A Method and System for Intelligent Human-Computer Interaction of Ebikes Based on Bluetooth Mesh Networking

Abstract

This invention proposes an intelligent human-computer interaction method and system for e-bikes based on Bluetooth Mesh networking. It belongs to the field of intelligent transportation and the Internet of Things (IoT) technology. The method includes: acquiring Bluetooth communication node information of various functional modules of the e-bike and the user's smart device; performing Mesh network topology analysis on the acquired Bluetooth communication node information to construct a Bluetooth Mesh network topology for the e-bike usage scenario; configuring and setting parameters for each node according to the constructed Bluetooth Mesh network topology to form a complete e-bike Bluetooth Mesh communication network; and connecting the functional modules of the e-bike and the user's smart device into an efficient and stable communication network through Bluetooth Mesh networking technology. This network has self-organizing capabilities and can automatically adjust the topology based on the addition or removal of nodes, ensuring communication stability and real-time performance, thereby significantly improving the overall system response speed and interaction efficiency.

Description

Technical Field

[0001] This invention proposes an intelligent human-computer interaction method and system for ebikes based on Bluetooth Mesh networking, belonging to the field of intelligent transportation and Internet of Things technology. Background Technology

[0002] With increasing public awareness of healthy and green travel, ebikes (electric bicycles) have become widely used as a convenient and environmentally friendly mode of transportation. However, traditional ebikes suffer from several shortcomings in human-computer interaction, such as limited interaction methods, limited information display, and the inability to achieve multi-device collaborative control. Bluetooth Mesh networking technology, with its advantages of low power consumption, self-organizing network, and multi-node communication, can effectively improve the user experience and intelligence level of ebikes when applied to intelligent human-computer interaction. However, there is currently no mature method for intelligent human-computer interaction of ebikes based on Bluetooth Mesh networking; therefore, it is necessary to develop a new method to address the aforementioned issues. Summary of the Invention

[0003] This invention provides an intelligent human-computer interaction method and system for ebikes based on Bluetooth Mesh networking, in order to solve the problems mentioned in the background art above:

[0004] The present invention proposes an intelligent human-computer interaction method for ebikes based on Bluetooth Mesh networking, the method comprising:

[0005] S1. Obtain Bluetooth communication node information of various functional modules of ebike and smart devices carried by users; perform Mesh network topology analysis on the obtained Bluetooth communication node information, and construct Bluetooth Mesh network topology for ebike usage scenarios; based on the constructed Bluetooth Mesh network topology, configure the network and set parameters for each node to form a complete ebike Bluetooth Mesh communication network.

[0006] S2. During the operation of ebike, data related to ebike is collected through various functional modules and smart devices. The collected data is preprocessed and classified according to different types; feature extraction is then performed on the classified data.

[0007] S3. Based on the extracted feature data, a user intent recognition model is built using machine learning algorithms; based on the recognized user intent, combined with the functions of ebike and the current usage scenario, the interaction mode is determined.

[0008] S4. After determining the interaction mode, the ebike and various smart devices will coordinate and interact through the Bluetooth Mesh network; and information from different devices will be merged and processed.

[0009] S5. Continuously optimize the user intent recognition model and interaction strategy based on actual user feedback and interaction effects; regularly evaluate the communication quality of the Bluetooth Mesh network and adjust the network parameters according to the evaluation results.

[0010] The present invention proposes an intelligent human-computer interaction system for ebike based on Bluetooth Mesh networking, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to implement the intelligent human-computer interaction method for ebike based on Bluetooth Mesh networking as described above.

[0011] The beneficial effects of this invention are as follows: By using Bluetooth Mesh networking technology, the functional modules of the e-bike (such as the motor controller, battery management system, and dashboard) are connected to the user's smart devices (such as smartphones and smartwatches) to form an efficient and stable communication network. This network has self-organizing capabilities and can automatically adjust its topology based on the addition or removal of nodes, ensuring communication stability and real-time performance, thereby significantly improving the overall system response speed and interaction efficiency.

[0012] By employing a multi-source data acquisition mechanism, multi-dimensional data, including riding status, battery status, user physiological indicators, and environmental trajectory, is obtained. Preprocessing and feature extraction are then performed to effectively remove noise and outliers, thereby improving data quality.

[0013] By training a user intent recognition model using historical data, the system can accurately predict a user's potential intentions (such as accelerating, decelerating, switching modes, viewing information, etc.) based on the current riding status, user physiological data, and environmental information, and dynamically select the most appropriate interaction method. This intelligent interaction method improves the user experience and reduces the operational burden.

[0014] After determining the interaction mode, the system enables collaborative interaction between the e-bike and multiple smart devices via a Bluetooth Mesh network. Simultaneously, it fuses information from different devices (such as GPS track and riding speed fusion, and heart rate and riding status fusion), further improving the completeness and accuracy of the information and helping to provide more personalized and precise services.

[0015] The system has a feedback-driven optimization mechanism that can continuously optimize the user intent recognition model and interaction strategy based on actual user usage and interaction effects. At the same time, it regularly evaluates the Bluetooth Mesh network communication quality and dynamically adjusts network parameters to cope with environmental changes or device aging, thereby ensuring the long-term stability, security and consistent user experience of the system.

[0016] Based on the analysis of user behavior and health data, the system can provide users with personalized riding suggestions, health guidance, energy management and other functions, which not only improves the intelligence level of e-bike, but also enhances the product's added value and market competitiveness. Attached Figure Description

[0017] Figure 1 This is a diagram illustrating the steps of the method described in this invention;

[0018] Figure 2 As described in this invention Figure 1 Detailed execution diagram of step S1. Detailed Implementation

[0019] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0020] One embodiment of the present invention, such as Figure 1 As shown, an intelligent human-computer interaction method for ebikes based on Bluetooth Mesh networking includes:

[0021] S1. Obtain Bluetooth communication node information of various functional modules of ebike and smart devices carried by users; perform Mesh network topology analysis on the obtained Bluetooth communication node information, and construct Bluetooth Mesh network topology for ebike usage scenarios; based on the constructed Bluetooth Mesh network topology, configure the network and set parameters for each node to form a complete ebike Bluetooth Mesh communication network.

[0022] S2. During ebike's operation, data related to ebike is collected through various functional modules and smart devices. The collected data is preprocessed and classified according to different types. Feature extraction is then performed on the classified data.

[0023] S3. Based on the extracted feature data, a user intent recognition model is built using machine learning algorithms; based on the recognized user intent, combined with the functions of ebike and the current usage scenario, the interaction mode is determined.

[0024] S4. After determining the interaction mode, the ebike and various smart devices will coordinate and interact through the Bluetooth Mesh network; and information from different devices will be merged and processed.

[0025] S5. Continuously optimize the user intent recognition model and interaction strategy based on actual user feedback and interaction effects; regularly evaluate the communication quality of the Bluetooth Mesh network and adjust the network parameters according to the evaluation results.

[0026] The working principle of the above technical solution is as follows: The system first collects Bluetooth communication node information from various functional modules on the e-bike (motor controller, battery management system, dashboard, etc.) and user smart devices (smartphones, smartwatches, etc.). This information forms the basis for building a Bluetooth Mesh network. The acquired node information is then analyzed to determine the Mesh network topology, taking into account factors such as the distance between nodes, communication signal strength, and the importance of node functions. The goal is to construct a topology that ensures stable and efficient communication between nodes, and this topology must have automatic adjustment capabilities to accommodate the addition or removal of nodes. Based on the constructed Bluetooth Mesh network topology, each node is configured and its parameters are set to enable communication according to the Mesh network protocol, thus forming a complete e-bike Bluetooth Mesh communication network.

[0027] During eBike operation, various functional modules and smart devices collect real-time data related to the eBike, including riding speed, motor speed, battery level, riding distance, user heart rate, and movement trajectory. This data reflects the eBike's operational status and user usage. The collected data undergoes preprocessing to remove noise and outliers, ensuring accuracy and reliability. The data is then categorized into different types for subsequent feature extraction and analysis. Feature extraction is performed on the categorized data, extracting representative and analyzable features from different data types. For example, features such as average speed, maximum speed, and speed change trends are extracted from riding speed data; remaining battery percentage and battery consumption rate are extracted from battery level data; and average heart rate and heart rate fluctuation range are extracted from user heart rate data. Through feature extraction, the raw data is transformed into a more valuable form, facilitating subsequent user intent recognition and interaction decision-making.

[0028] Based on extracted feature data, a user intent recognition model is constructed using machine learning algorithms. This model, through learning from a large amount of historical data, can accurately identify user intent based on current riding status, user physiological data, and environmental information, such as whether the user wants to accelerate, decelerate, switch riding modes, or view specific information. Based on the identified user intent, combined with the e-bike's functions and the current usage scenario, an appropriate interaction mode is determined. Interaction modes can include various methods such as voice interaction, dashboard display interaction, and mobile app interaction. For example, for different user intents, the most suitable interaction method is selected to meet user needs and improve user experience.

[0029] After determining the interaction mode, collaborative interaction between the ebike and various smart devices is achieved through a Bluetooth Mesh network. For example, when a user issues a control command via their smartphone, the command is transmitted to the corresponding functional module of the ebike through the Bluetooth Mesh network, enabling control of the ebike; simultaneously, the ebike's status information can also be fed back to the smartphone in real time for the user to view. This collaborative interaction allows users to control the ebike and obtain information through different devices. Information from different devices is fused to improve accuracy and completeness. For example, fusing riding speed information collected from the ebike's dashboard with motion trajectory information obtained from the smartphone's GPS positioning system can more accurately calculate the user's riding distance and speed distribution; fusing heart rate data collected from the user's smartwatch with the ebike's riding status data can analyze the impact of riding on the user's body and provide more personalized health advice. Information fusion can comprehensively utilize the data advantages of different devices to provide users with more comprehensive and accurate information.

[0030] Based on user feedback and interaction effects, the user intent recognition model and interaction strategies are continuously optimized. For example, if user satisfaction with a certain interaction method is low, or the recognition model's accuracy is low in certain situations, more user data and interaction records can be collected to retrain and adjust the model, thereby improving its performance and the adaptability of the interaction. The communication quality of the Bluetooth Mesh network is regularly evaluated, and network parameters are adjusted based on the evaluation results to ensure network stability and reliability. For example, if a node's communication signal is weak, its transmission power can be adjusted or the network topology optimized to ensure overall network communication quality. Continuous optimization continuously improves system performance and user experience, ensuring the system maintains good operating conditions during long-term use.

[0031] The effect of the above technical solution is that by building an efficient and stable Bluetooth Mesh network, e-bike can communicate efficiently with smart devices (such as smartphones, smartwatches, etc.) and support multiple interaction methods (voice, dashboard, APP, etc.), thereby improving the interactive experience between users and e-bike.

[0032] Leveraging the adaptive nature of Bluetooth Mesh networks, nodes can automatically adjust the network structure, reducing the need for manual debugging and maintenance. Furthermore, Mesh networks are highly scalable, easily adapting to e-bike systems of varying sizes.

[0033] By analyzing and extracting features from the collected data using machine learning algorithms, the system can accurately identify user intent, optimize the functions and control strategies of e-bikes, and make them more suitable for user needs and usage scenarios.

[0034] Bluetooth Mesh networks ensure real-time communication and data fusion between e-bikes and smart devices, enabling more accurate integration and analysis of cycling data (such as speed and heart rate) and external environmental information (such as GPS tracks), providing more personalized services and health advice.

[0035] By optimizing user feedback and interaction effects, the system can continuously improve the accuracy of intent recognition, enhance the adaptability and personalization of the e-bike system, and enable different users to obtain a customized riding experience.

[0036] By regularly evaluating and optimizing network communication quality, especially node signal optimization, the stability of the entire Mesh network has been improved, ensuring the reliable operation of e-bikes in complex environments.

[0037] One embodiment of the present invention, such as Figure 2 As shown, S1 includes:

[0038] S11. Obtain Bluetooth communication node information of ebike functional modules and smart devices through a scanning program. The node information includes node name, MAC address and signal strength.

[0039] S12. Construct a Bluetooth Mesh network topology for ebike usage scenarios using graph theory algorithms;

[0040] S13. A dynamic adjustment algorithm based on network topology automatically identifies new nodes when they join the network and integrates them into the existing topology according to their location and function. When existing nodes leave the network, the algorithm can detect the absence of nodes in a timely manner and replan the network path.

[0041] S14. Based on the constructed Bluetooth Mesh network topology, assign a unique network address to each node and set the node's communication parameters.

[0042] S15. Through the above configuration and settings, each node can communicate according to the Mesh network protocol to form a complete ebike Bluetooth Mesh communication network.

[0043] The working principle of the above technical solution is as follows: Bluetooth communication modules are installed on various functional modules of the ebike (such as the motor controller, battery management system, and dashboard), and these modules are ensured to be functioning properly. Simultaneously, the user's smart device (such as a smartphone or smartwatch) must have its Bluetooth function enabled. This is the basic hardware requirement for establishing Bluetooth communication. Through a scanning program, the system can actively search for and obtain Bluetooth communication node information from ebike functional modules and smart devices. This information includes the node name (used to identify the device type), MAC address (uniquely identifying each Bluetooth device), and signal strength (an important indicator of communication quality). By obtaining this information, the system can gain a preliminary understanding of the nodes present in the network.

[0044] Taking into account the distance between nodes, the impact of distance on communication quality is analyzed by measuring or estimating the physical distance between nodes and combining it with the propagation characteristics of Bluetooth signals (such as signal attenuation). Excessive distance may lead to severe signal attenuation, affecting the stability and reliability of communication. The communication signal strength of nodes is monitored in real time, and changes in signal strength over different time periods are recorded. This helps assess communication stability, promptly detect signal fluctuations or anomalies, and provide a basis for subsequent network optimization. Nodes are prioritized based on the importance of their functions; for example, motor controllers and battery management systems are crucial to the operation of e-bikes and should be given higher priority. When constructing the topology, the communication needs and connection stability of these important nodes are given priority. Graph theory algorithms (such as minimum spanning tree algorithm and shortest path algorithm) are used to construct a Bluetooth Mesh network topology for e-bike use cases. This structure aims to ensure stable and efficient communication between nodes, optimize communication paths, and improve the overall network performance by rationally planning the connection relationships between nodes.

[0045] Based on a dynamic adjustment algorithm for network topology, the system automatically identifies new nodes joining the network and integrates them appropriately into the existing topology according to their location and function. For example, when a user's smartwatch enters the coverage area of ​​the ebike Bluetooth Mesh network, the network can automatically add it to the topology and establish communication connections with other nodes. When an existing node leaves the network, the algorithm can promptly detect the node's absence and replan the network path. This ensures the stability and reliability of the network; even if a node leaves, it will not cause the entire network to collapse or communication to be interrupted.

[0046] Based on the constructed Bluetooth Mesh network topology, a unique network address is assigned to each node. This ensures that nodes can be accurately identified within the network, avoiding address conflicts and communication chaos. Node communication parameters, such as transmit power, communication frequency, and data transmission rate, are configured to meet the communication needs of different nodes. For example, for nodes at longer distances, the transmit power can be appropriately increased to enhance signal strength; for nodes requiring high-speed data transmission, a higher data transmission rate can be set. Node security parameters, such as encryption algorithms and keys, are configured to ensure the security of communication data. Encrypting communication data prevents data theft or tampering, ensuring the security of user information and eBike operational data.

[0047] Each node can communicate according to the Mesh network protocol, forming a complete ebike Bluetooth Mesh communication network. This network is stable, efficient, and secure, meeting the needs of intelligent human-computer interaction for ebikes and providing users with a convenient and comfortable riding experience.

[0048] The above technical solution achieves the following results: Through the Bluetooth Mesh network, e-bikes and users' smart devices can communicate efficiently and stably, supporting multiple interaction methods (such as voice, dashboard, and app). This enhances the user's interactive experience with the e-bike, making the riding process more personalized and convenient.

[0049] The adaptive nature of Bluetooth Mesh networks allows nodes to automatically adjust the network structure, reducing the need for manual debugging and maintenance. Furthermore, its strong scalability allows it to adapt to e-bike systems of varying sizes, lowering the difficulty of future expansion and maintenance.

[0050] By analyzing and extracting features from the collected data using machine learning algorithms, the system can accurately identify user intent, thereby optimizing the functions and control strategies of e-bikes, making the system more in line with user needs and usage scenarios, and improving the overall level of intelligence.

[0051] Bluetooth Mesh networks ensure real-time communication and data fusion between e-bikes and smart devices, enabling precise integration and analysis of cycling data (such as speed, heart rate, etc.) and external environmental information (such as GPS tracks) to provide users with personalized cycling advice and health services.

[0052] By continuously optimizing user feedback and interaction, the system has improved the accuracy of intent recognition and enhanced the adaptability and personalization of the e-bike system. Different users can enjoy a customized riding experience, meeting various individual needs.

[0053] Regularly assessing and optimizing communication quality, especially in node signal optimization, improves the stability of the entire Mesh network, ensuring that e-bikes can operate reliably in complex environments and reducing system failures caused by signal problems.

[0054] In one embodiment of the present invention, S13 includes:

[0055] S131. Based on the deployed node monitoring module, continuously scan devices within the Bluetooth signal range. When a new Bluetooth signal is detected, initiate the new node access process. By comparing the MAC address of the new node with the known node database, determine whether it is a newly added node. If it is a new node, further obtain its functional information.

[0056] S132. Based on the RSSI (Received Signal Strength Indication) value of the Bluetooth signal and combined with the location information of known nodes, make a preliminary judgment on the approximate location of the new node; based on the functional information of the new node, evaluate its role and importance in the ebike Bluetooth Mesh network;

[0057] S133. Based on the location and function evaluation results of the new nodes, the network topology is replanned using graph theory algorithms, and the load balancing of nodes in the network is considered. The network paths are adjusted to make the communication traffic evenly distributed among the nodes.

[0058] S134. Send network configuration information to the new node, update the network topology, include the new node in the network, and notify other nodes about the new node's addition; at the same time, update the network path.

[0059] S135. Monitor the communication status of nodes in the network in real time. When a node is detected to be unresponsive for a long time or its signal strength continues to decline, initiate the node departure detection process. Confirm whether the node has left the network by comparing it with the list of active nodes in the node database. If it is confirmed that the node has left the network, initiate the network reconstruction process and replan the network path.

[0060] S136. After a node joins or leaves the network, the stability of the network is evaluated in real time, and the network topology is further optimized based on the evaluation results.

[0061] The above technical solution works as follows: A node monitoring module deployed in the ebike Bluetooth Mesh network continuously scans for devices within Bluetooth signal range. Once a new Bluetooth signal is detected, the system immediately initiates the new node access process. By comparing the new node's MAC address with a known node database, the system can quickly determine if the node is a newly added node. This step ensures that only truly new nodes enter the subsequent access process, avoiding redundant processing. If it is determined to be a new node, the system further obtains its functional information, such as node type and supported communication protocols. This information is crucial for subsequent network topology planning and node configuration.

[0062] Based on the RSSI value of the Bluetooth signal and the location information of known nodes, the system can initially determine the approximate location of a new node. This helps in rationally arranging the connection relationships of new nodes in subsequent network topology planning. According to the functional information of the new node, the system assesses its role and importance in the ebike Bluetooth Mesh network. For example, if the new node is a sensor related to cycling safety, it should be given higher priority to ensure the stability and reliability of its communication.

[0063] Based on the location and functional evaluation results of the new nodes, the system uses graph theory algorithms to restructure the network topology. This step aims to ensure that the new nodes can be efficiently integrated into the existing network while optimizing the overall network communication performance. When restructuring the network topology, the system considers load balancing among nodes to prevent performance degradation caused by some nodes bearing excessive communication tasks. By adjusting network paths, communication traffic is evenly distributed among nodes, improving the overall efficiency of the network.

[0064] The system sends network configuration information to the new node, including network address, communication parameters, and security parameters. This information is fundamental for the new node to communicate according to the Mesh network protocol. The network topology is updated to include the new node in the network, and other nodes are notified of its addition. Simultaneously, network paths are updated to ensure efficient data transmission between the new node and existing nodes.

[0065] The system monitors the communication status of nodes in the network in real time. If a node is detected to be unresponsive for an extended period or its signal strength continues to decline, a node departure detection process is immediately initiated. By comparing the node's information with the list of active nodes in the node database, the system confirms whether the node has left the network. This step ensures that only nodes that have truly left trigger the network reconstruction process. If a node departure is confirmed, the system initiates the network reconstruction process and re-plans the network path. This helps maintain network stability and reliability, avoiding communication interruptions or performance degradation caused by node departures.

[0066] After a node joins or leaves the network, the system evaluates the network's stability in real time, including metrics such as communication latency and packet loss rate. These metrics reflect the network's communication performance and reliability. Based on the evaluation results, the system further optimizes the network topology, such as adjusting node transmission power and communication frequency. This step aims to continuously improve network stability and performance to meet the needs of ebike's intelligent human-computer interaction.

[0067] The benefits of the above technical solution are as follows: by automatically detecting and flexibly handling the addition of new nodes, the system can dynamically adjust the network structure according to actual needs, ensuring that the network can expand as the number of devices and nodes increases. This automated adjustment enables the e-bike Bluetooth Mesh network to efficiently adapt to different environments and user needs, improving the system's adaptability.

[0068] By automatically monitoring and adjusting node joining and leaving, the need for manual intervention is reduced, making network maintenance simpler and less costly. The system can manage the lifecycle of nodes autonomously, enhancing the network's self-healing capabilities.

[0069] By monitoring the signal strength and response status of nodes in real time, the system can react promptly and replan network paths when a node leaves or signal quality deteriorates. This enhances network robustness and avoids overall network performance degradation caused by a single node problem.

[0070] By optimizing the network topology using graph theory algorithms, communication tasks are evenly distributed among nodes, avoiding overload on some nodes and improving communication efficiency and data transmission stability.

[0071] When each new node joins, the system sends it security configuration information, such as encryption algorithms and keys, to ensure secure communication within the network. This enhances the overall network security and protects data transmission from attacks.

[0072] Based on the functional information of new nodes (such as safety sensors, power system modules, etc.), the system can assess their importance in the network and assign different priorities to nodes. Through this intelligent management, the operational performance of e-bikes can be effectively optimized, ensuring the stable operation of critical modules (such as safety monitoring).

[0073] The system can monitor and respond to dynamic changes in the network in real time, including the addition, departure, or signal changes of nodes. This enables the network to quickly adjust to different environments or user scenarios, ensuring efficient system operation.

[0074] In one embodiment of the present invention, S133 includes:

[0075] Using graph structure from graph theory, existing nodes and new nodes are represented as vertices in the graph, and the connections between vertices are represented as edges; based on the known node position information, the relative positions of each vertex in the graph are initially determined;

[0076] By combining the functional information of the new node, we analyze its communication needs with existing nodes. Based on graph theory algorithms (such as Dijkstra's algorithm), we calculate the optimal communication path between the new node and existing nodes according to the communication needs and location information between nodes.

[0077] The planned communication paths are quality-assessed, and the network topology is fine-tuned based on the path quality and stability assessment results. The quality of the communication paths is evaluated using the following formula:

[0078]

[0079] in, This represents the standardized basic metrics (length, hop count, latency, bandwidth, packet loss rate, all ∈ [0,1]).

[0080] H represents the weight of the basic indicator; H represents the number of hops in the path. Indicates the reliability of the k-th link (∈[0,1]); This represents the stability coefficient of m nodes (∈[0,1]); Represents the stability weight (∈[0,1]);

[0081] The adjusted network topology is displayed in a visual manner, and a detailed network topology update plan is generated, including changes to node connection relationships and adjustments to communication parameters.

[0082] The working principle of the above technical solution is as follows: It utilizes graph structures from graph theory to model the Bluetooth Mesh network, representing existing and new nodes as vertices in the graph. The connections between vertices (i.e., communication links) are represented as edges. This abstraction facilitates the mathematical handling of network topology problems. For example, in a simple e-bike Bluetooth Mesh network, the motor controller, battery management system, dashboard, and newly added smart devices are all vertices in the graph, and the Bluetooth communication links between them are the edges. Based on the known node location information, the relative positions of each vertex in the graph are initially determined. Although this is not a strict physical location mapping, it helps to understand the impact of spatial relationships between nodes on communication. For example, nodes that are closer together may be arranged in close positions in the graph, facilitating subsequent consideration of factors such as communication distance and signal strength.

[0083] Based on the functional information of the new node, analyze its communication requirements with existing nodes, including communication frequency and data volume. Different functional nodes have different communication requirements. For example, sensors related to cycling safety may require high-frequency, low-data-volume communication to transmit safety information in real time; while nodes used for data recording and analysis may require low-frequency, high-data-volume communication. Using graph theory algorithms (such as Dijkstra's algorithm), calculate the optimal communication path between the new node and existing nodes based on the communication requirements and location information. Dijkstra's algorithm finds the shortest path from a starting point to all other vertices; here, it can be understood as finding the optimal communication link that satisfies communication requirements while comprehensively considering factors such as location. During path planning, fully consider the load balancing of nodes in the network to avoid performance degradation caused by some nodes bearing too many communication tasks. Just as in a traffic network, one road cannot bear too much traffic flow, otherwise it will lead to congestion; the same applies to Bluetooth Mesh networks, where communication tasks must be allocated reasonably.

[0084] The planned communication path undergoes quality evaluation, including metrics such as communication latency, packet loss rate, and signal strength. Communication latency refers to the time required for data to travel from the sender to the receiver; packet loss rate refers to the proportion of data packets lost during data transmission; and signal strength directly affects the stability and reliability of communication. Evaluating these metrics reveals the performance of the communication path. Simulation or real-world testing is used to verify the path's stability in different scenarios. For example, testing the communication path's ability to maintain stable communication in various cycling environments (such as urban roads and mountain roads). Ensuring new nodes can be efficiently and stably integrated into the existing network is crucial, much like ensuring a new member can smoothly integrate and fulfill their role when joining a team.

[0085] Based on the path quality and stability assessment results, the network topology is fine-tuned, such as optimizing the connections between nodes and adjusting the transmission power of nodes. If the signal strength of a certain communication path is found to be weak, the transmission power of the relevant nodes can be adjusted; if some nodes are found to be overloaded, the connections between nodes can be optimized, and some communication tasks can be transferred to other nodes. The goal is to ensure that the adjusted network topology can meet the communication needs of the new nodes while maintaining the overall stability and performance of the network. This is similar to adjusting a mechanical structure, ensuring that the newly added components function properly while maintaining the stability and performance of the entire structure.

[0086] The adjusted network topology is displayed visually, allowing users to intuitively understand the network's connections and structure. For example, communication links between nodes, as well as node locations and statuses, are graphically represented. Detailed network topology update plans are generated, including changes to node connections and adjustments to communication parameters. This provides clear guidance for subsequent network maintenance and management, acting like a detailed blueprint to guide engineers on how to modify and upgrade the network.

[0087] The above technical solution achieves the following results: by calculating the optimal communication path using graph theory algorithms and combining it with the functional information and communication requirements of nodes, the system can select the most suitable path for data transmission. This not only reduces communication latency but also improves data transmission efficiency, ensuring that new nodes can be stably and efficiently integrated into the network.

[0088] During path planning, the system fully considers the load of each node to avoid performance degradation caused by some nodes undertaking too many communication tasks. This reduces network bottlenecks caused by node overload and improves the overall performance and stability of the network.

[0089] The planned communication path is validated through quality assessments (including metrics such as latency, packet loss rate, and signal strength) to ensure its stability under different scenarios. This enhances network stability and reduces data transmission errors or interruptions caused by path instability.

[0090] Based on path quality assessment, the system can fine-tune the network topology, adjusting connections and communication parameters between nodes, such as transmit power. This ensures the adjusted network topology better meets the communication needs of new nodes while maintaining overall network stability and performance.

[0091] By visualizing the network topology, users can clearly see the node connections and changes in communication parameters, helping technicians quickly understand the network status and perform debugging and optimization. This reduces the complexity of network maintenance and adjustment, and improves operational efficiency.

[0092] By combining the actual functions and communication needs of nodes for path planning, new nodes can dynamically select the most suitable communication path based on the actual scenario and task. This enables the system to intelligently optimize for different application scenarios, providing more flexible network adaptability.

[0093] Through simulation or real-world testing, the system can verify path stability and optimize network performance under different scenarios. This enhances the system's adaptability to changing environments and ensures stable network operation.

[0094] The formula above combines common network performance metrics such as length, hop count, latency, bandwidth, and packet loss rate (which typically affect network efficiency), while also incorporating two important factors: link reliability and node stability, making the evaluation criteria more comprehensive. The introduction of weighting factors and stability weights allows the formula to be flexibly adjusted to meet different needs. For example, in some network environments, bandwidth may be more important than latency, or for certain high-reliability applications, stability may be more critical than other factors. Through flexible weight allocation, the formula can adapt to diverse requirements.

[0095] The inclusion of link reliability and node stability considers not only the instantaneous performance of the path but also the long-term stability of the network. This is crucial because many network applications require stable performance over extended periods. The formula provides a quantifiable path quality Q, making network optimization more practical. By scoring each path, it's clear which paths are optimal, allowing for further optimization based on these results. Thus, path selection and fine-tuning are no longer experience-driven but rather based on data and analysis-driven scientific decisions. The optimized topology is visualized, enhancing the understandability and operability of the solution. Simultaneously, generating detailed update plans helps the team track changes during actual deployment, ensuring that network structure and performance improvements are synchronized.

[0096] In one embodiment of the present invention, S135 includes:

[0097] The system monitors the communication status of nodes in the network in real time. When a node's signal strength continuously decreases or it becomes unresponsive for an extended period, it is marked as a potentially abnormal node. The system further analyzes the historical communication data of the abnormal node to confirm whether any anomalies exist.

[0098] For nodes confirmed to be abnormal, initiate the node departure detection process; by comparing the list of active nodes in the node database, query the latest status information of the node to confirm whether it has actually left the network.

[0099] If a comparison confirms that a node has left the network, it is removed from the list of active nodes, and relevant information is recorded. At the same time, the network topology is updated, and the node is marked as having left the network.

[0100] Based on the changes in network topology after a node leaves, initiate the network reconstruction process; and use graph theory algorithms to replan network paths.

[0101] The reconstructed network is validated and tested; based on the validation results, the network topology is further optimized, information about node departures and network reconstruction is communicated to other nodes in the network, and relevant records in the network management system are updated.

[0102] The working principle of the above technical solution is as follows: A monitoring mechanism deployed in the network continuously monitors the communication status of each node in real time. This includes real-time tracking of node signal strength and continuous observation of node response. Signal strength is an important indicator of communication quality between nodes, while node response reflects whether the node is in normal working condition. When a node's signal strength continuously decreases or it becomes unresponsive for an extended period, it indicates a potential anomaly and is marked as a potentially anomalous node. For these marked potentially anomalous nodes, their historical communication data is further analyzed. Historical communication data contains the node's communication records over a past period. In-depth analysis of this data can confirm whether there have been communication interruptions, data loss, or other anomalies. For example, if a node is found to experience frequent communication interruptions over a period of time, or to have lost a large number of data packets during data transmission, it further confirms that the node is anomaly.

[0103] For nodes confirmed to be abnormal, the system initiates a node departure detection process. The purpose of this process is to accurately determine whether the node has actually left the network. The system queries the latest status information of the node by comparing it with the active node list in the node database. The active node list records information about nodes currently operating normally in the network; comparing this list confirms whether the node has actually left the network. If the node is not in the active node list, or if the status recorded in the list does not match the actual situation, then it can be confirmed that the node has left the network.

[0104] If a node is confirmed to have left the network through comparison, it is removed from the list of active nodes, and the departure time, reason, and other relevant information are recorded. This information is crucial for subsequent network management and troubleshooting. For example, analyzing the reason for a node's departure can identify potential problems in the network and allow for appropriate improvements. Simultaneously, the network topology is updated, marking the node as "left." The network topology reflects the connections between nodes in the network; marking left nodes allows the network topology to more accurately reflect the current network situation.

[0105] The network reconstruction process is initiated based on the changes in network topology after a node leaves. After a node leaves, the original network topology may no longer meet the network's communication requirements, necessitating a replanning of network paths to ensure efficient communication connections between the remaining nodes. Graph theory algorithms are used to replan network paths. These algorithms can calculate the optimal network path based on factors such as node positions and communication needs. For example, Dijkstra's algorithm finds the shortest path from a starting point to all other vertices. Applying these algorithms ensures more efficient and stable communication paths between the remaining nodes. The reconstructed network is then validated and tested, and the network topology is further optimized based on the results. Validation tests may include communication latency tests and packet loss rate tests to understand the performance of the reconstructed network. If unsatisfactory communication performance between some nodes is found, optimization can be achieved by adjusting parameters such as node transmit power and communication frequency to improve network communication quality and stability.

[0106] The system notifies other nodes in the network of node departures and network reconfigurations, ensuring they can update their routing tables and network topology information promptly. In a network, each node needs to select communication paths based on the network topology; therefore, timely updates to routing tables and network topology information are crucial for ensuring normal network operation. Relevant records in the network management system, including node departure and network reconfiguration information, are also updated. These records provide valuable information for subsequent network management and maintenance. For example, analyzing historical node departure records can identify potential weaknesses in the network and allow for appropriate reinforcement measures.

[0107] The effect of the above technical solution is that by monitoring the communication status of nodes in real time, potential abnormal nodes with declining signal strength or no response for a long time can be detected in a timely manner, reducing the risk of network interruption and enhancing network reliability.

[0108] After detecting an abnormal node, the node departure detection process is initiated. The historical data of the abnormal node is automatically analyzed, which can quickly confirm whether it has actually left the network, reducing the need for manual intervention and shortening the fault recovery time.

[0109] By using graph theory algorithms to reconstruct the network, the system can automatically adjust the network paths, ensure efficient communication connections between the remaining nodes, and enhance the network's adaptability after a node leaves.

[0110] The network topology can be adjusted through optimization algorithms after a node leaves, ensuring that the system continues to operate efficiently and stably after a node leaves.

[0111] By automatically updating node status, topology, and routing tables, network management is simplified, manual operations are reduced, and operational efficiency is improved.

[0112] Through automated processes such as real-time monitoring, historical data analysis, and path reconstruction, the network's self-diagnosis and repair capabilities are enhanced, making the network operation more intelligent.

[0113] By ensuring that abnormal nodes are detected and removed in a timely manner, the network can quickly adapt to node changes, reduce network performance degradation caused by node departure, and ensure the stability of data transmission.

[0114] By updating the network topology and path planning in a timely manner, communication between the remaining nodes is ensured to be unaffected by node departures, thereby reducing the risk of communication delays and path interruptions.

[0115] This solution supports the dynamic addition and removal of nodes, and by optimizing network topology and path planning, it can support larger-scale network expansion to meet future needs.

[0116] By updating the network topology and recording information such as departure time and reason, the system's ability to track abnormal node states is enhanced, thereby improving the security of network management.

[0117] In one embodiment of the present invention, S2 includes:

[0118] S21. During the operation of ebike, data related to ebike is collected through various functional modules and smart devices;

[0119] S22. Clean the collected data, use a filtering algorithm to smooth the cleaned data, and classify the smoothed data according to different types.

[0120] S23, and extract features from the classified data;

[0121] S24. Through feature extraction, the original data is transformed into representative and analyzable feature data.

[0122] The working principle of the above technical solution is as follows: During the operation of the ebike, data related to the ebike is comprehensively collected through functional modules distributed in different parts of the ebike and the smart devices carried by the user. This includes:

[0123] Motor Controller and Motor Data: A speed sensor is installed in the eBike's motor controller to collect riding speed data in real time and accurately. For example, when the rider accelerates or decelerates, the speed sensor can quickly detect and record the speed change. Simultaneously, a speed sensor is installed on the motor to obtain motor speed information. Motor speed is closely related to riding speed, and analyzing the motor speed reveals the vehicle's power output.

[0124] Battery Management System (BMS) Data: The BMS is responsible for collecting battery power data, including remaining percentage, voltage, and current. The remaining percentage allows riders to intuitively understand the battery's remaining capacity, while voltage and current data can be used to assess the battery's operating status and health. For example, a sudden drop in voltage or an abnormal increase in current may indicate a battery malfunction.

[0125] Dashboard data: The dashboard records cycling mileage data. Cycling mileage is one of the important indicators for measuring e-bike usage, as it reflects the distance the rider travels and the frequency of vehicle use.

[0126] Smart device data: The smartwatch worn by the user collects the user's heart rate data through its built-in heart rate sensor. This heart rate data reflects the cyclist's physical condition and exercise intensity. For example, the heart rate will increase when the cyclist is engaged in high-intensity cycling, and will decrease accordingly when the cycling intensity decreases. The mobile phone uses the GPS positioning system to obtain the user's movement trajectory information. The movement trajectory can record the cyclist's route and location information, providing geospatial data support for subsequent analysis.

[0127] The collected raw data may contain noise and outliers, thus requiring cleaning. Filtering algorithms are used to smooth the cleaned data, removing high-frequency noise and making the data more stable. For example, for cycling speed data, filtering algorithms can eliminate speed fluctuations caused by sensor measurement errors or road bumps, making the speed data more accurate and reliable. The smoothed data is then categorized according to different types for targeted analysis and processing. Cycling speed and motor speed are classified as motion data, primarily reflecting the e-bike's activity status and performance. Battery level and remaining range are classified as energy data, crucial for evaluating e-bike range and battery management. User heart rate and movement trajectory are classified as user health and location data, which can be used to analyze the rider's physical condition and travel habits.

[0128] Feature extraction is performed, including:

[0129] Motion data feature extraction: For cycling speed data, the average speed, maximum speed, and minimum speed within a certain time window are calculated. These features can reflect the cyclist's overall cycling speed level and speed fluctuation range over a period of time. Simultaneously, the trends in speed changes, such as acceleration and deceleration, are analyzed. Acceleration and deceleration reflect the cyclist's acceleration and deceleration behavior, which is of great significance for evaluating the cyclist's riding style and the vehicle's power performance.

[0130] Energy data feature extraction: Features such as remaining battery percentage, battery consumption rate (the amount of battery power reduced per unit time), and estimated remaining range are extracted from battery power data. The remaining battery percentage directly reflects the battery's remaining capacity, while the battery consumption rate can be used to assess battery efficiency and vehicle energy consumption. The estimated remaining range predicts the distance the vehicle can still travel based on the current battery level and riding conditions, providing a reference for the rider.

[0131] User health and location data feature extraction: For user heart rate data, calculate features such as average heart rate, heart rate fluctuation range (difference between maximum and minimum heart rate), and the correlation between heart rate and cycling intensity. Average heart rate reflects the cyclist's overall heart rate level during cycling, heart rate fluctuation range reflects the magnitude of heart rate changes, and the correlation between heart rate and cycling intensity can be used to analyze the impact of cycling intensity on heart rate.

[0132] Through the feature extraction process described above, the raw data is transformed into representative and analyzable feature data. This feature data can more accurately reflect information such as the e-bike's operational status, the user's physical condition, and travel habits, providing a foundation for subsequent user intent recognition and interaction. For example, by analyzing cycling speed and acceleration characteristics, it is possible to determine whether the cyclist is accelerating, decelerating, or maintaining a constant speed, thereby inferring the cyclist's intention; by analyzing the correlation between the user's heart rate and cycling intensity, personalized exercise suggestions can be provided to the cyclist, improving cycling safety and comfort.

[0133] The effect of the above technical solution is that by collecting multi-dimensional data related to electric bicycles through multiple sensors and smart devices, it can comprehensively reflect various information during the riding process, and improve the accuracy and breadth of data collection.

[0134] By cleaning and filtering the collected data, noise was effectively removed, ensuring data quality and accuracy, and reducing subsequent analysis errors caused by inaccurate or incomplete data.

[0135] By smoothing and extracting features from the cleaned data, the original data is transformed into representative and analyzable feature data, providing a more accurate foundation for user intent recognition and behavior prediction.

[0136] By extracting features from user heart rate data, such as calculating average heart rate and heart rate fluctuation range, it is possible to more accurately monitor the user's physical condition and perform correlation analysis with cycling intensity, thereby improving the real-time performance and accuracy of health monitoring.

[0137] By extracting characteristic data such as battery power, consumption rate, and remaining range, the battery usage can be monitored in real time, helping users understand the battery status, optimize riding strategies, and enhance the intelligence level of energy management.

[0138] By analyzing the characteristics of data such as riding trajectory and cycling speed, we can identify users' cycling habits and preferences, provide more personalized cycling suggestions and feedback, and improve users' cycling experience.

[0139] By extracting and classifying features from various types of data, the subsequent user intent recognition process is simplified, enabling more efficient analysis and prediction of user behavior and reducing the complexity of intent recognition.

[0140] By collecting and processing multi-dimensional data in real time, the system can quickly respond to user needs and environmental changes, improving the real-time performance and flexibility of the overall system.

[0141] By performing preprocessing steps such as data cleaning, filtering, and feature extraction, the raw data can be transformed into more compact and effective feature data, thereby improving the efficiency of data processing and the operability of subsequent analysis.

[0142] Transforming feature data into more representative indicators makes cycling data easier to analyze and present, facilitating user data viewing and decision-making, and enhancing user interaction experience and decision support capabilities.

[0143] In one embodiment of the present invention, S3 includes:

[0144] S31. Collect historical cycling data and user interaction records; label the collected data and determine the user intent corresponding to each data sample;

[0145] S32. Select a machine learning algorithm and train the model using labeled data; during the training process, continuously adjust the model parameters, evaluate the trained model using a test dataset, and optimize the model based on the evaluation results;

[0146] S33. Analyze the functions of ebike and determine the appropriate interaction method for each function; combine users' historical usage habits and preferences to recommend interaction modes;

[0147] S34. Based on the identified user intent and the above-mentioned interaction mode selection criteria, determine the appropriate interaction mode.

[0148] The working principle of the above technical solution is as follows: It collects a large amount of historical cycling data and user interaction records, covering cycling status (such as speed, motor speed, etc.), user physiological data (such as heart rate, etc.), environmental information (such as weather, road conditions, etc.), and corresponding user intentions (such as acceleration, deceleration, switching cycling modes, etc.). This data comprehensively reflects various factors during the cycling process, providing a rich information foundation for subsequent user intention recognition. For example, in sunny weather with good road conditions, users may be more inclined to accelerate; while in rainy weather with complex road conditions, users may slow down or switch to a safer cycling mode. The collected data is labeled to determine the user intention corresponding to each data sample. Data labeling is a key step in training machine learning models. Through labeling, the model can learn the correspondence between different data features and user intentions. For example, a set of data showing gradually increasing cycling speed, rising motor speed, and increasing user heart rate can be labeled as "acceleration intention."

[0149] Choose a suitable machine learning algorithm and train the model using labeled data. Common machine learning algorithms such as decision trees, support vector machines, and neural networks can all be used for user intent recognition. During training, the model learns feature patterns in the data and classifies new data based on these patterns, i.e., recognizing user intent. For example, neural networks can gradually improve the accuracy of user intent recognition through the learning and adjustment of multiple layers of neurons. Continuously adjust the model's parameters and evaluate the trained model using a test dataset. Parameter tuning is to optimize the model's performance, ensuring it performs well on different datasets. The test dataset is independent of the training dataset and is used to evaluate the model's generalization ability. Optimize the model based on the evaluation results. For example, if the model's accuracy on the test dataset is low, it may be necessary to adjust the model's parameters, increase the amount of training data, or change the algorithm to ensure that the model can accurately recognize the user's intent based on the current riding status, user physiological data, and environmental information.

[0150] Analyze the functions of the e-bike, such as acceleration, deceleration, switching riding modes, and viewing riding data, to determine the appropriate interaction method for each function. Different functions have different requirements for interaction methods. For example, acceleration and deceleration functions require fast and accurate operation, making them suitable for interaction methods such as handlebar buttons or voice commands; while viewing riding data can be done through the dashboard display or a mobile app. Consider the current usage scenario, such as when the user is riding at a relatively high speed. To avoid affecting riding safety, voice interaction or a simple dashboard display should be prioritized. Voice interaction allows users to obtain information or perform operations without distraction, while a simple dashboard display can quickly convey key information. When the user is parked, a wider range of mobile app interaction methods can be selected, such as viewing detailed riding data and setting riding parameters. Recommend interaction modes based on the user's historical usage habits and preferences. For example, if the user is accustomed to using the mobile app, then the mobile app interaction mode should be prioritized, while ensuring safety and convenience. By analyzing the user's historical usage records, we can understand the user's preferences and provide a more personalized interactive experience.

[0151] Based on the identified user intent and the aforementioned interaction mode selection criteria, a suitable interaction mode is determined. For example, when the user's intent is to check their cycling speed, the speed information is displayed in real-time digital form on the dashboard, and the user can also be prompted with voice commands about the current speed. This interaction method allows users to quickly obtain speed information without compromising cycling safety. When the user's intent is to switch cycling modes, the user is prompted with voice confirmation, such as "Are you sure you want to switch to Sport mode?", and the mode switch status and related information, such as changes in power output and range, are simultaneously displayed on the mobile app. This interaction method ensures that users are aware of relevant information before switching modes, avoiding inconvenience or safety hazards caused by accidental operation.

[0152] The above technical solution achieves the following results: by collecting a large amount of historical cycling data and user interaction records, and annotating the data, it can accurately identify user intent. The machine learning model is continuously optimized during training, improving the accuracy of user intent recognition and ensuring that the system can accurately determine user needs based on different cycling states, user physiological data, and environmental information.

[0153] When designing the interaction mode, safety under different riding conditions was fully considered. For example, when the user is riding at a higher speed, the system prioritizes voice interaction or a simple dashboard display, avoiding interference from complex operations and reducing the risk of user misoperation.

[0154] Based on users' historical usage habits and preferences, the system can intelligently recommend the most suitable interaction method. For example, if a user is accustomed to using a mobile app, the system will prioritize recommending the mobile app interaction mode, while simultaneously meeting security and convenience requirements, thus improving the personalization and intelligence of the interaction mode.

[0155] By combining different usage scenarios and interaction needs, the system can provide the optimal interaction method at the appropriate time. For example, when the user is parked, it recommends using a rich mobile app for interaction, while during cycling, it prioritizes a simple and fast interaction method, improving user convenience and interactive experience.

[0156] By combining real-time riding status, user physiological data, and environmental information, the system can promptly identify user intentions and respond during riding. For example, when a user wants to check their riding speed, the system displays the speed in real time on the dashboard and provides voice feedback, allowing the user to obtain the necessary information without distraction.

[0157] By clearly analyzing the interaction methods of different functions, the system can select the most appropriate interaction mode based on the specific function. For example, when the user intends to switch cycling modes, the system not only prompts the user for confirmation via voice, but also displays the mode switching information simultaneously on the mobile app, enhancing the operability and transparency of the function operation.

[0158] By automatically recognizing user intent and recommending appropriate interaction methods, users no longer need to perform complex manual operations, reducing operational complexity, simplifying the interaction process between users and the system, and improving the overall user experience.

[0159] The system continuously optimizes intent recognition and interaction method recommendation through machine learning algorithms, making the interaction between users and electric bicycles more intelligent, enhancing the system's adaptability and intelligence, and improving the overall interactive experience.

[0160] Through accurate intent recognition and personalized interactive recommendations, users can make cycling decisions more efficiently, such as switching cycling modes and viewing cycling data, reducing decision-making time and improving overall cycling efficiency.

[0161] Through continuous training and optimization of machine learning algorithms, the system can make corresponding adjustments based on constantly changing user needs and riding environments, thereby improving the system's maintainability and scalability to adapt to future new functions.

[0162] In one embodiment of the present invention, step S4 includes:

[0163] S41. After determining the interaction mode, when the user issues a control command through a smartphone, the smartphone sends the command to the corresponding functional module of the ebike through the Bluetooth Mesh network.

[0164] After receiving the instruction, the S42 and ebike's functional modules execute the corresponding operations and send the results back to the smartphone for the user to view.

[0165] S43. The cycling speed information collected by the ebike dashboard is fused with the motion trajectory information obtained by the smartphone GPS positioning system, and a more accurate cycling distance and speed distribution are calculated through the data fusion algorithm.

[0166] S44. Integrate heart rate data collected by the user's smartwatch with ebike riding status data to analyze the impact of cycling on the user's body; and provide users with more personalized health advice.

[0167] The working principle of the above technical solution is as follows: After determining the interaction mode, when the user issues control commands (such as acceleration, deceleration, switching riding modes, etc.) through a smartphone, the smartphone sends the commands to the corresponding functional modules of the ebike via the Bluetooth Mesh network. The Bluetooth Mesh network has advantages such as low power consumption, wide coverage, and multi-node connectivity, ensuring stable and efficient transmission of commands between the smartphone and the ebike. For example, when a user wants to accelerate while riding, they simply click the acceleration button on the mobile app, and the smartphone quickly sends the acceleration command to the ebike's motor controller via the Bluetooth Mesh network. Upon receiving the command, the ebike's functional module immediately executes the corresponding operation. Taking the acceleration command as an example, after receiving the command, the motor controller adjusts the motor's power supply parameters to increase the motor speed, thereby accelerating the vehicle. Simultaneously, the functional module feeds back the operation results to the smartphone for the user to view. For instance, after acceleration, the motor controller feeds back the accelerated speed information to the mobile app via the Bluetooth Mesh network, allowing the user to monitor the vehicle's speed in real time on the mobile app. This command execution and feedback mechanism enables real-time interaction between the user and the ebike, allowing the user to easily control the vehicle and obtain vehicle status information.

[0168] The speed information collected by the e-bike dashboard is real-time and accurate, reflecting the vehicle's speed at a specific moment. The motion trajectory information obtained from the smartphone's GPS positioning system includes the rider's location and route information, but may contain some positioning errors. By fusing these two types of data and using a data fusion algorithm, a more accurate calculation of riding distance and speed distribution can be achieved. The data fusion algorithm can determine the approximate riding location and route based on GPS positioning information, and combine this with the speed information from the dashboard to more accurately calculate the speed and riding distance for each segment. For example, on a straight road segment, GPS positioning information can determine the start and end points of the segment, while the speed information from the dashboard can record the vehicle's speed changes along that segment. The data fusion algorithm can then calculate the actual riding distance and average speed for that segment. This data fusion method compensates for the shortcomings of a single data source, improving the accuracy and reliability of riding data.

[0169] Heart rate data collected by a user's smartwatch reflects their physical condition, while eBike riding status data (such as riding speed and motor power) reflects the intensity and difficulty of the ride. There is a correlation between these two types of data; for example, the higher the riding intensity, the higher the user's heart rate may be. By fusing heart rate data with riding status data, the impact of riding on the user's body can be analyzed. By analyzing heart rate, riding speed, and motor power, it's possible to determine whether the user's riding intensity is appropriate. If the heart rate is too high, and the riding speed and motor power are also high, it indicates that the user's riding intensity may be too high. In this case, the system can provide more personalized health advice, such as, "Based on your current heart rate and riding intensity, we suggest you appropriately reduce your speed to maintain healthy riding." This data fusion and analysis method helps users better understand their physical condition and riding intensity, rationally plan their riding schedules, and ensure riding safety.

[0170] The effect of the above technical solution is that, through the Bluetooth Mesh network, the smartphone can quickly transmit the user's control commands to the functional modules of the electric bicycle, reducing the delay in command transmission, improving the system's response speed and accuracy, and ensuring timely feedback of user operations.

[0171] Through a clear division of functional modules and a command feedback mechanism, electric bicycles can efficiently execute user commands and send the results back to the mobile app. This interaction method simplifies the operation process, reduces malfunctions caused by user errors, and improves system reliability and user experience.

[0172] By fusing cycling speed information from the eBike dashboard with motion trajectory information obtained from the smartphone's GPS positioning system, a data fusion algorithm is used to calculate more accurate cycling distance and speed distribution. This technology enhances the accuracy and real-time nature of cycling data, providing more precise cycling information and route analysis.

[0173] By integrating heart rate data collected from the user's smartwatch with riding status data from the electric bicycle, the system can analyze the user's physical responses during riding and provide more personalized health advice. For example, when the system determines that the user's riding intensity is too high, it will promptly prompt the user to reduce speed, helping the user maintain an appropriate riding intensity and reduce health risks.

[0174] When a user issues a control command, the smartphone can provide real-time feedback based on the riding status and the result of the operation, such as speed information after acceleration, enhancing the user's interactive experience with the system. The system's feedback is not only a single operation result, but also incorporates the user's real-time status, improving the intelligence and dynamism of the feedback.

[0175] By integrating cycling data from multiple sources, the system can provide users with more accurate cycling information and offer reasonable health advice. Users can then make more effective cycling decisions based on this data, avoiding over-cycling or other behaviors that may lead to health problems.

[0176] The system seamlessly collaborates with various devices via a Bluetooth Mesh network, reducing reliance on external hardware. Users can obtain cycling data and health information without needing numerous separate devices, improving the system's ease of use and accessibility.

[0177] By integrating cycling data (such as speed and distance), health data (such as heart rate), and external positioning data (such as GPS), the system achieves seamless integration of multiple functions, providing users with a more comprehensive and richer cycling experience. This multi-functional integration enhances the system's comprehensiveness and practicality, improving the overall user experience.

[0178] By simplifying the transmission process of control commands and optimizing the command feedback mechanism, users can more easily control electric bicycles via their mobile phones and check the riding status in real time. This optimization improves the system's operability and smoothness of interaction, reducing lag or delays during operation.

[0179] Through deep integration with devices such as smartphones and smartwatches, users can obtain various data and health feedback about cycling in real time, enhancing their sense of control and participation in the cycling process and improving the overall user experience.

[0180] In one embodiment of the present invention, S43 includes:

[0181] The cycling speed information collected from the ebike dashboard is preprocessed, and the motion trajectory information obtained from the smartphone GPS positioning system is also preprocessed.

[0182] Key features are extracted from the preprocessed cycling speed information, and features related to cycling speed are extracted from the motion trajectory information. These features are then matched with the cycling speed features to identify the speed characteristics of different road sections.

[0183] Based on the feature matching results, a data fusion model is constructed. Using the constructed data fusion model, the cycling distance and speed distribution are calculated based on cycling speed information and motion trajectory information.

[0184] Based on the road segment divisions in the motion trajectory information, the cycling process is divided into multiple road segments; for each road segment, the cycling distance and average speed of that segment are calculated using a data fusion model; the average speed is obtained using the following formula:

[0185]

[0186] in, This represents the average speed (m / s) of the i-th road segment. This represents the instantaneous velocity function collected by the instrument panel;

[0187] Represents the instantaneous velocity function calculated by GPS; Indicates the time interval of road segment i; Indicates the fusion weights;

[0188] The cycling distance and speed distribution calculated by the data fusion algorithm are compared and verified with the actual situation. Based on the verification results, the data fusion model is optimized and adjusted.

[0189] The working principle of the above technical solution is as follows: The cycling speed information collected by the e-bike dashboard may be subject to noise interference, such as sensor measurement errors and instantaneous speed fluctuations caused by road bumps. By removing noise and filtering smoothing, these interference factors can be effectively eliminated, making the speed data more stable and accurate. For example, a moving average filtering algorithm is used to average the speed data within a certain time window, thereby smoothing out instantaneous speed fluctuations and obtaining data closer to the true speed. The motion trajectory information obtained by the smartphone's GPS positioning system may have problems such as inconsistent coordinate systems and timestamps. To ensure consistency with the cycling speed information in time and space, the motion trajectory information needs to be preprocessed. Coordinate transformation can unify the position information under different coordinate systems into the same coordinate system, facilitating subsequent data analysis and fusion. Timestamp alignment matches the speed information and trajectory information in chronological order, ensuring that the corresponding speed and position information can be obtained at the same point in time.

[0190] Key features, such as peak speed and rate of change of speed, are extracted from the preprocessed cycling speed information. Peak speed reflects the maximum speed during cycling, while the rate of change of speed reflects the degree of increase or decrease in speed. These features describe the overall trend and characteristics of cycling speed changes. Features related to cycling speed, such as road segment length and turning angle, are extracted from the trajectory information. Road segment length can be obtained by calculating and summing the distances between adjacent points on the trajectory, while turning angle can be determined by analyzing the directional changes of points on the trajectory. Matching these trajectory features with cycling speed features allows for the identification of speed characteristics in different road segments. For example, on longer straight road segments, cycling speed may be relatively high and change less; while on turning segments, cycling speed may decrease, and the rate of change of speed may increase.

[0191] Based on feature matching results, a data fusion model is constructed. This model comprehensively considers cycling speed information and trajectory information, fusing the two through specific algorithms and rules. For example, a weighted average method can be used, assigning different weights based on the importance of speed and trajectory features to calculate more accurate cycling distance and speed distribution. According to the road segment divisions in the trajectory information, the cycling process is divided into multiple segments. For each segment, the data fusion model calculates the cycling distance and average speed. Simultaneously, the variation in cycling speed within the segment is considered; further analysis of the speed data distribution and trends yields more detailed speed distribution information. For example, the proportion of different speed ranges within a segment can be calculated to understand the distribution of cycling speed within that segment.

[0192] The cycling distance and speed distribution calculated by the data fusion algorithm are compared and verified with the actual situation. The accuracy of the data fusion results can be evaluated through on-site measurements or by referring to other reliable data sources, such as professional cycling recording equipment or officially released road information. Based on the verification results, the data fusion model is optimized and adjusted. If a significant deviation is found between the calculated results and the actual situation, it may be necessary to adjust the model's parameters, algorithm, or feature extraction method to improve the model's accuracy and reliability. For example, if the road segment length calculation is found to be inaccurate, the trajectory feature extraction algorithm can be optimized to improve the accuracy of the road segment length calculation.

[0193] The effect of the above technical solution is that by preprocessing the cycling speed information collected by the ebike dashboard and the motion trajectory information obtained by the smartphone GPS positioning system (such as noise reduction, filtering, coordinate transformation, etc.), the temporal and spatial consistency between different data sources is ensured, thereby greatly improving the accuracy and reliability of the final calculation results.

[0194] By extracting key features from preprocessed cycling speed and trajectory information and performing feature matching, the system can identify the speed characteristics of different road segments. The establishment of the data fusion model enables the system to calculate more accurate cycling distance and speed distribution based on these precise features, improving the intelligence and accuracy of data fusion.

[0195] Based on road segmentation, the data fusion model can provide detailed cycling data analysis for each segment, including the cycling distance, average speed, and speed variation. This detailed analysis improves the accuracy of cycling route analysis, allowing users to obtain more precise cycling information and speed distribution.

[0196] By comparing and verifying cycling data with actual conditions (such as on-site measurements or reference to other reliable data sources), deviations and errors in the data can be identified and corrected in a timely manner, thereby further improving the accuracy of the data fusion model and reducing the negative impact of inaccurate data.

[0197] By utilizing the real-time cycling distance and speed distribution calculated through a data fusion model, the system can dynamically reflect cycling status and route characteristics, providing users with more real-time and personalized cycling suggestions, thus enhancing the interactivity and real-time nature of the cycling experience. Through validation and optimization of the data fusion model, it has acquired stronger adaptive capabilities, enabling adjustments and optimizations based on different cycling environments and actual conditions. This continuous optimization capability improves the model's long-term effectiveness and applicability.

[0198] By accurately calculating the cycling distance and speed distribution for each road segment, the system can provide users with clearer route and speed planning, helping them make more scientific and reasonable cycling decisions, thereby reducing the risks and uncertainties that may be encountered during cycling.

[0199] This data fusion model can dynamically adjust according to different road characteristics and riding conditions, and has strong scalability and adaptability. It can be applied to different types of riding scenarios or environments, thus enhancing the versatility of the system.

[0200] By integrating and verifying multiple data sources, the system can ensure the credibility and reliability of cycling data, enhancing users' trust and reliance on the data analysis results, thereby further improving the user experience.

[0201] By integrating and analyzing multi-dimensional data such as cycling speed, road segment information, and turning angle, the system's ability to comprehensively and deeply analyze the cycling process has been improved, providing users with more comprehensive and detailed cycling information and enhancing the overall performance and analytical capabilities of the system.

[0202] The formula above combines two data sources: speed information collected from the dashboard and speed calculated by the GPS positioning system. This allows them to complement each other's advantages. The dashboard typically provides high instantaneous accuracy, but may be affected by environmental or equipment errors, while GPS provides relatively stable location information, but may have errors in environments such as urban high-rise buildings and tunnels. By combining these two, a more accurate average speed estimate can be obtained.

[0203] The fusion weights play a crucial role in the formula, enabling dynamic adjustment of the weights of the two data sources based on the characteristics of the road segment. For some road segments, dashboard speed may be more reliable (e.g., on flat roads), while in complex environments (e.g., densely populated urban areas), GPS speed data may be more valuable. Through adaptive fusion weights, the optimal data source can be used for each road segment.

[0204] By integrating, this formula captures speed changes over a time interval, not just instantaneous speed. This helps reduce the volatility of instantaneous data, providing a more stable and reasonable average speed estimate. The formula's design ensures that the final average speed is not merely a theoretical value, but a fusion of data from actual riding conditions. By comparing and validating with real-world scenarios, the model can be continuously optimized, thereby improving the accuracy of calculations and the reliability of practical applications.

[0205] During the calculation process, the fusion model is continuously optimized through comparison and verification with actual conditions. This not only improves the model's predictive ability but also makes it more adaptable to different environments, especially complex road conditions, thus providing cyclists with more accurate cycling data.

[0206] In one embodiment of the present invention, step S5 includes:

[0207] S51. Regularly collect user feedback to understand user satisfaction with the interaction methods and effects;

[0208] S52. Continuously optimize the user intent recognition model based on user feedback and actual interaction effects;

[0209] S53. Regularly evaluate the communication quality of the Bluetooth Mesh network and adjust the network parameters based on the evaluation results;

[0210] S54. At the same time, establish a network fault early warning mechanism. When network anomalies occur, issue an alarm in a timely manner and take corresponding measures to repair them.

[0211] The working principle of the above technical solution is as follows: User feedback is collected regularly through various methods such as questionnaires, user reviews, and online customer service. These feedback channels provide comprehensive information on user satisfaction with the interaction methods and effects. Questionnaires can be designed with targeted questions to understand users' preferences and needs for interaction methods in different scenarios; user reviews can directly reflect users' views on the overall interaction experience; and online customer service can promptly obtain specific problems and suggestions encountered by users during use. For example, a questionnaire could include the question, "How satisfied are you with the current voice interaction method?", allowing users to evaluate the interaction method based on their choices. The user intent recognition model is continuously optimized based on user feedback and actual interaction effects. When user satisfaction with a certain interaction method is found to be low, or the recognition model's accuracy is not high in certain situations, more relevant user data and interaction records are collected. This data contains user behavior and intent information in different scenarios, allowing for model retraining and adjustment, which improves the model's adaptability and recognition accuracy in various situations. For example, if users report that intent recognition is inaccurate under certain road conditions (such as bumpy roads), more riding data, user operation records, and corresponding actual intents under those road conditions are collected to train the model specifically, enabling the model to better recognize user intents under those road conditions.

[0212] Regularly evaluate the communication quality of the Bluetooth Mesh network, monitoring key indicators such as node signal strength, data transmission latency, and packet loss rate. Signal strength reflects the strength of the signal between nodes; low signal strength can lead to unstable data transmission. Data transmission latency refers to the time required for data to travel from the sending node to the receiving node; excessive latency can affect system real-time performance. Packet loss rate indicates the proportion of data packets lost during data transmission; a high packet loss rate can result in incomplete data. Monitoring these indicators provides a comprehensive understanding of the Bluetooth Mesh network's communication status. For example, use professional network monitoring tools to regularly test the communication signal strength of each node and record data transmission latency and packet loss rate. Adjust network parameters based on the evaluation results to ensure network stability and reliability. When a node's communication signal is found to be weak, its transmission power can be adjusted to enhance signal strength; or the network topology can be optimized, and communication paths can be redesigned to reduce signal interference and transmission latency. For example, a communication path that previously involved multiple nodes can be changed to direct communication, or the node's position can be adjusted to facilitate smoother communication with other nodes. These adjustments will improve the communication quality of Bluetooth Mesh networks and ensure stable and efficient data transmission.

[0213] Establish a network fault early warning mechanism. By monitoring various indicators of the Bluetooth Mesh network in real time, the system will promptly issue alerts when network anomalies occur. For example, if the communication signal strength of a node suddenly drops below a preset threshold, or if data transmission delay and packet loss rate exceed normal ranges, the system will automatically trigger an alarm and notify relevant personnel for handling. Simultaneously with issuing the alarm, appropriate repair measures will be taken. Depending on the specific fault, measures such as restarting the node, replacing hardware, or adjusting network configurations can be implemented. For instance, if the communication anomaly is caused by a node software fault, the node can be restarted; if the hardware is damaged, it needs to be replaced promptly; and if the network configuration is an issue, relevant parameters can be adjusted.

[0214] The effect of the above technical solution is that by regularly collecting user feedback and adjusting the user intent recognition model based on the feedback, the system's interaction method can be optimized in real time, improving the accuracy and adaptability of the user experience and ensuring that the system can be flexibly adjusted according to different user needs.

[0215] By analyzing user feedback and continuously optimizing the recognition model, the system can effectively improve user satisfaction with the interaction. In particular, when the recognition accuracy is low or the interaction method is not ideal, it can respond quickly and make targeted improvements, thereby enhancing user satisfaction and trust.

[0216] To address specific issues raised by users regarding different road conditions and rough surfaces (such as bumpy roads), the accuracy and robustness of the user intent recognition model were improved through retraining with more relevant data. This ensures the system maintains high recognition performance in varying environments. Continuous optimization and adjustment of the recognition model enhances the system's dynamic adaptability, enabling it to respond to changes in different usage scenarios and environments. This strengthens the system's flexibility and operability, providing customized interactive experiences for different users.

[0217] Regularly assessing network communication quality and adjusting network parameters based on the assessment results ensures the communication stability of the Bluetooth Mesh network, reduces signal interference and transmission delay, improves the overall communication quality of the system, and ensures the continuous and efficient operation of the system.

[0218] By establishing a network failure early warning mechanism, the system can issue alarms and take corrective measures in a timely manner when anomalies occur, thereby enhancing the system's self-repair capability, reducing the impact of network failures, and ensuring the continuity and stability of services.

[0219] Timely alerts and repairs reduce response time after a failure occurs, improve the timeliness and effectiveness of fault handling, and reduce service interruptions and user experience issues caused by network anomalies.

[0220] By continuously monitoring node communication signal strength, data transmission latency, packet loss rate, etc., and adjusting network parameters, data loss and latency in network communication can be effectively reduced, ensuring smooth system operation and accurate data transmission.

[0221] By optimizing the network topology and communication paths, signal interference was reduced, network communication efficiency was improved, the communication capabilities of the entire Bluetooth Mesh network were enhanced, and the overall network performance was improved.

[0222] Regularly evaluating and optimizing the performance of the user intent recognition model and the Bluetooth Mesh network enhances system reliability, ensuring that the system maintains high performance and maintainability during actual use, and providing stable and high-quality service.

[0223] One embodiment of the present invention provides an ebike intelligent human-computer interaction system based on Bluetooth Mesh networking, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to implement the ebike intelligent human-computer interaction method based on Bluetooth Mesh networking as described above.

[0224] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent human-computer interaction method for eBikes based on Bluetooth Mesh networking, characterized in that, The method includes: S1. Obtain Bluetooth communication node information of various functional modules of ebike and smart devices carried by users; perform Mesh network topology analysis on the obtained Bluetooth communication node information, and construct Bluetooth Mesh network topology for ebike usage scenarios; based on the constructed Bluetooth Mesh network topology, configure the network and set parameters for each node to form a complete ebike Bluetooth Mesh communication network. S2. During the operation of ebike, data related to ebike is collected through various functional modules and smart devices. The collected data is preprocessed and classified according to different types; feature extraction is then performed on the classified data. S3. Based on the extracted feature data, a user intent recognition model is built using machine learning algorithms; based on the recognized user intent, combined with the functions of ebike and the current usage scenario, the interaction mode is determined. S4. After determining the interaction mode, the ebike and various smart devices will coordinate and interact through the Bluetooth Mesh network; and information from different devices will be merged and processed. S5. Continuously optimize the user intent recognition model and interaction strategy based on actual user feedback and interaction effects; regularly evaluate the communication quality of the Bluetooth Mesh network and adjust the network parameters according to the evaluation results; The S4 includes: S41. After determining the interaction mode, when the user issues a control command through a smartphone, the smartphone sends the command to the corresponding functional module of the ebike through the Bluetooth Mesh network. After receiving the instruction, the S42 and ebike's functional modules execute the corresponding operations and send the results back to the smartphone for the user to view. S43. The cycling speed information collected by the ebike dashboard is fused with the motion trajectory information obtained by the smartphone GPS positioning system, and the accurate cycling distance and speed distribution are calculated through the data fusion algorithm. S44. Integrate heart rate data collected by the user's smartwatch with ebike riding status data to analyze the impact of cycling on the user's body; and provide users with more personalized health advice. S1 includes: S11. Obtain Bluetooth communication node information of ebike functional modules and smart devices through a scanning program. The node information includes node name, MAC address and signal strength. S12. Construct a Bluetooth Mesh network topology for ebike usage scenarios using graph theory algorithms; S13. A dynamic adjustment algorithm based on network topology automatically identifies new nodes when they join the network and integrates them into the existing topology according to their location and function; when existing nodes leave the network, it detects the absence of nodes and replans the network path. S14. Based on the constructed Bluetooth Mesh network topology, assign a unique network address to each node and set the node's communication parameters. S15. Through the above configuration and settings, each node can communicate according to the Mesh network protocol to form a complete ebike Bluetooth Mesh communication network.

2. The intelligent human-computer interaction method for ebike based on Bluetooth Mesh networking according to claim 1, characterized in that, S13 includes: S131. Based on the deployed node monitoring module, continuously scan devices within the Bluetooth signal range. When a new Bluetooth signal is detected, initiate the new node access process. By comparing the MAC address of the new node with the known node database, determine whether it is a newly added node. If it is a new node, further obtain its functional information. S132. Based on the RSSI value of the Bluetooth signal and combined with the known location information of the nodes, make a preliminary determination of the location of the new node; evaluate the functional information of the new node; S133. Based on the location and functional evaluation results of the new nodes, the network topology is re-planned using graph theory algorithms; S134. Send network configuration information to the new node, update the network topology, include the new node in the network, and notify other nodes about the new node's addition; at the same time, update the network path. S135. Monitor the communication status of nodes in the network in real time. When a node is detected to be unresponsive for a long time or its signal strength continues to decline, initiate the node departure detection process. Confirm whether the node has left the network by comparing it with the list of active nodes in the node database. If it is confirmed that the node has left the network, initiate the network reconstruction process and replan the network path. S136. After a node joins or leaves the network, the stability of the network is evaluated in real time, and the network topology is further optimized based on the evaluation results.

3. The ebike intelligent human-computer interaction method based on Bluetooth Mesh networking according to claim 2, characterized in that, S133 includes: Using graph structure from graph theory, existing nodes and new nodes are represented as vertices in the graph, and the connections between vertices are represented as edges; based on the known node position information, the relative positions of each vertex in the graph are initially determined; By combining the functional information of the new node, we analyze its communication needs with existing nodes. Based on graph theory algorithms, we calculate the optimal communication path between the new node and existing nodes according to the communication needs and location information between nodes. The planned communication paths are evaluated for quality, and the network topology is fine-tuned based on the results of the path quality and stability evaluation. The adjusted network topology is displayed in a visual manner, and a detailed network topology update plan is generated.

4. The ebike intelligent human-computer interaction method based on Bluetooth Mesh networking according to claim 2, characterized in that, S135 includes: The system monitors the communication status of nodes in the network in real time. When a node's signal strength continuously decreases or it becomes unresponsive for an extended period, it is marked as a potentially abnormal node. The system further analyzes the historical communication data of the abnormal node to confirm whether any anomalies exist. For nodes confirmed to be abnormal, initiate the node departure detection process; by comparing the list of active nodes in the node database, query the latest status information of the node to confirm whether it has actually left the network. If a comparison confirms that a node has left the network, it is removed from the list of active nodes, and relevant information is recorded. At the same time, the network topology is updated, and the node is marked as having left the network. Based on the changes in network topology after a node leaves, initiate the network reconstruction process; and use graph theory algorithms to replan network paths. The reconstructed network is validated and tested; based on the validation results, the network topology is further optimized, information about node departures and network reconstruction is communicated to other nodes in the network, and relevant records in the network management system are updated.

5. The intelligent human-computer interaction method for ebike based on Bluetooth Mesh networking according to claim 1, characterized in that, S2 includes: S21. During the operation of ebike, data related to ebike is collected through various functional modules and smart devices; S22. Clean the collected data, use a filtering algorithm to smooth the cleaned data, and classify the smoothed data according to different types. S23, and extract features from the classified data; S24. Through feature extraction, the original data is transformed into representative and analyzable feature data.

6. The ebike intelligent human-computer interaction method based on Bluetooth Mesh networking according to claim 1, characterized in that, The S3 includes: S31. Collect historical cycling data and user interaction records; label the collected data and determine the user intent corresponding to each data sample; S32. Select a machine learning algorithm and train the model using labeled data; during the training process, continuously adjust the model parameters, evaluate the trained model using a test dataset, and optimize the model based on the evaluation results; S33. Analyze the functions of ebike and determine the appropriate interaction method for each function; combine users' historical usage habits and preferences to recommend interaction modes; S34. Based on the identified user intent and the above-mentioned interaction mode selection criteria, determine the appropriate interaction mode.

7. The ebike intelligent human-computer interaction method based on Bluetooth Mesh networking according to claim 1, characterized in that, The S5 includes: S51. Regularly collect user feedback to understand user satisfaction with the interaction methods and effects; S52. Continuously optimize the user intent recognition model based on user feedback and actual interaction effects; S53. Regularly evaluate the communication quality of the Bluetooth Mesh network and adjust the network parameters based on the evaluation results; S54. At the same time, establish a network fault early warning mechanism. When network anomalies occur, issue an alarm in a timely manner and take corresponding measures to repair them.

8. An eBike intelligent human-computer interaction system based on Bluetooth Mesh networking, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the ebike intelligent human-computer interaction method based on Bluetooth Mesh networking as described in any one of claims 1-7.

Images