Intelligent bicycle control system and application method

Abstract

The present invention relates to the technical field of intelligent control systems, and in particular to an intelligent bicycle control system and an application method. In the present invention, a long short-term memory network is used to analyze the road conditions ahead, determine whether the road is unobstructed, and automatically adjust the electric power output to optimize the riding experience. When encountering road blockage or obstacles, a convolutional neural network is used to timely determine the obstacle threat, reduce the power output and generate an adjustment strategy. A deep Q learning algorithm is used to adjust the electric power mode according to the change of the navigation path. At the same time, according to the low battery warning, the vehicle automatically switches to the energy-saving mode, reduces the electric power output, and extends the battery life. In energy management, a Bayesian optimization algorithm is used to evaluate the battery health and remaining power in real time to ensure that the battery energy can meet the riding needs, thereby enhancing the convenience and safety of riding and the effectiveness of battery management, and ensuring the best state of the riding experience.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control systems, and in particular to an intelligent bicycle control system and an application method thereof. Background Art

[0002] The field of intelligent control system technology aims to achieve efficient, accurate management and optimization of complex systems through automated and intelligent control methods. The goal is to collect environmental and system status data in real time through sensors, actuators and control algorithms, conduct intelligent analysis and decision-making, and thus automatically adjust control strategies to improve system performance, reliability and safety.

[0003] The intelligent bicycle control system integrates sensors, control units and actuators to monitor and adjust the various performance of the bicycle in real time. The purpose is to achieve functions such as automatic speed change, brake optimization, and path planning through intelligent control means, improve the convenience and safety of riding, and ensure the best riding experience under various road conditions.

[0004] Existing technologies lack the ability to integrate and process multiple performance indicators in depth and in real time, especially their adaptability in complex environments is weak. The control algorithm is mainly based on the adjustment of a single factor and cannot fully consider the interactive effects of multiple factors such as roads, vehicles coming from behind, navigation paths and battery health. The riding experience cannot be flexibly and accurately adjusted in real time under different road conditions or complex environments. Existing technologies also fail to optimize electric power assistance and path planning at the same time, resulting in insufficient overall riding safety and may even affect riding distance and safety due to insufficient battery power. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an intelligent bicycle control system and an application method.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: The intelligent bicycle control system includes: Front road condition feedback module: Through the front road condition sensor, the road slope, road type and obstacle information are collected in real time, and data is filtered and analyzed to identify road blockage and obstacle existence. When there is no obstacle, the road is confirmed to be unobstructed and a result of good road condition ahead is generated. When there is an obstacle, the impact of the obstacle is determined and a warning result of road obstacle ahead is generated; Rear vehicle warning module: Through the rear sensor, the speed and distance data of the vehicles on the left and right are obtained, and real-time comparison and judgment are performed. When the vehicle is too close and too fast, a warning signal is generated; Navigation warning module: Receives route traffic data in real time, analyzes factors such as traffic flow and road slope, and determines whether there are blocked sections or sections that need to be detoured. If the route is congested, generates a navigation route adjustment warning signal; if there are no obstacles, generates a route unblocked result; Battery management module: collects battery power and health status in real time, calculates the remaining battery power, and makes judgments based on riding needs. When the battery power is insufficient, it generates a low battery warning message; Intelligent adjustment module: when receiving the result that the road ahead is in good condition, it uses a long short-term memory network to determine that the road is clear and increase the electric power output; when receiving the result of the road ahead obstacle warning, it determines the obstacles encountered and the road section that is not suitable for passing, reduces the power output, and generates a road condition adjustment strategy; According to the vehicle warning reminder signal, a convolutional neural network is used to adjust the vibration intensity of the vibration motor, reduce the power output, and generate a vehicle warning adjustment strategy when the threat of the vehicle coming from behind is greater; According to the navigation path adjustment reminder signal, combined with deep Q learning, when the navigation system prompts a path change, the electric power assist mode is adjusted and a path adjustment strategy is generated; According to the low battery warning information, when the battery power is low, the power is switched to energy saving mode, the electric power output is reduced, and an electric power adjustment strategy is generated; Energy management module: Based on the electric power adjustment strategy, the battery power, the road conditions ahead and the battery health information are obtained, and Bayesian optimization is used to determine whether the battery energy meets the riding needs. When the battery power is insufficient, an external charging signal is sent through the charging interface. When the battery is in poor health, a battery maintenance suggestion is generated and an energy distribution control strategy is executed.

[0007] As a further solution of the present invention, the front road condition feedback module includes: Slope recognition submodule: Based on the real-time road data collected by the front road condition sensor, the slope of the road surface is calculated. By comparing the height difference of multiple consecutive collection points, the slope change is obtained, and the road is judged whether it is too steep, and the road slope analysis results are generated; Road surface type identification submodule: Based on the ground feature data obtained by the front road condition sensor, it performs signal decoding and classification analysis, identifies the physical properties of the road surface, determines whether it affects the traffic capacity, and generates a road surface type identification result; Obstacle detection submodule: Based on the road slope analysis results and road surface type recognition results, combined with the data of the obstacle sensor, detect whether there is an obstacle ahead, and judge the type, size and impact by comparing with the preset obstacle standards. If there is no obstacle, generate a result that the road ahead is in good condition; if there is an obstacle, generate a result that the road ahead is in good condition.

[0008] As a further solution of the present invention, the rear vehicle warning module includes: Oncoming vehicle detection submodule: Through the real-time data collected by the rear sensor, it continuously reads the signal, calculates and analyzes the distance change, determines whether there is an oncoming vehicle and obtains its position, and generates the oncoming vehicle presence detection result; Speed ​​judgment submodule: based on the oncoming vehicle presence detection result, obtain the speed data of the oncoming vehicle from behind, and compare it with the set safety speed value. When the speed of the oncoming vehicle is greater than the preset threshold, generate the oncoming vehicle speed judgment result; Warning signal generation submodule: Based on the oncoming vehicle presence detection result and the oncoming vehicle speed judgment result, if the oncoming vehicle is too close and the speed is too fast, a warning reminder signal for the oncoming vehicle is generated; if the oncoming vehicle speed and distance information are normal, a normal oncoming vehicle information result is generated.

[0009] As a further solution of the present invention, the navigation warning reminder module includes: Traffic flow analysis submodule: Based on the real-time traffic data received, the traffic flow information of the route is obtained, and by continuously analyzing the traffic flow changes, whether the current route is congested is determined, and the traffic flow analysis results are generated; Road condition judgment submodule: Based on the traffic flow analysis results, combined with factors such as road slope and weather, it analyzes whether the path is blocked or needs to be detoured. If the path is blocked, it generates a path blockage judgment result; Path adjustment reminder submodule: based on the path blocking judgment result and the path unobstructed judgment result, if the path is blocked, a navigation path adjustment reminder signal is generated; if the path is unobstructed, a path unobstructed result is generated.

[0010] As a further solution of the present invention, the battery management module includes: Battery data acquisition submodule: monitors the battery status through sensors, reads the battery power, health status and temperature data in real time, records and stores the current power information, and generates battery status data; A battery power calculation submodule: based on the battery status data, obtains the remaining battery power information, compares the real-time power with the preset safety threshold, calculates the remaining power, estimates it in combination with the riding demand, and generates a calculation result of the remaining battery power; The battery warning generation submodule: based on the calculation result of the remaining battery power, detects whether the battery power is lower than a preset safety threshold, and generates a low battery warning message when it is lower than the threshold.

[0011] As a further solution of the present invention, the intelligent adjustment module includes: Road condition judgment submodule: Based on the good road condition result ahead and the road obstacle warning result ahead, a long short-term memory network is used to judge whether the road is unobstructed. If the road is unobstructed, the electric power assist output is increased; if obstacles and unsuitable road sections are encountered, the electric power assist power is reduced, and a road condition adjustment strategy is generated; Oncoming vehicle warning adjustment submodule: Based on the oncoming vehicle warning reminder signal, a convolutional neural network is used to judge the threat level of the vehicle coming from behind. If the threat of the vehicle coming from behind is large, the vibration intensity of the vibration motor is adjusted and the electric power assist power is reduced to generate an oncoming vehicle warning adjustment strategy; Path adjustment submodule: based on the navigation path adjustment reminder signal, deep Q learning is used to determine whether the navigation system prompts a path change. If the path changes, the electric power assist mode is adjusted accordingly, the power assist output is optimized, and a path adjustment strategy is generated; Battery energy-saving adjustment submodule: Based on the low battery warning information, determine whether the battery power is lower than the preset threshold. If the battery power is insufficient, switch to energy-saving mode, reduce the electric power output, and generate an electric power adjustment strategy.

[0012] As a further solution of the present invention, in the energy-saving mode, when the battery power is lower than a preset threshold, the electric assist power is reduced, unnecessary energy consumption is reduced, the current and voltage outputs are dynamically adjusted according to the remaining battery power, and the assist intensity is gradually reduced until a complete energy-saving state is entered.

[0013] As a further solution of the present invention, the energy management module includes: Battery status evaluation submodule: based on the electric power adjustment strategy, obtains the battery power, the road condition ahead and the battery health information, calculates the remaining power by reading the battery power, and compares it with the riding demand to determine whether the current battery meets the riding demand and generate a battery energy demand evaluation result; Battery charging control submodule: based on the battery energy demand evaluation result, obtain the current battery power information, determine whether the battery meets the riding demand, and if the battery power is insufficient, generate an external charging signal and perform charging operation by controlling the charging interface; Battery maintenance suggestion submodule: Based on the battery energy demand assessment results and battery health information, Bayesian optimization is used to determine whether the battery has health problems. If the battery is in poor health, battery maintenance suggestions are generated, and the user is prompted to perform battery maintenance and replacement operations based on the maintenance suggestions.

[0014] As a further solution of the present invention, the Bayesian optimization is performed according to the formula: in: is the posterior probability of the battery health state, When the battery is in a healthy state, When the battery temperature information is given, When the remaining battery power information is given, is the prior distribution of battery voltage, is the joint marginal likelihood of battery performance data, temperature, remaining capacity and voltage.

[0015] A smart bicycle application method is implemented based on the smart bicycle control system, and includes the following steps: By integrating multiple sensors and processing modules, it monitors and analyzes road conditions, obstacles, vehicles coming from behind, and battery health status in real time, intelligently determines whether the road is clear, and generates warning information based on the degree of impact of obstacles. It also detects the speed and distance of vehicles coming from behind in real time, and generates a warning signal when a dangerous vehicle is found. Adjust the electric power output according to the feedback from the road ahead, use long short-term memory network and convolutional neural network, combined with deep Q learning, to automatically adjust the electric power mode, vibration motor intensity and navigation path, while judging the traffic flow and road congestion in real time, and providing path adjustment reminders; When the battery is low on power, the system uses a Bayesian optimization algorithm to analyze the battery status, determine whether the battery meets riding needs, and trigger an external charging signal or battery maintenance recommendation when necessary.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are: 1. In the present invention, the road conditions ahead are analyzed through the long short-term memory network to determine whether the road is smooth and automatically adjust the electric power output to optimize the riding experience; 2. In the present invention, when encountering road blockage or obstacles, the convolutional neural network is used to timely judge the obstacle threat, reduce power output and generate an adjustment strategy; 3. In the present invention, the electric power assist mode is adjusted according to the change of the navigation path through the deep Q learning algorithm, and the low battery warning is automatically used to switch to the energy-saving mode, reduce the electric power assist output, and extend the endurance time; 4. In the present invention, the Bayesian optimization algorithm is used in energy management to evaluate the battery health and remaining power in real time, ensuring that the battery energy can meet the riding needs, enhancing the convenience and safety of riding and the effectiveness of battery management, and ensuring the best riding experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a system flow chart of the present invention; Figure 2 It is a schematic diagram of the system framework of the present invention; Figure 3 It is a schematic diagram of the method steps of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] See also Figure 1 The present invention provides a technical solution: an intelligent bicycle control system comprises: Front road condition feedback module: Through the front road condition sensor, the module collects the road slope, road type and obstacle information in real time, and performs data filtering and analysis to identify the road blockage and the existence of obstacles. When there are no obstacles, the module confirms that the road is unobstructed and generates a result that the road ahead is in good condition. When there are obstacles, the module determines the degree of impact of the obstacles and generates a warning result for the road ahead obstacles. Rear vehicle warning module: Through the rear sensor, the speed and distance data of the vehicles on the left and right are obtained, and real-time comparison and judgment are performed. When the vehicle is too close and too fast, a warning signal is generated; Navigation warning module: Receives route traffic data in real time, analyzes factors such as traffic flow and road slope, and determines whether there are blocked sections or sections that need to be detoured. If the route is congested, generates a navigation route adjustment warning signal; if there are no obstacles, generates a route unblocked result; Battery management module: collects battery power and health status in real time, calculates the remaining battery power, and makes judgments based on riding needs. When the battery power is insufficient, it generates a low battery warning message; Intelligent adjustment module: When receiving a result indicating that the road ahead is in good condition, the long short-term memory network is used to determine that the road is clear and increase the electric power output. When receiving a warning result indicating that the road ahead is in good condition, the module determines the obstacles encountered and the road sections that are not suitable for travel, reduces the power output, and generates a road condition adjustment strategy. According to the vehicle warning signal, a convolutional neural network is used to adjust the vibration intensity of the vibration motor, reduce the power output, and generate a vehicle warning adjustment strategy when the threat of the vehicle coming from behind is greater; According to the navigation path adjustment reminder signal, combined with deep Q learning, when the navigation system prompts the path change, the electric power assist mode is adjusted and the path adjustment strategy is generated; According to the low battery warning information, when the battery power is low, it switches to energy-saving mode, reduces the electric power output, and generates an electric power adjustment strategy; Energy management module: Based on the electric power adjustment strategy, the battery power, road conditions ahead and battery health information are obtained, and Bayesian optimization is used to determine whether the battery energy meets the riding needs. When the battery power is insufficient, an external charging signal is sent through the charging interface. When the battery is in poor health, battery maintenance suggestions are generated and energy distribution control strategies are executed.

[0020] See also Figure 2 , the front road condition feedback module includes: Slope recognition submodule: Based on the real-time road data collected by the front road condition sensor, the slope of the road surface is calculated. By comparing the height difference of multiple consecutive collection points, the slope change is obtained, and the road is judged whether it is too steep, and the road slope analysis results are generated; Road surface type identification submodule: Based on the ground feature data obtained by the front road condition sensor, it performs signal decoding and classification analysis, identifies the physical properties of the road surface, determines whether it affects the traffic capacity, and generates a road surface type identification result; Obstacle detection submodule: Based on the road slope analysis results and road type recognition results, combined with the data of the obstacle sensor, it detects whether there is an obstacle ahead, and judges the type, size and impact by comparing with the preset obstacle standards. If there is no obstacle, it generates a result that the road ahead is in good condition. If there is an obstacle, it generates a result that the road ahead is in good condition. Slope recognition submodule: Based on the real-time road data collected by the front road condition sensor, the height difference of multiple consecutive collection points is calculated using the differential method. The height difference between two adjacent points is calculated point by point and the average value is calculated to obtain the slope change of the road. Combined with the slope change range, it is judged whether the road is too steep and the road slope analysis result is generated; Road surface type identification submodule: Based on the ground feature data obtained by the front road condition sensor, signal decoding and classification analysis are performed. The ground feature data is extracted and input into the support vector machine model for training. The trained model is used to classify and distinguish the collected data, determine the physical properties of the road surface, and determine whether it affects the traffic capacity based on the category label, and generate the road surface type identification result; Obstacle detection submodule: Based on the results of road slope analysis and road type recognition, combined with the data of obstacle sensors, the real-time data obtained by the obstacle sensors is input, and the multi-layer structure of the feedforward neural network is used for feature extraction. The weights are adjusted using the back propagation algorithm to determine whether there are obstacles ahead, and by comparing with the preset obstacle standards, the type, size and impact of the obstacle are determined. If there are no obstacles, a good road condition result is generated. If there are obstacles, a road obstacle warning result is generated.

[0021] See also Figure 2, the rear vehicle warning module includes: Oncoming vehicle detection submodule: Through the real-time data collected by the rear sensor, it continuously reads the signal, calculates and analyzes the distance change, determines whether there is an oncoming vehicle and obtains its position, and generates the oncoming vehicle presence detection result; Speed ​​judgment submodule: Based on the oncoming vehicle presence detection result, obtain the speed data of the rear vehicle, and compare it with the set safe speed value. When the speed of the oncoming vehicle is greater than the preset threshold, generate the oncoming vehicle speed judgment result; Warning signal generation submodule: Based on the oncoming vehicle presence detection result and the oncoming vehicle speed judgment result, if the oncoming vehicle is too close and the speed is too fast, a warning signal for the oncoming vehicle is generated; if the oncoming vehicle speed and distance information are normal, a normal result of the oncoming vehicle information is generated; Oncoming vehicle detection submodule: Based on the real-time data collected by the rear sensor, the Kalman filter is used to continuously read and filter the signal. By predicting the current signal state and comparing it with historical data, the distance change is calculated to determine whether there is an oncoming vehicle. The relative position of the oncoming vehicle is calculated through the position estimation algorithm to generate the oncoming vehicle presence detection result. Speed ​​judgment submodule: Based on the oncoming vehicle presence detection result, the speed data of the oncoming vehicle from behind is obtained, and a threshold-based judgment algorithm is used to compare the collected speed data with the set safety speed threshold. When the speed of the oncoming vehicle exceeds the set threshold, a conditional judgment statement is used to generate the oncoming vehicle speed judgment result; Warning signal generation submodule: Based on the oncoming vehicle presence detection results and the oncoming vehicle speed judgment results, a fuzzy logic control algorithm is used to determine whether the oncoming vehicle is too close and too fast based on the speed and distance information of the oncoming vehicle through the set rules and membership functions. If the conditions are met, an oncoming vehicle warning reminder signal is generated. If the information is normal, a normal oncoming vehicle information result is generated, and an oncoming vehicle warning reminder signal is generated.

[0022] See also Figure 2 , the navigation warning module includes: Traffic flow analysis submodule: Based on the real-time traffic data received, the traffic flow information of the route is obtained, and by continuously analyzing the traffic flow changes, whether the current route is congested is determined, and the traffic flow analysis results are generated; Road condition judgment submodule: Based on the traffic flow analysis results, combined with factors such as road slope and weather, it analyzes whether the path is blocked or needs to be detoured. If the path is blocked, it generates a path blockage judgment result; Path adjustment reminder submodule: based on the path blocking judgment results and the path unobstructed judgment results, if the path is blocked, a navigation path adjustment reminder signal is generated; if the path is unobstructed, a path unobstructed result is generated; Traffic flow analysis submodule: Based on the real-time traffic data, the weighted moving average method is used to smooth the traffic flow information. By setting the window size to 5 data points, weighted calculation is performed on each data point to obtain the traffic flow information of the path, and multiple consecutive traffic data are analyzed to determine whether the current path is congested. The flow change rate is determined by setting a threshold and the traffic flow analysis results are generated; Road condition judgment submodule: Based on the traffic flow analysis results, combined with influencing factors such as road slope and weather, a multivariate linear regression analysis method is used to perform regression modeling on factors such as traffic flow, slope, and weather. Relevant data is input, and the weighted value of each factor is calculated. A comprehensive analysis is conducted to determine whether the path is blocked or needs to be detoured. If the path blocking risk value exceeds the set threshold, a path blocking judgment result is generated. Path adjustment reminder submodule: Based on the path blocking judgment results and the path unobstructed judgment results, the decision tree algorithm is used to establish a decision tree model, input the path status and traffic flow information, and judge whether the path blocking conditions are met. When the path is blocked, the rules are used to generate a path adjustment reminder signal. If the path is unobstructed, the path unobstructed result is generated, and a navigation path adjustment reminder signal is generated.

[0023] See also Figure 2 , the battery management module includes: Battery data acquisition submodule: monitors the battery status through sensors, reads the battery power, health status and temperature data in real time, records and stores the current power information, and generates battery status data; Battery power calculation submodule: Based on the battery status data, obtain the remaining battery power information, compare the real-time power with the preset safety threshold, calculate the remaining power, and make an estimate based on the riding demand to generate the remaining battery power calculation result; Battery warning generation submodule: based on the calculation result of the remaining battery power, detect whether the battery power is lower than the preset safety threshold. If it is lower than the threshold, generate a low battery warning message; Battery data acquisition submodule: monitors the battery status based on sensors, communicates with the battery management system using the I2C protocol, reads the battery power, health status and temperature data in real time, collects and records the current battery power information through multiple data points obtained every second, stores battery status data, and generates battery status data; Battery power calculation submodule: Based on the battery status data, obtain the remaining battery power information, use the Kalman filter algorithm to compare the real-time power with the preset safety threshold, and obtain more accurate power information by smoothing the power data according to the filter. Combined with the riding demand data, the linear regression model is used to estimate the remaining power and generate the battery remaining power calculation result; Battery warning generation submodule: Based on the calculation result of the remaining battery power, detect whether the battery power is lower than the preset safety threshold, and use the if-else conditional judgment statement to determine whether the current power is less than the set threshold. When the power is lower than the set threshold, a low battery warning message is generated through the system trigger.

[0024] See also Figure 2 , the intelligent adjustment module includes: Road condition judgment submodule: Based on the good road condition results and the road obstacle warning results, the long short-term memory network is used to judge whether the road is smooth. If the road is smooth, the electric power output is increased. If there are obstacles and unsuitable sections, the electric power output is reduced to generate a road condition adjustment strategy. Oncoming vehicle warning adjustment submodule: Based on the oncoming vehicle warning reminder signal, a convolutional neural network is used to determine the threat level of the vehicle coming from behind. If the threat of the vehicle coming from behind is large, the vibration intensity of the vibration motor is adjusted and the electric power assist power is reduced to generate an oncoming vehicle warning adjustment strategy; Path adjustment submodule: Based on the navigation path adjustment reminder signal, deep Q learning is used to determine whether the navigation system prompts a path change. If the path changes, the electric power assist mode is adjusted accordingly to optimize the power assist output and generate a path adjustment strategy; Battery energy-saving adjustment submodule: Based on the low battery warning information, it determines whether the battery power is lower than the preset threshold. If the battery power is insufficient, it switches to energy-saving mode, reduces the electric power output, and generates an electric power adjustment strategy; Road condition judgment submodule: Based on the good road condition results ahead and the road obstacle warning results ahead, a long short-term memory network is used to input the road condition data in the past period of time, and the memory function of the LSTM network is used to predict the road condition. The output signal of the model is combined to determine whether the road is smooth. If the road is smooth, the electric power output is increased. If there are obstacles and sections that are not suitable for traffic, the electric power output is reduced to generate a road condition adjustment strategy. Oncoming vehicle warning adjustment submodule: Based on the oncoming vehicle warning reminder signal, a convolutional neural network is used to input the distance and speed data of the rear-coming vehicle into the convolutional neural network, and the threat of the oncoming vehicle is extracted by using multiple convolutional layers. Combined with the dimension reduction operation of the pooling layer, the threat level of the oncoming vehicle is analyzed. If the threat is large, the vibration intensity of the vibration motor is adjusted, and the electric power assist power is reduced to generate an oncoming vehicle warning adjustment strategy; Path adjustment submodule: Based on the navigation path adjustment reminder signal, deep Q learning is used to explore the action space by inputting the current path status, traffic information and navigation change data. The maximum Q value is used to update the strategy to determine whether the navigation system prompts a path change. If the path changes, the electric power assist mode is adjusted and the power assist output is optimized to generate a path adjustment strategy. Battery energy-saving adjustment submodule: Based on the low battery warning information, it is determined whether the battery power is lower than the preset threshold. The if-else condition is used to judge and the threshold is set to 20%. When the battery power is lower than this threshold, it switches to energy-saving mode, reduces the electric power output, and generates an electric power adjustment strategy. Energy-saving mode: when the battery power is lower than the preset threshold, the electric assist power is reduced, unnecessary power consumption is reduced, the current and voltage output are dynamically adjusted according to the remaining battery power, and the assist intensity is gradually reduced until it enters a completely energy-saving state.

[0025] See also Figure 2 , the energy management module includes: Battery status assessment submodule: Based on the electric power adjustment strategy, obtain the battery power, road conditions ahead and battery health information, read the battery power, calculate the remaining power, and compare it with the riding demand to determine whether the current battery meets the riding demand and generate the battery energy demand assessment result; Battery charging control submodule: Based on the battery energy demand assessment results, obtain the current battery power information to determine whether the battery meets the riding requirements. If the battery power is insufficient, generate an external charging signal and perform charging operations by controlling the charging interface; Battery maintenance suggestion submodule: Based on the battery energy demand assessment results and battery health information, Bayesian optimization is used to determine whether the battery has health problems. If the battery is in poor health, battery maintenance suggestions are generated, and the user is prompted to perform battery maintenance and replacement operations based on the maintenance suggestions; Battery status assessment submodule: Based on the electric power adjustment strategy, the battery power, road conditions ahead and battery health information are obtained. A prediction algorithm based on regression analysis is used to compare the real-time battery power data with the riding demand parameters. By calculating the difference between the remaining power and the set riding demand value, it is analyzed whether the battery can meet the demand. The remaining power of the battery is accurately predicted using linear regression to generate a battery energy demand assessment result. Battery charging control submodule: Based on the battery energy demand assessment result, obtain the current battery power information, use the if-else judgment algorithm to compare the current battery power with the preset minimum power threshold, and if the battery power is insufficient, generate an external charging signal, send it to the charging interface through a control command, implement the battery charging operation, and generate an external charging signal; Battery maintenance recommendation submodule: Based on the battery energy demand assessment results and battery health information, the Bayesian optimization algorithm is used to analyze the battery health status data. Multiple influencing factors such as battery power, temperature, and number of uses are input, and the Bayesian reasoning method is used to update the probability distribution of battery health to determine whether the battery has health problems. If the battery is in poor health, battery maintenance recommendations are generated based on the optimization model. Combined with the set threshold, the user is prompted to perform battery maintenance and replacement operations to generate battery maintenance recommendations.

[0026] Bayesian optimization, according to the formula: in: is the posterior probability of the battery health state, When the battery is in a healthy state, When the battery temperature information is given, When the remaining battery power information is given, is the prior distribution of battery voltage, is the joint marginal likelihood of battery performance data, temperature, remaining capacity and voltage; Execution process: First, we need to calculate the likelihood function , that is, assuming that the battery health state is Get battery performance data The probability of Information, use As a prior distribution of the battery health state, it reflects the impact of temperature on battery health, and then takes into account the remaining battery power Impact on battery health, use To further adjust the health status assessment, the battery voltage Information through To provide additional prior data about the battery health status, after integrating the information, the posterior probability of the battery health status is obtained ,Finally, based on the posterior probability, it is determined whether the battery has health problems and provides users with specific maintenance or replacement suggestions.

[0027] See also Figure 3 , the intelligent bicycle application method comprises the following steps: By integrating multiple sensors and processing modules, it monitors and analyzes road conditions, obstacles, vehicles coming from behind, and battery health status in real time, intelligently determines whether the road is clear, and generates warning information based on the degree of impact of obstacles. It also detects the speed and distance of vehicles coming from behind in real time, and generates a warning signal when a dangerous vehicle is found. Adjust the electric power output according to the feedback from the road ahead, use long short-term memory network and convolutional neural network, combined with deep Q learning, to automatically adjust the electric power mode, vibration motor intensity and navigation path, while judging the traffic flow and road congestion in real time, and providing path adjustment reminders; When the battery is low on power, the system uses a Bayesian optimization algorithm to analyze the battery status, determine whether the battery meets riding needs, and trigger an external charging signal or battery maintenance recommendation when necessary.

[0028] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. Intelligent bicycle control system, characterized in that: The system comprises: Front road condition feedback module: Through the front road condition sensor, the road slope, road type and obstacle information are collected in real time, and data is filtered and analyzed to identify road blockage and obstacle existence. When there is no obstacle, the road is confirmed to be unobstructed and a result of good road condition ahead is generated. When there is an obstacle, the impact of the obstacle is determined and a warning result of road obstacle ahead is generated; Rear vehicle warning module: Through the rear sensor, the speed and distance data of the vehicles on the left and right are obtained, and real-time comparison and judgment are performed. When the vehicle is too close and too fast, a warning signal is generated; Navigation warning module: Receives route traffic data in real time, analyzes factors such as traffic flow and road slope, and determines whether there are blocked sections or sections that need to be detoured. If the route is congested, generates a navigation route adjustment warning signal; if there are no obstacles, generates a route unblocked result; Battery management module: collects battery power and health status in real time, calculates the remaining battery power, and makes judgments based on riding needs. When the battery power is insufficient, it generates a low battery warning message; Intelligent adjustment module: when receiving the result that the road ahead is in good condition, it uses a long short-term memory network to determine that the road is clear and increase the electric power output; when receiving the result of the road ahead obstacle warning, it determines the obstacles encountered and the road section that is not suitable for passing, reduces the power output, and generates a road condition adjustment strategy; According to the vehicle warning reminder signal, a convolutional neural network is used to adjust the vibration intensity of the vibration motor, reduce the power output, and generate a vehicle warning adjustment strategy when the threat of the vehicle coming from behind is greater; According to the navigation path adjustment reminder signal, combined with deep Q learning, when the navigation system prompts a path change, the electric power assist mode is adjusted and a path adjustment strategy is generated; According to the low battery warning information, when the battery power is low, the power is switched to energy saving mode, the electric power output is reduced, and an electric power adjustment strategy is generated; Energy management module: Based on the electric power adjustment strategy, the battery power, the road conditions ahead and the battery health information are obtained, and Bayesian optimization is used to determine whether the battery energy meets the riding needs. When the battery power is insufficient, an external charging signal is sent through the charging interface. When the battery is in poor health, a battery maintenance suggestion is generated and an energy distribution control strategy is executed.

2. The intelligent bicycle control system according to claim 1, characterized in that: The front road condition feedback module includes: Slope recognition submodule: Based on the real-time road data collected by the front road condition sensor, the slope of the road surface is calculated. By comparing the height difference of multiple consecutive collection points, the slope change is obtained, and the road is judged whether it is too steep, and the road slope analysis results are generated; Road surface type identification submodule: Based on the ground feature data obtained by the front road condition sensor, it performs signal decoding and classification analysis, identifies the physical properties of the road surface, determines whether it affects the traffic capacity, and generates a road surface type identification result; Obstacle detection submodule: Based on the road slope analysis results and road surface type recognition results, combined with the data of the obstacle sensor, detect whether there is an obstacle ahead, and judge the type, size and impact by comparing with the preset obstacle standards. If there is no obstacle, generate a result that the road ahead is in good condition; if there is an obstacle, generate a result that the road ahead is in good condition.

3. The intelligent bicycle control system according to claim 1, characterized in that: The rear vehicle warning module comprises: Oncoming vehicle detection submodule: Through the real-time data collected by the rear sensor, it continuously reads the signal, calculates and analyzes the distance change, determines whether there is an oncoming vehicle and obtains its position, and generates the oncoming vehicle presence detection result; Speed ​​judgment submodule: based on the oncoming vehicle presence detection result, obtain the speed data of the oncoming vehicle from behind, and compare it with the set safety speed value. When the speed of the oncoming vehicle is greater than the preset threshold, generate the oncoming vehicle speed judgment result; Warning signal generation submodule: Based on the oncoming vehicle presence detection result and the oncoming vehicle speed judgment result, if the oncoming vehicle is too close and the speed is too fast, a warning reminder signal for the oncoming vehicle is generated; if the oncoming vehicle speed and distance information are normal, a normal oncoming vehicle information result is generated.

4. The intelligent bicycle control system according to claim 1, characterized in that: The navigation warning reminder module includes: Traffic flow analysis submodule: Based on the real-time traffic data received, the traffic flow information of the route is obtained, and by continuously analyzing the traffic flow changes, whether the current route is congested is determined, and the traffic flow analysis results are generated; Road condition judgment submodule: Based on the traffic flow analysis results, combined with factors such as road slope and weather, it analyzes whether the path is blocked or needs to be detoured. If the path is blocked, it generates a path blockage judgment result; Path adjustment reminder submodule: based on the path blocking judgment result and the path unobstructed judgment result, if the path is blocked, a navigation path adjustment reminder signal is generated; if the path is unobstructed, a path unobstructed result is generated.

5. The intelligent bicycle control system according to claim 1, characterized in that: The battery management module comprises: Battery data acquisition submodule: monitors the battery status through sensors, reads the battery power, health status and temperature data in real time, records and stores the current power information, and generates battery status data; A battery power calculation submodule: based on the battery status data, obtains the remaining battery power information, compares the real-time power with the preset safety threshold, calculates the remaining power, estimates it in combination with the riding demand, and generates a calculation result of the remaining battery power; The battery warning generation submodule: based on the calculation result of the remaining battery power, detects whether the battery power is lower than a preset safety threshold, and generates a low battery warning message when it is lower than the threshold.

6. The intelligent bicycle control system according to claim 1, characterized in that: The intelligent adjustment module comprises: Road condition judgment submodule: Based on the good road condition result ahead and the road obstacle warning result ahead, a long short-term memory network is used to judge whether the road is unobstructed. If the road is unobstructed, the electric power assist output is increased; if obstacles and unsuitable road sections are encountered, the electric power assist power is reduced, and a road condition adjustment strategy is generated; Oncoming vehicle warning adjustment submodule: Based on the oncoming vehicle warning reminder signal, a convolutional neural network is used to judge the threat level of the vehicle coming from behind. If the threat of the vehicle coming from behind is large, the vibration intensity of the vibration motor is adjusted and the electric power assist power is reduced to generate an oncoming vehicle warning adjustment strategy; Path adjustment submodule: based on the navigation path adjustment reminder signal, deep Q learning is used to determine whether the navigation system prompts a path change. If the path changes, the electric power assist mode is adjusted accordingly, the power assist output is optimized, and a path adjustment strategy is generated; Battery energy-saving adjustment submodule: Based on the low battery warning information, determine whether the battery power is lower than a preset threshold. If the battery power is insufficient, switch to energy-saving mode, reduce the electric power output, and generate an electric power adjustment strategy.

7. The intelligent bicycle control system according to claim 6, characterized in that: In the energy-saving mode, when the battery power is lower than a preset threshold, the electric assist power is reduced, unnecessary power consumption is reduced, the current and voltage output are dynamically adjusted according to the remaining battery power, and the assist intensity is gradually reduced until a complete energy-saving state is entered.

8. The intelligent bicycle control system according to claim 1, characterized in that: The energy management module comprises: Battery status evaluation submodule: based on the electric power adjustment strategy, obtains the battery power, the road condition ahead and the battery health information, calculates the remaining power by reading the battery power, and compares it with the riding demand to determine whether the current battery meets the riding demand and generate a battery energy demand evaluation result; Battery charging control submodule: based on the battery energy demand evaluation result, obtain the current battery power information, determine whether the battery meets the riding demand, and if the battery power is insufficient, generate an external charging signal and perform charging operation by controlling the charging interface; Battery maintenance suggestion submodule: Based on the battery energy demand assessment results and battery health information, Bayesian optimization is used to determine whether the battery has health problems. If the battery is in poor health, battery maintenance suggestions are generated, and the user is prompted to perform battery maintenance and replacement operations based on the maintenance suggestions.

9. The intelligent bicycle control system according to claim 1, characterized in that: The Bayesian optimization is based on the formula: in: is the posterior probability of the battery health state, When the battery is in a healthy state, For a given battery temperature information, When the remaining battery power information is given, is the prior distribution of battery voltage, is the joint marginal likelihood of battery performance data, temperature, remaining capacity and voltage.

10. The smart bicycle application method is characterized in that: The intelligent bicycle control system according to any one of claims 1 to 9 comprises the following steps: By integrating multiple sensors and processing modules, it monitors and analyzes road conditions, obstacles, vehicles coming from behind, and battery health status in real time, intelligently determines whether the road is clear, and generates warning information based on the degree of impact of obstacles. It also detects the speed and distance of vehicles coming from behind in real time, and generates a warning signal when a dangerous vehicle is found. Adjust the electric power output according to the feedback from the road ahead, use long short-term memory network and convolutional neural network, combined with deep Q learning, to automatically adjust the electric power mode, vibration motor intensity and navigation path, while judging the traffic flow and road congestion in real time, and providing path adjustment reminders; When the battery is low on power, the system uses a Bayesian optimization algorithm to analyze the battery status, determine whether the battery meets riding needs, and trigger an external charging signal or battery maintenance recommendation when necessary.

Images