter. S42. Based on the terrain slope information and path traffic status information in the activity environment and path constraint dataset, and combined with the statistical analysis results of terrain adaptation characteristics by the travel feature extraction model, the terrain complexity parameter is obtained by using the comprehensive index calculation method of slope and undulation. S43. Based on the trajectory offset data, the deviation data between the actual execution time and the expected trip time in the trip execution feedback dataset, and the user interaction operation information generated during the trip execution, and combined with the trip feature extraction model to analyze the trajectory offset evolution characteristics, execution rhythm change characteristics, and user active intervention behavior characteristics, the cumulative counting and weighted statistical method of execution deviation is used to calculate the cumulative number of trajectory offset exceeding the limit, the number of execution time deviation exceeding the limit, and the number of user active intervention operations within a unit trip, and obtain the risk accumulation parameter.

[0012] Furthermore, step four also includes: S44. After dimensionless processing of the obtained environmental risk intensity parameters, terrain complexity parameters, and risk accumulation parameters, the travel environment risk coupling coefficient is calculated. S45. By setting a preset travel environment risk threshold and comparing the travel environment risk coupling coefficient with the travel environment risk threshold, the second assessment result is obtained, including: When the coupling coefficient of the trip environment risk is less than or equal to the trip environment risk threshold, it means that the current trip environment risk is within an acceptable range, and the current trip can continue to be executed while continuous monitoring continues. When the travel environment risk coupling coefficient is greater than the travel environment risk threshold, it indicates that the current travel environment risk exceeds the standard, and there is a risk of activity interruption or safety. This triggers a second warning instruction and generates a second strategy: replace the routes of the risky travel segments; dynamically adjust the travel time window to avoid risky periods; after adjustment, call the AI ​​dynamic adjustment model to generate updated travel planning results, and recalculate after completing the travel segment adjustment until the travel environment risk coupling coefficient is less than or equal to the travel environment risk threshold.

[0013] Furthermore, step five includes: S51. Based on the deviation data between the actual execution time and the expected execution time of the trip in the trip execution feedback dataset, and combined with the analysis results of the trip rhythm characteristics by the AI ​​dynamic adjustment model, the time difference calculation method is used to obtain the trip execution time deviation parameter. S52. Based on the trajectory offset data in the trip execution feedback dataset and combined with the analysis results of the trajectory offset evolution trend by the AI ​​dynamic adjustment model, the trajectory offset historical cumulative calculation method is used to obtain the trajectory offset cumulative parameters. S53. Based on the user interaction operation information generated by the activity objects in the trip execution feedback dataset during the trip execution process, and combined with the analysis results of the user's active intervention behavior characteristics by the AI ​​dynamic adjustment model, the user operation intervention frequency parameter is obtained by using the operation count method.

[0014] Furthermore, step five also includes: S54. After dimensionless processing of the obtained trip execution time deviation parameters, trajectory offset cumulative parameters, and user operation intervention frequency parameters, calculate and obtain the trip execution deviation feedback coefficient. S55. By setting a preset travel execution deviation threshold, and comparing and analyzing the travel execution deviation feedback coefficient with the travel execution deviation threshold, the third evaluation result is obtained, including: When the travel execution deviation feedback coefficient is less than or equal to the travel execution deviation threshold, it indicates that the current travel execution deviation is within a controllable range and will be continuously monitored. When the trip execution deviation feedback coefficient exceeds the trip execution deviation threshold, it indicates that the current trip execution deviation is uncontrollable, posing a risk of trip planning becoming disconnected from actual execution, user experience deterioration, or trip interruption. This triggers a third warning instruction and generates a third strategy: based on real-time collected trip execution feedback data and combined with the analysis results of trip deviation characteristics from the trip feature extraction model, the AI ​​dynamic adjustment model is adaptively corrected. This includes: reducing the continuous travel intensity of subsequent trip segments to match actual travel capabilities; adjusting path selection rules to reduce path types that deviate from the actual trajectory; optimizing environmental constraint matching conditions to reduce the probability of triggering user intervention; after the correction is completed, the trip plan is regenerated, and the trip execution deviation feedback coefficient is continuously calculated during subsequent execution to form a closed-loop optimization.

[0015] Furthermore, a dynamic generation system for outdoor activity itineraries based on multi-source heterogeneous data includes: The multi-source trip status acquisition module is used to monitor the activity objects in real time during hiking, mountaineering and camping outdoor activities, and to collect the location and trajectory data of the activity objects, activity behavior and movement characteristic data, activity environment and path constraint data, and trip execution feedback data. The trip feature modeling module is used to train multi-source heterogeneous trip data using convolutional neural networks to build a trip feature extraction model; based on the intermediate layer feature vectors, it characterizes travel preferences, path selection and environmental adaptation characteristics, and integrates trip execution feedback for enhancement, building an AI dynamic adjustment model; The trip continuity assessment module is used to calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing the risky path, and calling the AI ​​dynamic adjustment model to replan the trip are adopted. The trip environment risk assessment module is used to calculate the trip environment risk coupling coefficient and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If it is not, it will provide strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. The trip execution deviation feedback adjustment module is used to calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If it is not, an adaptive correction strategy for trip intensity, path selection and environmental matching is given.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention performs phased and quantitative evaluation of the trip execution process based on the construction of a continuous adaptation coefficient, an environmental risk coupling coefficient, and a trip execution deviation feedback coefficient. It also triggers targeted adjustment strategies in unqualified states, enabling trip planning to be dynamically updated according to the status of the activity object, path conditions, and execution deviations. This effectively avoids the problems of unstable travel, path deviation, and trip interruption caused by traditional static planning.

[0017] This invention also integrates multi-source heterogeneous data such as BeiDou positioning trajectory data, activity behavior characteristics, environmental constraint information, and trip execution feedback for unified modeling, and uses convolutional neural networks to extract trip feature vectors. This can accurately characterize the travel ability, path preference, and environmental adaptability of the activity object, and realize the transformation of trip planning from static recommendation based on experience rules to dynamic decision-making based on BeiDou positioning data and intelligent models, significantly improving the personalization and intelligence level of trip planning.

[0018] This invention also introduces an execution deviation feedback coefficient during the trip execution phase to quantify the deviation between the actual execution and the planned results, and adaptively corrects subsequent trips based on the feedback results, forming a closed-loop optimization mechanism of "monitoring-evaluation-adjustment-re-evaluation", thereby continuously improving trip feasibility, safety and overall user experience. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall method steps of the present invention; Figure 2 This is a schematic diagram of the overall system flow of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0022] Example 1 Please see Figure 1 This invention provides a technical solution: a method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data, the specific steps of which include: Step 1: By monitoring the participants in hiking, mountaineering and camping outdoor activities in real time, collect data on the location and trajectory of the participants, their behavior and movement characteristics, the environment and path constraints, and the feedback data on the execution of the trip. Step 2: Train a multi-source heterogeneous travel data using a convolutional neural network to build a travel feature extraction model; characterize travel preferences, path selection, and environmental adaptation based on intermediate layer feature vectors, and enhance these characteristics by integrating travel execution feedback to build an AI dynamic adjustment model; Step 3: Calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing risky paths, and calling AI to dynamically adjust the model to replan the trip will be adopted. Step 4: Calculate the coupling coefficient of the trip environment risk and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If not, implement strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. Step 5: Calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If not, provide an adaptive correction strategy for trip intensity, path selection, and environmental matching.

[0023] In this embodiment, the outdoor activity itinerary is evaluated in three levels: continuity adaptability, environmental risk, and execution deviation. When the threshold conditions are not met, the AI ​​is linked to dynamically adjust the model to implement targeted itinerary corrections. This achieves real-time closed-loop optimization of the itinerary planning and actual execution process, thereby effectively improving the safety, feasibility, and stability of the itinerary and reducing the risk of itinerary interruption and experience degradation.

[0024] Example 2 Please see Figure 1 In this embodiment, as explained in Embodiment 1, specifically, step one includes: S11. Real-time monitoring of the spatial location and trajectory status of participants during hiking, mountaineering, and camping outdoor activities. By configuring satellite positioning equipment on the participants' terminals and using BeiDou satellite navigation and positioning technology, the real-time location coordinates, positioning update timestamps, and continuous trajectory point sequences of the participants are obtained; and a dataset of the participants' location and trajectory is established. S12. Real-time monitoring of the behavior and movement characteristics of participants during outdoor activities. Through the accelerometer and angular velocity sensor built into the terminal device, the movement speed, direction of movement, and duration of stay of the participants are continuously collected. Combined with historical activity records in the cloud, the path selection behavior and historical trajectory characteristics of the participants in similar outdoor activities are obtained. A dataset of activity behavior and movement characteristics is established. S13. Monitor the environmental conditions and path conditions of the outdoor activity area in real time. Obtain weather information in the activity area through the external environment data interface, and obtain terrain slope information, path accessibility information and environmental risk warning information corresponding to the activity area through the path status interface; establish an activity environment and path constraint dataset. S14. Monitor the actual execution process of outdoor activities in real time. By comparing and analyzing the real-time trajectory of the activity subject obtained based on Beidou satellite positioning technology with the current planned route, obtain trajectory offset data; synchronously collect the deviation data between the actual execution time and the expected duration of the trip, as well as the user interaction operation information generated by the activity subject during the trip execution; and establish a trip execution feedback dataset.

[0025] In this embodiment, by collecting and modeling the location trajectory, movement characteristics, environmental status, and itinerary execution feedback of the activity object in a multi-dimensional and synchronous manner, a complete data foundation for the entire process of outdoor activities is constructed. This enables itinerary planning to no longer rely on single location or static information, but to truly reflect the behavior, environmental changes, and execution deviations, providing highly reliable and multi-source data support for subsequent itinerary evaluation and dynamic adjustment.

[0026] Example 3 Please see Figure 1 In the explanation of Example 2, this embodiment specifically includes the following steps: S21. Using convolutional neural networks, construct an initial convolutional neural network model for outdoor activity itineraries. Use the activity object location and trajectory dataset, activity behavior and movement feature dataset, activity environment and path constraint dataset, and itinerary execution feedback dataset as model inputs to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as the itinerary feature extraction model. S22. Based on the trip feature extraction model, extract the intermediate layer output as the trip feature vector, analyze the travel preferences, path selection characteristics and environmental adaptation characteristics of the activity object; and enhance the trip feature vector by combining real-time trajectory offset data, trip execution time deviation data and user interaction operation information. S23. Based on the enhanced trip feature vector, train and test the trip feature extraction model. Use the trained trip feature extraction model as the AI ​​dynamic adjustment model to dynamically generate and predict outdoor activity trips, and output trip planning results that match the current activity status and environmental constraints.

[0027] In this embodiment, a convolutional neural network is introduced to uniformly model multi-source heterogeneous travel data, and intermediate layer feature vectors are used to characterize travel preferences, path selection and environmental adaptation characteristics. Then, feature enhancement is performed by combining real-time execution feedback, so that the generated travel plan can dynamically fit the real behavior and environmental changes of the activity object, thereby significantly improving the accuracy and personalized adaptation level of travel prediction and adjustment.

[0028] Example 4 Please see Figure 1 In the explanation of Example 3, this embodiment specifically includes the following steps: S31. Based on the real-time location coordinates and positioning update timestamps of the active objects in the active object location and trajectory dataset, and combined with the analysis results of trajectory time series features by the travel feature extraction model, the velocity change rate calculation method of adjacent sampling points is used to quantify the velocity fluctuation of the active object during its movement, and obtain the movement speed stability parameter, denoted as Sv. S32. Based on the continuous trajectory point sequence in the active object location and trajectory dataset, and combined with the statistical analysis results of the path direction change characteristics by the travel feature extraction model, the direction offset angle distribution statistical method is used to obtain the travel direction consistency parameter, denoted as Sd. S33. Based on the real-time trajectory offset data in the trip execution feedback dataset, and combined with the fusion analysis results of the path execution deviation features by the trip feature extraction model, the trajectory-path spatial deviation calculation method is used to obtain the path offset parameter, denoted as Sp.

[0029] In this embodiment, by introducing the travel speed stability parameter Sv, the travel direction consistency parameter Sd, and the path deviation parameter Sp in step three, the speed fluctuation, direction change, and path deviation of the active object during the journey execution process are evaluated in a multi-dimensional quantitative manner. This enables a fine characterization of the stability of the travel state and the reliability of the path execution, thereby more accurately identifying abnormal behavior and providing reliable data for subsequent dynamic adjustment and optimization decisions.

[0030] Example 5 Please see Figure 1 In the explanation of Example 4, specifically, step three further includes: S34. After obtaining the travel speed stability parameter Sv, travel direction consistency parameter Sd, and path offset parameter Sp, the travel continuity adaptation coefficient, denoted as XLS, is calculated after dimensionless processing, as follows: In the formula, w1, w2, and w3 represent weighting coefficients; The stability of travel speed has a high weighting on its impact on the continuity of the journey. The continuous stability of travel speed directly reflects the physical exertion level and rhythm matching degree of the participant, and is a key indicator for judging whether the journey may be interrupted or whether there is a risk of fatigue. The second highest weighting is given to the impact of consistency of travel direction on the continuity of the journey; consistency of direction reflects the ability of the activity object to execute the given path and the degree of navigation adaptation, and is highly correlated with the risk of getting lost and the behavior of frequently adjusting the path. The path offset characterizes the impact of travel continuity adaptability and has a relatively low weight. Path offset reflects local environmental disturbances or temporary adjustment behaviors, and its impact on overall continuity adaptability is relatively indirect, so it is given a low weight. By weighted fusion of speed stability, directional consistency, and path deviation, the level of continuous adaptation of the journey can be comprehensively evaluated from the dual perspectives of motion rhythm and path execution, providing a quantitative basis for journey shortening, path replacement, and dynamic replanning.

[0031] S35. By setting a preset travel continuity adaptation threshold, denoted as Xth, and comparing the travel continuity adaptation coefficient XLS with the travel continuity adaptation threshold Xth, the first evaluation result is obtained, including: When the trip continuity adaptation coefficient XLS ≥ the trip continuity adaptation threshold Xth, it indicates that the current trip continuity adaptation is qualified and there is no risk of trip interruption. Continuous monitoring is required. When the trip continuity adaptation coefficient XLS < the trip continuity adaptation threshold Xth, it indicates that the current trip continuity adaptation is unqualified, and there is a risk of unstable travel and directional deviation, leading to trip interruption or decreased experience. This triggers the first warning instruction and generates the first strategy: shorten the distance of subsequent trip segments to reduce the burden of continuous travel; replace trip segments with path deviation risk and select alternative paths; after completing the trip segment adjustment, regenerate subsequent trip planning segments based on the AI ​​dynamic adjustment model to match them with the current activity state and environmental constraints; recalculate until the trip continuity adaptation coefficient XLS ≥ the trip continuity adaptation threshold Xth.

[0032] The method for obtaining the trip continuity adaptation threshold Xth is as follows: Statistical analysis of historical operational data from a large number of hiking, mountaineering, and camping outdoor activities is conducted to extract the distribution range of the trip continuity adaptation coefficient between states of good continuity adaptation and states of decreased adaptation such as unstable movement or directional deviation. This is combined with the identification results of features such as speed stability, direction consistency, and path deviation from the trip feature extraction model, and further considered in conjunction with outdoor activity safety guidelines and the experience of professional outdoor planners, to determine a reasonable trip continuity adaptation judgment threshold. This threshold is used to effectively distinguish between a state of satisfactory trip continuity adaptation and a state with the risk of trip interruption or decreased experience.

[0033] In this embodiment, by constructing a trip continuity adaptation coefficient XLS and weighting and fusing the stability of travel speed, consistency of direction, and degree of path deviation with threshold determination, the trip continuity adaptation level can be quantified in real time during trip execution. When the adaptation is insufficient, an early warning is automatically triggered and the trip segment length and path structure are dynamically adjusted, thereby effectively reducing the risk of trip interruption and improving the stability, continuity, and overall experience quality of trip execution.

[0034] Example 6 Please see Figure 1 In the explanation of Example 5, specifically, step four includes: S41. Based on the weather status information and environmental risk warning information in the activity environment and path constraint dataset, and combined with the analysis results of environmental adaptability characteristics by the itinerary feature extraction model, the weighted risk scoring method is used to quantify the environmental risk intensity of the current activity area and obtain the environmental risk intensity parameter, denoted as Re; S42. Based on the terrain slope information and path traffic status information in the activity environment and path constraint dataset, and combined with the statistical analysis results of terrain adaptation characteristics by the travel feature extraction model, the terrain complexity parameter, denoted as Rt, is obtained by using the comprehensive index calculation method of slope and undulation. S43. Based on the trajectory offset data, the deviation data between the actual execution time and the expected trip time in the trip execution feedback dataset, and the user interaction operation information generated during the trip execution, and combined with the trip feature extraction model to analyze the trajectory offset evolution characteristics, execution rhythm change characteristics, and user active intervention behavior characteristics, the cumulative counting and weighted statistical method of execution deviation is used to calculate the cumulative number of trajectory offset exceeding the limit, the number of execution time deviation exceeding the limit, and the number of user active intervention operations within a unit trip, and obtain the risk accumulation parameter, denoted as Rr.

[0035] In this embodiment, by quantifying the intensity of environmental risks, the complexity of terrain, and the accumulation of risks during the trip, and by integrating and analyzing multi-source environmental constraints and execution feedback information, the impact of the external environment and actual execution on the safety and controllability of the trip can be comprehensively and dynamically depicted during the trip. This provides a reliable basis for subsequent risk assessment and strategy adjustment, thereby effectively reducing trip interruptions and safety hazards caused by sudden environmental factors.

[0036] Example 7 Please see Figure 1 In the explanation of Example Six, specifically, step four further includes: S44. After obtaining the environmental risk intensity parameter Re, terrain complexity parameter Rt, and risk accumulation parameter Rr, and performing dimensionless processing, calculate the trip environmental risk coupling coefficient, denoted as HOX, as follows: In the formula, a1, a2, and a3 represent weighting coefficients; The environmental risk intensity characterizes the impact of the trip's environmental risk and has the highest weight; sudden weather changes, environmental warnings, and other factors are often sudden and irreversible, and are the primary external causes of trip safety accidents. The complexity of terrain represents the impact of the trip's environmental risks and has a medium weight; the slope and undulation of the terrain directly affect the difficulty of the journey and the physical exertion, and are long-term fundamental factors of environmental risks. This represents the impact of risk accumulation on the environmental risk of the trip, and has a secondary weight; risk accumulation reflects the cumulative effect of multiple small deviations or interventions, and its impact on environmental risk has a lag. By constructing a weighted coupling model of environmental risk intensity, terrain complexity, and risk accumulation, we can comprehensively characterize the combined impact of environmental factors on trip safety and provide quantitative triggering conditions for route replacement and time window adjustment.

[0037] S45. By setting a preset travel environment risk threshold, denoted as Hth, and comparing the travel environment risk coupling coefficient HOX with the travel environment risk threshold Hth, the second assessment results are obtained, including: When the trip environment risk coupling coefficient HOX ≤ the trip environment risk threshold Hth, it means that the current trip environment risk is within an acceptable range, and the current trip can continue to be executed and continuously monitored. When the travel environment risk coupling coefficient HOX is greater than the travel environment risk threshold Hth, it indicates that the current travel environment risk exceeds the standard, and there is a risk of activity interruption or safety. This triggers a second warning instruction and generates a second strategy: replace the routes of the risky travel segments; dynamically adjust the travel time window to avoid the risky period; after adjustment, call the AI ​​dynamic adjustment model to generate updated travel planning results, and recalculate after completing the travel segment adjustment until the travel environment risk coupling coefficient HOX is less than or equal to the travel environment risk threshold Hth.

[0038] The method for obtaining the trip environmental risk threshold Hth is as follows: Statistical analysis is conducted on historical environmental data and trip execution data of a large number of outdoor activities under different weather conditions, terrain complexity, and risk exposure levels. The range of variation of the trip environmental risk coupling coefficient under controllable environmental risk and under conditions of safety risk or trip interruption risk is extracted. Combined with the analysis results of environmental adaptability and risk accumulation characteristics from the trip feature extraction model, and with reference to outdoor activity safety regulations, risk warning standards, and the experience judgment of professionals, a reasonable trip environmental risk judgment threshold is determined. This threshold is used to distinguish between acceptable and unsafe trip environmental risk conditions.

[0039] In this embodiment, by dimensionlessly fusing environmental risk intensity, terrain complexity, and risk accumulation, a trip environmental risk coupling coefficient is constructed. Based on threshold comparison, hierarchical assessment and linkage decision-making are realized. Before the risk exceeds the standard, dynamic adjustment of the path and time window can be proactively triggered, enabling trip planning to adaptively optimize in real time with environmental changes, thereby significantly improving the safety, continuity, and overall feasibility of outdoor activity trips.

[0040] Example 8 Please see Figure 1 In the explanation of Example 7, specifically, step five includes: S51. Based on the deviation data between the actual execution time and the expected execution time of the trip in the trip execution feedback dataset, and combined with the analysis results of the trip rhythm characteristics by the AI ​​dynamic adjustment model, the trip execution time deviation parameter is obtained by using the time difference calculation method, denoted as Dt. S52. Based on the trajectory offset data in the trip execution feedback dataset, and combined with the analysis results of the trajectory offset evolution trend by the AI ​​dynamic adjustment model, the trajectory offset historical cumulative calculation method is used to obtain the trajectory offset cumulative parameter, denoted as Ds. S53. Based on the user interaction operation information generated by the activity object in the trip execution feedback dataset during the trip execution process, and combined with the analysis results of the user's active intervention behavior characteristics by the AI ​​dynamic adjustment model, the user operation intervention frequency parameter is obtained by using the operation number statistics method, denoted as Do.

[0041] In this embodiment, by quantifying the deviation in trip execution time, the cumulative degree of trajectory deviation, and the frequency of user intervention, and combining the AI ​​dynamic adjustment model with a comprehensive analysis of the trip rhythm and execution status, the deviations and unstable factors in the trip execution process can be fully characterized, providing accurate and quantifiable feedback for subsequent trip adjustments, thereby improving the pertinence and effectiveness of trip dynamic adjustments.

[0042] Example 9 Please see Figure 1 In the explanation of Embodiment Eight, specifically, step five further includes: S54. After dimensionless processing of the obtained trip execution time deviation parameter Dt, trajectory offset cumulative parameter Ds, and user operation intervention frequency parameter Do, the trip execution deviation feedback coefficient, denoted as ZPF, is calculated, as follows: In the formula, s1, s2 and s3 represent weighting coefficients; The time deviation characterizes the impact of trip execution time deviation on execution deviation feedback and has a high weight; the time deviation directly reflects the mismatch between the trip rhythm and actual capacity, and is an important indicator of planning failure. The cumulative trajectory offset characterizes the impact of execution deviation feedback, accounting for the second highest weight; the continuous accumulation of trajectory offset reflects the insufficient adaptation of path selection to the real environment, which has a significant impact on trip quality and safety. This represents the impact of the frequency of user intervention on the feedback of execution deviation, and has a relatively low weight. User intervention is mostly a subjective corrective behavior, which can reflect the downward trend of experience, but its impact is relatively indirect. By weighted fusion of time deviation, spatial offset, and user intervention behavior, a closed-loop feedback indicator reflecting the trip execution status can be formed, providing a quantitative basis for the adaptive correction of the AI ​​dynamic adjustment model.

[0043] S55. By setting a preset travel execution deviation threshold, denoted as Zth, and comparing the travel execution deviation feedback coefficient ZPF with the travel execution deviation threshold Zth, the third evaluation results are obtained, including: When the travel execution deviation feedback coefficient ZPF ≤ travel execution deviation threshold Zth, it indicates that the current travel execution deviation is within a controllable range and will be continuously monitored. When the trip execution deviation feedback coefficient ZPF > the trip execution deviation threshold Zth, it indicates that the current trip execution deviation is uncontrollable, and there is a risk of trip planning being disconnected from actual execution, user experience deterioration, or trip interruption. This triggers a third warning instruction and generates a third strategy: based on real-time collected trip execution feedback data and combined with the analysis results of trip deviation characteristics from the trip feature extraction model, the AI ​​dynamic adjustment model is adaptively corrected, including: reducing the continuous travel intensity of subsequent trip segments to match actual travel capacity; adjusting path selection rules to reduce the types of paths that deviate from the actual trajectory; optimizing environmental constraint matching conditions to reduce the probability of triggering user intervention; after the correction is completed, the trip plan is regenerated, and the trip execution deviation feedback coefficient ZPF is continuously calculated during subsequent execution to form a closed-loop optimization.

[0044] The method for obtaining the trip execution deviation threshold Zth is as follows: Statistical analysis is conducted on a large amount of data on actual outdoor activity trip execution, including execution time deviations, cumulative trajectory offsets, and user intervention behaviors. The distribution range of the trip execution deviation feedback coefficient is extracted between a controllable state of trip execution deviation and a state where there is a significant disconnect between planned and executed trips. This is combined with the AI ​​dynamic adjustment model's identification results of trip rhythm changes and user intervention characteristics, and with reference to relevant empirical rules for outdoor trip planning and execution evaluation, to determine a reasonable trip execution deviation judgment threshold. This threshold is used to effectively distinguish between a state where trip execution deviation is within a controllable range and a state that requires triggering adaptive adjustment.

[0045] In this embodiment, by constructing a trip execution deviation feedback coefficient and comparing it with a threshold to trigger a graded early warning and adaptive correction mechanism, the risk of a disconnect between planning and actual execution can be identified in time before the trip execution deviation becomes uncontrollable. Based on real-time feedback, the travel intensity, path selection and environmental constraints are dynamically optimized, thereby forming a closed-loop adjustment, which significantly improves the stability, continuity and user experience of trip execution.

[0046] Example 10 Please refer to the following: A system for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data. Figure 2 Specifically, including: The multi-source trip status acquisition module is used to monitor the activity objects in real time during hiking, mountaineering and camping outdoor activities, and to collect the location and trajectory data of the activity objects, activity behavior and movement characteristic data, activity environment and path constraint data, and trip execution feedback data. The trip feature modeling module is used to train multi-source heterogeneous trip data using convolutional neural networks to build a trip feature extraction model; based on the intermediate layer feature vectors, it characterizes travel preferences, path selection and environmental adaptation characteristics, and integrates trip execution feedback for enhancement, building an AI dynamic adjustment model; The trip continuity assessment module is used to calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing the risky path, and calling the AI ​​dynamic adjustment model to replan the trip are adopted. The trip environment risk assessment module is used to calculate the trip environment risk coupling coefficient and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If it is not, it will provide strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. The trip execution deviation feedback adjustment module is used to calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If it is not, an adaptive correction strategy for trip intensity, path selection and environmental matching is given.

[0047] In this embodiment, by decoupling and coordinating the functions of the multi-source trip status acquisition module, trip feature modeling module, trip continuous adaptation evaluation module, trip environment risk assessment module, and trip execution deviation feedback adjustment module, a hierarchical and interconnected trip management system structure is formed. This enables the trip perception, evaluation, and adjustment to achieve modular operation and closed-loop control at the system level, avoiding response lag and strategy fragmentation caused by independent judgment of a single module, and improving the real-time performance, stability, and overall system reliability of trip planning adjustments.

[0048] It should be noted that all calculation formulas in this application employ regression analysis, including but not limited to machine learning algorithms, to deeply analyze the collected parameters and identify their natural trends and interrelationships. Specialized software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Then, cross-validation and other methods are used to objectively evaluate the model performance, and continuous feedback and optimization are combined to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula undergo dimensionless processing within a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless processing techniques include, but are not limited to, min-max-normalization and Z-score standardization. The algorithm of this invention is implemented as a Python script. Before executing the core logic, the program first executes a data loading module (e.g., using the widely used pandas library in Python) configured to read the aforementioned spreadsheet file and load its contents into the program's working memory (e.g., a DataFrame data structure). Subsequent algorithm steps will directly query and retrieve the required configuration parameters from this in-memory data structure.

[0049] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data, characterized in that, The specific steps include: Step 1: By monitoring the participants in hiking, mountaineering and camping outdoor activities in real time, collect data on the location and trajectory of the participants, their behavior and movement characteristics, the environment and path constraints, and the feedback data on the execution of the trip. Step 2: Train a multi-source heterogeneous travel data using a convolutional neural network to build a travel feature extraction model; characterize travel preferences, path selection, and environmental adaptation based on intermediate layer feature vectors, and enhance these characteristics by integrating travel execution feedback to build an AI dynamic adjustment model; Step 3: Calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing risky paths, and calling AI to dynamically adjust the model to replan the trip will be adopted. Step 4: Calculate the coupling coefficient of the trip environment risk and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If not, implement strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. Step 5: Calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If not, provide an adaptive correction strategy for trip intensity, path selection, and environmental matching.

2. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 1, characterized in that: Step one includes: S11. Real-time monitoring of the spatial location and trajectory status of participants during hiking, mountaineering, and camping outdoor activities. By configuring satellite positioning equipment on the participants' terminals and using BeiDou satellite navigation and positioning technology, the real-time location coordinates, positioning update timestamps, and continuous trajectory point sequences of the participants are obtained; and a dataset of the participants' location and trajectory is established. S12. Real-time monitoring of the behavior and movement characteristics of participants during outdoor activities. Through the accelerometer and angular velocity sensor built into the terminal device, the movement speed, direction of movement, and duration of stay of the participants are continuously collected. Combined with historical activity records in the cloud, the path selection behavior and historical trajectory characteristics of the participants in similar outdoor activities are obtained. A dataset of activity behavior and movement characteristics is established. S13. Monitor the environmental conditions and path conditions of the outdoor activity area in real time. Obtain weather information in the activity area through the external environment data interface, and obtain terrain slope information, path accessibility information and environmental risk warning information corresponding to the activity area through the path status interface; establish an activity environment and path constraint dataset. S14. Monitor the actual execution process of outdoor activities in real time. By comparing and analyzing the real-time trajectory of the activity subject obtained based on Beidou satellite positioning technology with the current planned route, obtain trajectory offset data; synchronously collect the deviation data between the actual execution time and the expected duration of the trip, as well as the user interaction operation information generated by the activity subject during the trip execution; and establish a trip execution feedback dataset.

3. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 2, characterized in that: Step two includes: S21. Using convolutional neural networks, construct an initial convolutional neural network model for outdoor activity itineraries. Use the activity object location and trajectory dataset, activity behavior and movement feature dataset, activity environment and path constraint dataset, and itinerary execution feedback dataset as model inputs to train and test the initial convolutional neural network model. Use the trained initial convolutional neural network model as the itinerary feature extraction model. S22. Based on the trip feature extraction model, extract the intermediate layer output as the trip feature vector, analyze the travel preferences, path selection characteristics and environmental adaptation characteristics of the activity object; and enhance the trip feature vector by combining real-time trajectory offset data, trip execution time deviation data and user interaction operation information. S23. Based on the enhanced trip feature vector, train and test the trip feature extraction model. Use the trained trip feature extraction model as the AI ​​dynamic adjustment model to dynamically generate and predict outdoor activity trips, and output trip planning results that match the current activity status and environmental constraints.

4. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 3, characterized in that: Step three includes: S31. Based on the real-time location coordinates and positioning update timestamps of the active objects in the active object location and trajectory dataset, and combined with the analysis results of trajectory time series features by the travel feature extraction model, the speed change rate calculation method of adjacent sampling points is used to quantify the speed fluctuation of the active object during its movement and obtain the movement speed stability parameter. S32. Based on the continuous trajectory point sequence in the active object location and trajectory dataset, and combined with the statistical analysis results of the path direction change characteristics by the travel feature extraction model, the direction offset angle distribution statistical method is used to obtain the travel direction consistency parameter. S33. Based on the real-time trajectory offset data in the trip execution feedback dataset, and combined with the fusion analysis results of the path execution deviation features by the trip feature extraction model, the trajectory-path spatial deviation calculation method is used to obtain the path offset parameter.

5. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 4, characterized in that: Step three also includes: S34. After dimensionless processing of the obtained travel speed stability parameters, travel direction consistency parameters, and path offset parameters, the travel continuity adaptation coefficient is calculated. S35. By setting a preset travel continuity adaptation threshold and comparing the travel continuity adaptation coefficient with the travel continuity adaptation threshold, the first evaluation result is obtained, including: When the trip continuity adaptation coefficient is greater than or equal to the trip continuity adaptation threshold, it indicates that the current trip continuity adaptation is qualified and there is no risk of trip interruption. Continuous monitoring is required. When the trip continuity adaptation coefficient is less than the trip continuity adaptation threshold, it indicates that the current trip continuity adaptation is unqualified, and there is a risk of unstable travel and directional deviation, which may lead to trip interruption or decreased experience. This triggers the first warning instruction and generates the first strategy: shorten the distance of subsequent trip segments to reduce the burden of continuous travel; replace trip segments with path deviation risk and select alternative paths; after completing the trip segment adjustment, regenerate subsequent trip planning segments based on the AI ​​dynamic adjustment model to match them with the current activity status and environmental constraints; recalculate until the trip continuity adaptation coefficient is greater than or equal to the trip continuity adaptation threshold.

6. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 5, characterized in that: Step four includes: S41. Based on the weather status information and environmental risk warning information in the activity environment and path constraint dataset, and combined with the analysis results of environmental adaptability characteristics by the itinerary feature extraction model, the weighted risk scoring method is used to quantify the environmental risk intensity of the current activity area and obtain the environmental risk intensity parameter. S42. Based on the terrain slope information and path traffic status information in the activity environment and path constraint dataset, and combined with the statistical analysis results of terrain adaptation characteristics by the travel feature extraction model, the terrain complexity parameter is obtained by using the comprehensive index calculation method of slope and undulation. S43. Based on the trajectory offset data, the deviation data between the actual execution time and the expected trip time in the trip execution feedback dataset, and the user interaction operation information generated during the trip execution, and combined with the trip feature extraction model to analyze the trajectory offset evolution characteristics, execution rhythm change characteristics, and user active intervention behavior characteristics, the cumulative counting and weighted statistical method of execution deviation is used to calculate the cumulative number of trajectory offset exceeding the limit, the number of execution time deviation exceeding the limit, and the number of user active intervention operations within a unit trip, and obtain the risk accumulation parameter.

7. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 6, characterized in that: Step four also includes: S44. After dimensionless processing of the obtained environmental risk intensity parameters, terrain complexity parameters, and risk accumulation parameters, the travel environment risk coupling coefficient is calculated. S45. By setting a preset travel environment risk threshold and comparing the travel environment risk coupling coefficient with the travel environment risk threshold, the second assessment result is obtained, including: When the coupling coefficient of the trip environment risk is less than or equal to the trip environment risk threshold, it means that the current trip environment risk is within an acceptable range, and the current trip can continue to be executed while continuous monitoring continues. When the travel environment risk coupling coefficient is greater than the travel environment risk threshold, it indicates that the current travel environment risk exceeds the standard, and there is a risk of activity interruption or safety. This triggers a second warning instruction and generates a second strategy: replace the routes of the risky travel segments; dynamically adjust the travel time window to avoid risky periods; after adjustment, call the AI ​​dynamic adjustment model to generate updated travel planning results, and recalculate after completing the travel segment adjustment until the travel environment risk coupling coefficient is less than or equal to the travel environment risk threshold.

8. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 7, characterized in that: Step five includes: S51. Based on the deviation data between the actual execution time and the expected execution time of the trip in the trip execution feedback dataset, and combined with the analysis results of the trip rhythm characteristics by the AI ​​dynamic adjustment model, the time difference calculation method is used to obtain the trip execution time deviation parameter. S52. Based on the trajectory offset data in the trip execution feedback dataset and combined with the analysis results of the trajectory offset evolution trend by the AI ​​dynamic adjustment model, the trajectory offset historical cumulative calculation method is used to obtain the trajectory offset cumulative parameters. S53. Based on the user interaction operation information generated by the activity objects in the trip execution feedback dataset during the trip execution process, and combined with the analysis results of the user's active intervention behavior characteristics by the AI ​​dynamic adjustment model, the user operation intervention frequency parameter is obtained by using the operation count method.

9. The method for dynamically generating outdoor activity itineraries based on multi-source heterogeneous data according to claim 8, characterized in that: Step five also includes: S54. After dimensionless processing of the obtained trip execution time deviation parameters, trajectory offset cumulative parameters, and user operation intervention frequency parameters, calculate and obtain the trip execution deviation feedback coefficient. S55. By setting a preset travel execution deviation threshold, and comparing and analyzing the travel execution deviation feedback coefficient with the travel execution deviation threshold, the third evaluation result is obtained, including: When the travel execution deviation feedback coefficient is less than or equal to the travel execution deviation threshold, it indicates that the current travel execution deviation is within a controllable range and will be continuously monitored. When the trip execution deviation feedback coefficient exceeds the trip execution deviation threshold, it indicates that the current trip execution deviation is uncontrollable, posing a risk of trip planning becoming disconnected from actual execution, user experience deterioration, or trip interruption. This triggers a third warning instruction and generates a third strategy: based on real-time collected trip execution feedback data and combined with the analysis results of trip deviation characteristics from the trip feature extraction model, the AI ​​dynamic adjustment model is adaptively corrected. This includes: reducing the continuous travel intensity of subsequent trip segments to match actual travel capabilities; adjusting path selection rules to reduce path types that deviate from the actual trajectory; optimizing environmental constraint matching conditions to reduce the probability of triggering user intervention; after the correction is completed, the trip plan is regenerated, and the trip execution deviation feedback coefficient is continuously calculated during subsequent execution to form a closed-loop optimization.

10. A dynamic generation system for outdoor activity itineraries based on multi-source heterogeneous data, applied to the dynamic generation method for outdoor activity itineraries based on multi-source heterogeneous data as described in any one of claims 1 to 9, characterized in that, include: The multi-source trip status acquisition module is used to monitor the activity objects in real time during hiking, mountaineering and camping outdoor activities, and to collect the location and trajectory data of the activity objects, activity behavior and movement characteristic data, activity environment and path constraint data, and trip execution feedback data. The trip feature modeling module is used to train multi-source heterogeneous trip data using convolutional neural networks to build a trip feature extraction model; The model is constructed by characterizing travel preferences, path selection and environmental adaptation based on intermediate layer feature vectors and enhancing them by integrating trip execution feedback. The trip continuity assessment module is used to calculate the trip continuity adaptation coefficient and compare it with the trip continuity adaptation threshold to determine whether the current trip continuity adaptation is qualified. If it is not qualified, strategies such as shortening the trip, replacing the risky path, and calling the AI ​​dynamic adjustment model to replan the trip are adopted. The trip environment risk assessment module is used to calculate the trip environment risk coupling coefficient and compare it with the trip environment risk threshold to determine whether the current trip environment risk is within an acceptable range. If it is not, it will provide strategies such as route replacement, time window adjustment, and AI dynamic adjustment model to repeatedly replan the trip. The trip execution deviation feedback adjustment module is used to calculate the trip execution deviation feedback coefficient and compare it with the trip execution deviation threshold to determine whether the current trip execution deviation is within a controllable range. If it is not, an adaptive correction strategy for trip intensity, path selection and environmental matching is given.

Images