A smart cycling helmet computer control software

Abstract

The application discloses a kind of intelligent riding helmet computer control software, run in helmet main control processor or mobile terminal.Software acquisition GPS positioning, elevation, speed, road type, dangerous area, network state, GPS drift, camera definition, weather and wind noise data, calculate slope angle, downhill risk value and external perception reliability.When detecting that rider is at high speed downhill and there is dangerous area such as mountain path, city alley, stair, water area in front, control AI voice module issues deceleration warning alarm;When rain, strong wind or communication anomaly cause network perception, positioning perception or regional form perception failure, trigger the highest priority limit cycling alarm, remind rider to return starting point along original route or public safety facilities route.The application can realize cycling risk active identification, voice alarm, return prompt and log trace, improve the software safety control ability of intelligent cycling helmet.

Description

Technical Field

[0001] This invention belongs to the field of computer software, specifically a computer control software for an intelligent cycling helmet. Background Technology

[0002] With the development of smart wearable devices, cycling, and short-distance urban commuting, cycling helmets have evolved from traditional passive protective gear towards intelligent, safer, and more interactive features. Current smart cycling helmets typically integrate Bluetooth communication, music playback, video recording, GPS positioning, intercom communication, turn signal indicators, remote control operation, and AI voice interaction into a basic helmet structure. This allows riders to access navigation prompts, recording video, track data, voice communication, light warnings, and entertainment during rides, thereby improving convenience and safety. Especially in nighttime cycling, multi-person cycling, urban commuting, and outdoor cycling scenarios, smart helmets can provide auxiliary reminders to riders through lights, voice prompts, and location data, demonstrating significant application value.

[0003] However, the software control methods of existing smart cycling helmets are mostly still at the stage of single-function control or simple module stacking. For example, the Bluetooth music playback module is mainly used for audio output, the camera module is mainly used for cycling video recording, the GPS module is mainly used for positioning and track recording, the turn signal module is mainly controlled by remote control buttons, and the intercom module is mainly used for voice communication. Although the above functions can operate independently, there is a lack of a unified data fusion, scene judgment, and safety priority control mechanism between the modules, which makes it difficult for the helmet to actively identify cycling risks in complex cycling environments. When the rider is going downhill at high speed, or when there are dangerous areas such as mountain trails, urban alleys, stairs, waterways, sharp bends, or roads with cliffs ahead, existing helmets usually cannot combine GPS track, speed, elevation changes, and road risk information to actively issue deceleration and risk avoidance warnings, which can easily cause riders to be in danger due to excessive speed, insufficient reaction distance, or sudden changes in road conditions.

[0004] Furthermore, in situations such as rain, strong winds, heavy fog, network outages, GPS drift, or cameras being obscured by rain or experiencing reduced image clarity, the network perception, positioning perception, and external terrain perception capabilities of smart helmets will significantly decrease. Current technologies typically continue to operate in normal navigation, music playback, video recording, or intercom modes, lacking control strategies to proactively enter the highest priority safety protection state when perception capabilities are insufficient. They cannot promptly remind riders to abandon continuing on unknown routes, nor can they provide return-trip prompts based on the original riding trajectory or public safety facility routes.

[0005] Therefore, there is an urgent need for intelligent cycling helmet computer control software that can collect cycling status data, calculate downhill high-speed risks, assess the reliability of external perception, and combine AI voice alarms and safety priority arbitration mechanisms to achieve proactive identification, alarm prompts, and return control for dangerous scenarios on high-speed downhill slopes and scenarios where environmental perception fails. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides intelligent cycling helmet computer control software to solve the problems mentioned in the background section.

[0007] A computer control software for an intelligent cycling helmet, characterized in that: the computer control software runs on the main control processor of the intelligent cycling helmet or a mobile terminal that is communicatively connected to the intelligent cycling helmet; the intelligent cycling helmet includes a Bluetooth communication module, a camera module, a GPS positioning module, an intercom communication module, a turn signal module, a speaker module, a microphone module, a wireless network module, a remote control receiving module, and an AI voice processing module. The computer control software includes a device access management module, a riding status acquisition module, a GPS trajectory processing module, a downhill high-speed danger warning module, a forward danger area identification module, an environmental perception reliability assessment module, a riding restriction return alarm module, a safety priority arbitration module, an AI voice alarm module, and a log recording and playback module. The cycling status acquisition module is used to collect GPS positioning data, elevation data, speed data, road type data, dangerous area data, network connection status, GPS drift, camera image clarity, weather conditions, wind noise intensity, and user control commands. The GPS trajectory processing module is used to calculate the cycling distance, cycling speed, cycling direction angle, and slope angle based on GPS positioning data and elevation data; The downhill high-speed danger warning module is used to determine whether the rider is in a high-speed downhill state based on the riding speed and slope angle; The forward danger zone identification module is used to identify mountain trails, urban alleys, sharp bends, stairwells, water edges, bridge edges, cliffside roads, and road interruption areas within a preset distance in front of the rider. The environmental perception reliability assessment module is used to calculate the reliability of external perception based on network connection status, GPS drift, camera image clarity, weather conditions, and wind noise intensity. The restricted cycling return alarm module is used to trigger the highest priority restricted cycling return alarm when the reliability of external perception is lower than a preset threshold and continues for more than a preset time. The safety priority arbitration module is used to control the execution, suspension, degradation, or delay of each functional module according to the priority order of environmental perception failure limiting cycling alarm events, downhill high-speed danger warning events, turn signal events, navigation reminder events, intercom communication events, AI voice feedback events, camera recording events, and Bluetooth music playback events.

[0008] Preferably, the cycling status acquisition module establishes a cycling status table, which includes at least timestamp, latitude, longitude, plane coordinates, elevation, cycling speed, cycling direction angle, slope angle, road type, type of danger zone ahead, network status, GPS drift, camera image clarity, weather status, wind noise intensity, downhill risk value, external perception reliability, alarm level, and voice alarm content. The cycling status table is continuously updated according to a preset sampling period and serves as input data for the GPS trajectory processing module, downhill high-speed danger warning module, forward danger area identification module, environmental perception reliability assessment module, and safety priority arbitration module.

[0009] Preferably, the GPS trajectory processing module is used to convert GPS latitude and longitude into local planar coordinates with the starting point of cycling as the origin, and to calculate cycling distance, cycling speed and cycling direction angle based on the positional changes between adjacent sampling points; The local planar coordinates are used to provide a unified spatial coordinate basis for downhill judgment, judgment of dangerous areas ahead, and playback of cycling logs.

[0010] Preferably, the GPS trajectory processing module calculates the slope angle based on the elevation difference and planar distance between adjacent sampling points; When the riding speed is greater than a preset speed threshold, the slope angle is less than a negative slope threshold, and this state is maintained continuously for more than a preset number of samplings, the downhill high-speed danger warning module determines that the rider is in a high-speed downhill state.

[0011] Preferably, the forward danger zone identification module determines the forward risk search distance based on the riding speed, and identifies the danger zone within the forward risk search distance; The dangerous areas include mountain trails, urban alleyways, sharp bends, stairwells, water edges, bridge edges, roads near cliffs, road interruptions, and low-light areas. When both a high-speed downhill condition and a dangerous area exist simultaneously, the downhill high-speed danger warning module generates a downhill high-speed danger warning event.

[0012] Preferably, the downhill highway hazard warning module calculates the downhill risk value, which is obtained by weighting the speed risk factor, the slope risk factor, and the hazard area risk factor. When the downhill risk value is greater than or equal to the preset downhill risk threshold, the safety priority arbitration module sets the event as a high-priority safety event and controls the AI ​​voice alarm module and speaker module to issue a deceleration and risk avoidance alarm.

[0013] Preferably, the environmental perception reliability assessment module calculates the external perception reliability based on network reliability, GPS positioning reliability, visual perception reliability, weather reliability, and wind noise reliability. When the reliability of external perception is lower than a preset reliability threshold, the environmental perception reliability assessment module determines that the smart cycling helmet is in a state of decreased external perception capability.

[0014] Preferably, when the duration of the reduced external perception capability exceeds a preset failure determination time, the cycling return trip restriction alarm module generates an environmental perception failure cycling restriction alarm event. The environmental perception failure limiting cycling alarm event has a higher priority than downhill high-speed danger warning events, turn signal events, navigation reminder events, intercom communication events, AI voice feedback events, camera recording events, and Bluetooth music playback events.

[0015] Preferably, during the triggering of the environmental perception failure restriction riding alarm event, the safety priority arbitration module controls the Bluetooth music control module to pause music playback, controls the camera recording module to lock the current recording segment, controls the turn signal module to enter a fast flashing warning state, controls the intercom communication module to enter a low-occupancy listening state, and controls the AI ​​voice alarm module to periodically issue a restriction riding return alarm. The restricted return trip alert is used to remind riders to abandon continuing to ride on unknown sections of road and return to the starting point along the original route or a route with public safety facilities.

[0016] Preferably, the log recording and playback module stores GPS trajectory, cycling speed, elevation, slope angle, road type, dangerous area type, downhill risk value, external perception reliability, alarm level, voice alarm content, turn signal status, camera recording file index, Bluetooth music status, intercom communication status, and AI voice command records according to timestamps. After the ride ends, the log recording and playback module generates a riding safety log on the mobile terminal, which is used to replay, analyze and trace the downhill high-speed danger warning process, the environmental perception failure limiting riding alarm process, and the riding route.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention, through a GPS trajectory processing module, a downhill high-speed hazard warning module, and a forward hazard area identification module, jointly judges riding speed, elevation changes, slope angle, riding direction, and forward hazard areas. It can identify complex risk scenarios where "high speed, downhill, and dangerous terrain" coexist. Compared with ordinary speeding warnings based solely on speed, this invention can further determine whether the rider is near high-risk areas such as mountain trails, urban alleys, sharp bends, stairs, water edges, or roads near cliffs. Based on the downhill risk value, it actively triggers a voice alarm, allowing the rider to slow down or avoid the hazard area before reaching it, thereby reducing the risk of losing control at high speeds, rushing into water, hitting stairs, or veering off the road.

[0018] 2. This invention comprehensively evaluates network connectivity, GPS drift, camera image clarity, rain / fog weather conditions, and wind noise intensity through an environmental perception reliability assessment module and a ride-return warning module. When rain, strong winds, heavy fog, or communication failures prevent the helmet from reliably perceiving the external terrain, the software stops operating in normal navigation, music, or intercom modes. Instead, it sets this state as a highest-priority safety event and prompts the rider to abandon continuing to ride into unknown areas and return to the starting point along the original route or a route with public safety facilities by pausing music, locking recording, activating flashing warning lights, and providing continuous voice reminders. This improves riding safety in adverse weather conditions and scenarios where perception fails. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the downhill high-speed danger warning process in the computer control software of an intelligent cycling helmet according to the present invention. Figure 2 This is a flowchart illustrating the environmental perception failure limiting riding alarm process in the computer control software of an intelligent cycling helmet according to the present invention. Figure 3 This is a simulation diagram of the cycling path and danger zone in the computer control software of an intelligent cycling helmet according to the present invention; Figure 4 This is a schematic diagram of the simulation interface of the computer control software for an intelligent cycling helmet according to the present invention. Figure 5 This is a comprehensive control flow diagram of the computer control software for an intelligent cycling helmet according to the present invention. Detailed Implementation

[0020] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0021] like Figures 1 to 5In this embodiment, the intelligent cycling helmet computer control software refers to the control program running on the helmet's main control processor, a mobile terminal APP, or a collaborative environment of both. This software is not a single-button control interface, but a complete control system that includes cycling data acquisition, trajectory processing, slope judgment, danger zone identification, environmental perception reliability assessment, safety priority arbitration, voice alarms, light control, camera recording, and log playback.

[0022] Furthermore, let Pi be the i-th cycling state point, then Pi can be represented as: ; Where ti represents the sampling time of the i-th sampling point, lati represents latitude, loni represents longitude, xi represents the horizontal coordinate in the local plane coordinate system, yi represents the vertical coordinate in the local plane coordinate system, hi represents elevation, vi represents cycling speed, ψi represents cycling direction angle, αi represents slope angle, Ri represents downhill risk value, and Qi represents external perception reliability.

[0023] After the software starts, the device access management module first checks the working status of the Bluetooth communication module, GPS positioning module, camera module, wireless network module, microphone module, speaker module, turn signal module, and remote control receiver module. After completing the device detection, the riding status acquisition module collects riding status data according to a preset sampling period.

[0024] In this embodiment, the sampling period is preferably 1 second. Within each sampling period, the software collects data including GPS latitude and longitude, elevation, current time, camera image clarity, network connection status, GPS drift, weather conditions, wind noise intensity, road type, and hazardous area information.

[0025] The purpose of this step is to establish a unified data entry point, ensuring that subsequent speed calculations, slope calculations, downhill risk identification, and environmental perception reliability assessments are all processed based on data from the same timestamp, avoiding time synchronization issues caused by independent sampling by different modules. This step provides the foundational data for the GPS trajectory processing in the second step.

[0026] Since the GPS module outputs latitude and longitude data, while speed, distance, direction, and danger zone search need to be calculated in planar coordinates, the software first converts the latitude and longitude into local planar coordinates with the starting point of the ride as the origin.

[0027] Furthermore, let the latitude of the cycling starting point be lat0, the longitude of the starting point be lon0, the latitude of the i-th sampling point be lati, the longitude be loni, and the Earth's radius be Re. Then the planar coordinates xi and yi corresponding to the i-th sampling point can be expressed as:

[0028]

[0029] Where xi represents the lateral distance of the i-th sampling point relative to the starting point of cycling, yi represents the longitudinal distance of the i-th sampling point relative to the starting point of cycling, Re represents the Earth's radius, and lat0, lon0, lati, and loni are all calculated using radians.

[0030] The reason for this design is that latitude and longitude cannot be directly used to determine whether stairs, water bodies, or sharp bends ahead are in the rider's direction of travel. However, after being converted to local planar coordinates, distance, direction, slope, and the search area ahead can be directly calculated. This step provides the coordinate basis for the speed and direction angle calculations in the third step.

[0031] The software calculates the cycling distance based on the positional changes between two adjacent sampling points. Further, let the planar coordinates of the i-th sampling point be xi, yi, and the planar coordinates of the (i-1)-th sampling point be xi-1, yi-1, then the planar distance di between adjacent sampling points can be expressed as:

[0032] Where di represents the actual planar distance between the i-th sampling point and the (i-1)-th sampling point.

[0033] Furthermore, let the time interval between two adjacent sampling points be Δt, then the cycling speed vi corresponding to the i-th sampling point can be expressed as:

[0034] Where vi represents cycling speed in km / h; di is in meters; Δt is in seconds; and the coefficient 3.6 is used to convert m / s to km / h.

[0035] Furthermore, the riding direction angle ψi can be expressed as:

[0036] Where ψi represents the rider's current riding direction angle, and arctan2 represents the arctangent function in the four quadrants.

[0037] The purpose of this step is to obtain the rider's actual speed and direction of movement. Speed ​​is used to determine if high-speed riding is occurring, and direction angle is used to determine if the danger zone is located in front of the rider, rather than to the side or behind. This step provides technical support for the gradient calculation in step four and the forward search of the danger zone in step five.

[0038] Since the first innovation of this invention focuses on the risks of high-speed cycling downhill, the software needs to determine whether the rider is in a downhill state based on elevation changes. Further, let the elevation of the i-th sampling point be hi, and the elevation of the (i-1)-th sampling point be hi-1, then the slope angle αi corresponding to the i-th sampling point can be expressed as:

[0039] Where αi represents the slope angle; hi-hi-1 represents the elevation change between adjacent sampling points; and di represents the planar distance between adjacent sampling points.

[0040] When αi is positive, it indicates that the rider is going uphill; when αi is negative, it indicates that the rider is going downhill. To avoid misjudgment caused by errors from a single GPS point or elevation point, the software does not directly determine downhill based on a single sampling point, but requires that multiple consecutive sampling points meet the downhill condition.

[0041] Furthermore, assuming a preset negative slope threshold of Ath and a continuous downhill count of Cdown, the condition for determining a continuous downhill state can be expressed as:

[0042]

[0043] Wherein, Ath represents the negative slope threshold, preferably 5°; Cdown represents the number of consecutive samplings that meet the negative slope condition; Ndown represents the preset number of consecutive downhill samplings, preferably 3.

[0044] The reason for this design is that GPS elevation data may fluctuate briefly during cycling. If a downhill judgment is triggered based on only one change in gradient, it is easy to cause false alarms. Judging by continuous sampling points can improve the stability of downhill status recognition. This step provides technical support for the high-speed downhill status judgment in the fifth step.

[0045] After obtaining the riding speed and gradient angle, the downhill high-speed danger warning module determines whether the rider is in a high-speed downhill state. Further, assuming a preset speed threshold of Vth, the judgment condition for a high-speed downhill state can be expressed as:

[0046]

[0047]

[0048] Where vi represents the current cycling speed; Vth represents the high-speed threshold, preferably 25km / h; αi represents the slope angle; Ath represents the negative slope threshold; Cdown represents the continuous downhill count; and Ndown represents the continuous downhill count threshold.

[0049] When the above conditions are met simultaneously, the software determines that the rider is on a high-speed downhill slope. This step is designed to combine the two conditions of "high speed" and "downhill" to avoid falsely triggering downhill alarms during high-speed riding on flat roads or short-term speed fluctuations. This step provides the triggering condition for the sixth step, the forward risk search.

[0050] After determining that the rider is on a high-speed downhill slope, the software needs to assess whether there are dangerous areas ahead, such as stairs, water, sharp bends, or mountain trails. Considering that the faster the riding speed, the longer the rider's reaction distance, the forward hazard search distance is not a fixed value but is dynamically determined based on the current speed.

[0051] Furthermore, let the forward risk search distance be Li, the minimum search distance be Lmin, and the preset reaction time be Tr, then Li can be expressed as:

[0052] Where Li represents the forward risk search distance; Lmin represents the minimum search distance, preferably 20m; vi represents the current riding speed; Tr represents the preset reaction time, preferably 6s; vi / 3.6 is used to convert the speed from km / h to m / s.

[0053] The reason for this design is that at lower speeds, a shorter search distance is sufficient for the rider's reaction time; at higher speeds, if only a fixed search distance is used, it may be impossible to detect dangerous areas in advance. Dynamic search distance ensures that risks are identified earlier during high-speed downhill descents. This step provides the search range for the forward determination of dangerous areas in step seven.

[0054] The software obtains the coordinates of dangerous areas from map data, locally preset risk points, historical cycling data, or mobile terminal data. Further, let the center point of a dangerous area be Hj, with coordinates xHj and yHj, and the rider's current position be xi and yi. Then, the vector AHj from the rider to the dangerous area can be expressed as:

[0055] Furthermore, let ui be the unit vector of the rider's current direction, then ui can be represented as:

[0056] Furthermore, the forward distance Fj of the danger zone relative to the rider can be expressed as:

[0057] Furthermore, the lateral distance Bj of the danger zone relative to the rider can be expressed as:

[0058] Where Fj represents the distance of the danger zone in the rider's direction of travel; Bj represents the lateral distance of the danger zone from the rider's direction of travel.

[0059] A dangerous area is considered to be within the effective risk range in front of the rider when the following conditions are met:

[0060]

[0061] Where Li represents the forward risk search distance; Bth represents the lateral tolerance threshold, preferably 15m.

[0062] The purpose of this step is to avoid misclassifying areas such as stairs or water bodies behind or to the side of the rider as dangerous areas ahead. By combining forward and lateral distances, the spatial accuracy of dangerous area identification can be improved. This step provides the dangerous area input for the downhill risk value calculation in step eight.

[0063] To ensure that the warning is not simply a "yes or no" judgment, this invention sets a downhill risk value Ri. The downhill risk value is obtained by weighting the speed risk factor, the slope risk factor, and the danger zone risk factor.

[0064] Furthermore, let the speed risk factor be Sv, the slope risk factor be Ss, and the danger zone risk factor be Sh, then the downhill risk value Ri can be expressed as:

[0065] Where Ri represents the downhill risk value of the i-th sampling point; Sv represents the speed risk factor; Ss represents the slope risk factor; and Sh represents the danger zone risk factor.

[0066] Furthermore, the speed risk factor Sv can be expressed as:

[0067] Where vi represents the current speed; Vth represents the high-speed threshold; and Vmax represents the normalized maximum speed, preferably 45 km / h.

[0068] Furthermore, the slope risk factor Ss can be expressed as:

[0069] Where αi represents the slope angle; Ath represents the negative slope threshold; and Amax represents the normalized maximum slope value, preferably 15°.

[0070] Furthermore, the risk factor Sh for hazardous areas is assigned a value based on the type of hazardous area. It is 0.5 for mountain trails, 0.4 for urban alleyways, 0.6 for sharp bends, 0.9 for stairwell areas, 1.0 for water edges, 0.8 for bridge edges, 1.0 for roads near cliffs, 1.0 for road interruptions, and 0.5 for low-light areas.

[0071] Furthermore, assuming a preset downhill risk threshold of Rth, the triggering condition for a downhill highway hazard warning event can be expressed as:

[0072] Rth is preferably 0.65.

[0073] The reason for this design is that actual cycling dangers are not determined by a single factor. The faster the speed, the steeper the slope, and the more dangerous the area ahead, the higher the overall risk. By using a weighted risk value, multiple risk factors can be unified into a single, quantifiable value, allowing the safety priority arbitration module to decide whether to trigger a voice alarm based on this risk value. This step provides the basis for the AI ​​voice alarm in step nine.

[0074] When a high-speed downhill condition is established, and there is a dangerous area ahead, and the downhill risk value Ri is greater than or equal to the preset downhill risk threshold Rth, the safety priority arbitration module sets the event as a high-priority safety event.

[0075] In this state, the software performs the following control actions: First, the Bluetooth music control module lowers the music volume to avoid the music obscuring the safety warnings; second, the camera recording module continues recording and marks the current segment as a risky segment; third, the turn signal module maintains the current light status and enters the taillight enhancement warning status when necessary; fourth, the AI ​​voice alarm module outputs a deceleration and avoidance alarm through the speaker.

[0076] The purpose of this step is to transform the risk identification results into voice feedback that riders can directly perceive, while simultaneously reducing the music volume to ensure the alarm can be heard by the rider. This step completes the first innovation: closed-loop control from data collection and risk calculation to voice alarm.

[0077] The second innovation addresses the issue of helmets being unable to reliably perceive the external terrain in rainy, windy, foggy, or communication-stricken conditions. To this end, the environmental perception reliability assessment module collects data on network status, GPS drift, camera image clarity, weather conditions, and wind noise intensity.

[0078] The purpose of this step is to transform the question of "whether the helmet can still reliably understand the external environment" into computable data. This step provides the data source for the reliability calculation of external perception in step eleven.

[0079] Furthermore, let the network connection reliability be Qnet, the GPS positioning reliability be Qgps, the visual perception reliability be Qvis, the weather reliability be Qweather, and the wind noise reliability be Qwind, then the external perception reliability Q can be expressed as:

[0080] Where Q represents external perception reliability; Qnet represents network connection reliability; Qgps represents GPS positioning reliability; Qvis represents camera visual perception reliability; Qweather represents weather reliability; and Qwind represents wind noise reliability.

[0081] Furthermore, Qnet is set to 1 when the network connection is normal and to 0 when the network connection is interrupted.

[0082] Furthermore, assuming the current GPS drift is Dgps and the maximum allowable GPS drift is Dmax, the GPS positioning reliability Qgps can be expressed as:

[0083] Where Dgps represents the current GPS drift, and Dmax represents the maximum allowable GPS drift, preferably 30m. The larger the Dgps, the more unstable the positioning; the lower the Qgps.

[0084] Furthermore, assuming the current image sharpness score of the camera is Cvis and the standard image sharpness threshold is Cstd, the visual perception reliability Qvis can be expressed as:

[0085] Here, Cvis represents the current image sharpness score of the camera, and Cstd represents the standard image sharpness threshold, which is preferably 0.7. The lower the Cvis, the more significantly the camera is affected by rain, fog, low light, or obstruction.

[0086] Furthermore, Qweather is assigned a value based on the weather conditions. Sunny days are assigned 1, light rain is assigned 0.8, moderate rain is assigned 0.6, heavy rain is assigned 0.3, and torrential rain or heavy fog is assigned 0.1.

[0087] Furthermore, assuming the current wind noise intensity is Wnoise and the maximum allowable wind noise intensity is Wmax, the wind noise reliability Qwind can be expressed as:

[0088] Here, Wnoise represents the current wind noise intensity, and Wmax represents the maximum allowable wind noise intensity, preferably 90dB. The larger the Wnoise, the stronger the interference of strong winds on voice recognition, intercom communication, and environmental perception.

[0089] The reason for this design is that relying solely on network interruptions or blurry camera images to determine perception failure is too simplistic, while the reliability Q of external perception can comprehensively reflect the helmet's current perception capabilities of the external environment, road conditions, and communication status. This step provides a quantitative basis for the restricted riding alert in step twelve.

[0090] Furthermore, assuming the external perception reliability threshold is Qth, the duration of the perceived anomaly is T, and the preset failure judgment time is Tfail, the judgment condition for the environmental perception failure state can be expressed as:

[0091]

[0092] Where Q represents the perceived reliability; Qth represents the reliability threshold, preferably 0.45; T represents the duration during which the perceived reliability is below the threshold; and Tfail represents the preset failure determination time, preferably 10s.

[0093] When all of the above conditions are met simultaneously, the software determines that the smart cycling helmet is in a state of reduced external perception capability. This state includes at least one of the following: network perception failure, GPS positioning perception failure, map terrain perception failure, camera visual perception failure, rain obstruction perception abnormality, and strong wind noise perception abnormality.

[0094] This step is designed to avoid false alarms caused by momentary network fluctuations or brief rain obstructions. The restricted return trip alarm is only triggered when the perceived reliability falls below a threshold and persists for a certain period. This step provides the triggering condition for the highest-priority safety arbitration in step thirteen.

[0095] When the environmental perception failure state is established, the safety priority arbitration module sets the environmental perception failure riding restriction alarm event as the highest priority event. This event has a higher priority than downhill high-speed hazard warning events, turn signal events, navigation reminder events, intercom communication events, AI voice feedback events, camera recording events, and Bluetooth music playback events.

[0096] Furthermore, security priority can be expressed as:

[0097] Among them, Pfail indicates the priority of the environmental perception failure limiting riding alarm; Pdown indicates the priority of the downhill high-speed danger warning; Pturn indicates the priority of the turn signal event; Pnav indicates the priority of the navigation reminder; Ptalk indicates the priority of the intercom communication; Pai indicates the priority of the AI ​​voice feedback; Pcam indicates the priority of the camera recording prompt; and Pmusic indicates the priority of the Bluetooth music playback.

[0098] When the highest priority event is triggered, the software executes the following control actions: First, the Bluetooth music control module pauses music playback; second, the intercom communication module enters a low-occupancy listening state; third, the camera recording module locks the current recording segment; fourth, the turn signal module controls the taillight or warning light to enter a fast-flashing state; fifth, the AI ​​voice alarm module periodically issues an alarm to restrict the return trip.

[0099] The purpose of this step is to prevent the rider from relying on incomplete navigation or incorrect environmental information when the helmet cannot reliably perceive the terrain. Instead, the system proactively instructs the rider to stop proceeding into unknown areas. This step completes the second innovation: closed-loop control from detecting perception failure to issuing a warning to restrict the return journey.

[0100] When the restricted return trip alert is triggered, the software prioritizes recommending that riders return along the original route. If the mobile device can still access some map data, the software further compares the original route with the route to public safety facilities.

[0101] Furthermore, let the candidate return route be Mk, the route safety score be Sk, the public safety facility coverage factor be Fk, the lighting or guardrail factor be Lk, the map reliability factor be Mkq, and the route distance factor be Dk, then the route safety score Sk can be expressed as:

[0102] Where Sk represents the safety score of the k-th candidate return route; Fk represents the coverage of public safety facilities such as streetlights, surveillance cameras, public roads, and traffic signs; Lk represents the completeness of lighting, guardrails, or road isolation facilities; Mkq represents the reliability of map data; and Dk represents the reasonableness of route distance.

[0103] The software selects the route with the highest safety rating as the suggested return route. The method for this selection can be expressed as follows:

[0104] Where M* represents the final selected return route.

[0105] When both network and map data are unavailable, the software stops calculating candidate route scores and instead retrieves historical tracks saved in the log recording and playback module to generate a reverse return route prompt. This design is based on the fact that the original route is a path already traversed by the rider and recorded by the helmet, offering greater reversibility and interpretability compared to unknown routes. This step provides the rider with an executable return strategy in scenarios where perception fails.

[0106] Within each sampling period, the log recording and playback module saves riding status data, risk assessment results, and control actions according to timestamps.

[0107] The log entries include: time, latitude, longitude, x, y, elevation, speed, heading, slope, road_type, hazard_type, network_state, gps_drift, vision_score, weather_state, wind_noise, Q, R, alert_level, event, and warning_text.

[0108] Where Q represents the reliability of external perception, R represents the downhill risk value, alert_level represents the current alarm level, and warning_text represents the content of the voice alarm.

[0109] The purpose of this step is to ensure that every alarm is based on data. After the ride, users can replay their route on their mobile devices to see when, where, and what speed, gradient, danger zone, or sensor failure condition triggered the alarm. This step provides data support for subsequent accident investigation, riding behavior analysis, and software control effectiveness verification.

[0110] In one specific embodiment, the rider enters a continuous downhill section from a mountain road. The software collects the current speed vi as 36 km / h, and the slope angles at three consecutive sampling points are -8.2°, -9.1° and -10.0°, which meet the conditions for judging a high-speed downhill state.

[0111] At this point, the software calculates the forward risk search distance Li based on the current speed. Let Lmin be 20m and Tr be 6s, then:

[0112]

[0113] The software detected a dangerous area at the edge of the water within a range of 80m ahead; therefore, the danger zone risk factor Sh is set to 1.0. Let Vth be 25km / h, Vmax be 45km / h, Ath be 5°, and Amax be 15°, then the speed risk factor Sv and the slope risk factor Ss are respectively:

[0114]

[0115] Furthermore, the downhill risk value Ri is:

[0116]

[0117] Because Ri is greater than the preset downhill risk threshold of 0.65, the software triggers a downhill highway hazard warning event. The safety priority arbitration module lowers the music volume, the camera recording module locks the current recording segment, and the AI ​​voice alarm module outputs "Currently, this is a highway downhill section. There is a dangerous water area ahead. Please slow down and avoid it immediately."

[0118] In another specific embodiment, the rider enters an unfamiliar road in heavy rain and strong winds. The software detects a network connection interruption, so Qnet is set to 0; the GPS drift Dgps is 28m, and Dmax is 30m; the camera image clarity Cvis is 0.25, and Cstd is 0.7; the weather is heavy rain, so Qweather is set to 0.3; the wind noise intensity Wnoise is 85dB, and Wmax is 90dB.

[0119] Furthermore, the GPS positioning reliability Qgps is:

[0120]

[0121] Furthermore, the visual perception reliability Qvis is:

[0122]

[0123] Furthermore, the wind noise reliability Qwind is:

[0124]

[0125] Therefore, the reliability Q of external perception is:

[0126]

[0127] Since Q is less than the preset reliability threshold of 0.45, and this state lasts for more than 10 seconds, the software determines that the helmet is in a state of reduced external perception. The safety priority arbitration module immediately triggers the highest priority restricted cycling return alarm, pauses music playback, controls the warning lights to flash rapidly, locks the current video clip, and issues a voice prompt through the speaker: "Current rain or strong winds are affecting the helmet's perception ability. Please abandon continuing cycling and return to the starting point along the original route or the route of public safety facilities."

[0128] All embodiments of the present invention are within the scope of protection of this patent.

[0129] The embodiments of the present invention are given for the purposes of illustration and description. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A computer control software for an intelligent cycling helmet, characterized in that: The computer control software runs on the main control processor of the smart cycling helmet or on a mobile terminal that is connected to the smart cycling helmet. The smart cycling helmet includes a Bluetooth communication module, a camera module, a GPS positioning module, a walkie-talkie communication module, a turn signal module, a speaker module, a microphone module, a wireless network module, a remote control receiving module, and an AI voice processing module. The computer control software includes a device access management module, a riding status acquisition module, a GPS trajectory processing module, a downhill high-speed danger warning module, a forward danger area identification module, an environmental perception reliability assessment module, a riding restriction return alarm module, a safety priority arbitration module, an AI voice alarm module, and a log recording and playback module. The cycling status acquisition module is used to collect GPS positioning data, elevation data, speed data, road type data, dangerous area data, network connection status, GPS drift, camera image clarity, weather conditions, wind noise intensity, and user control commands. The GPS trajectory processing module is used to calculate the cycling distance, cycling speed, cycling direction angle, and slope angle based on GPS positioning data and elevation data; The downhill high-speed danger warning module is used to determine whether the rider is in a high-speed downhill state based on the riding speed and slope angle; The forward danger zone identification module is used to identify mountain trails, urban alleys, sharp bends, stairwells, water edges, bridge edges, cliffside roads, and road interruption areas within a preset distance in front of the rider. The environmental perception reliability assessment module is used to calculate the reliability of external perception based on network connection status, GPS drift, camera image clarity, weather conditions, and wind noise intensity. The restricted cycling return alarm module is used to trigger the highest priority restricted cycling return alarm when the reliability of external perception is lower than a preset threshold and continues for more than a preset time. The safety priority arbitration module is used to control the execution, suspension, degradation, or delay of each functional module according to the priority order of environmental perception failure limiting cycling alarm events, downhill high-speed danger warning events, turn signal events, navigation reminder events, intercom communication events, AI voice feedback events, camera recording events, and Bluetooth music playback events.

2. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The cycling status acquisition module establishes a cycling status table, which includes at least timestamp, latitude, longitude, plane coordinates, elevation, cycling speed, cycling direction angle, slope angle, road type, type of danger zone ahead, network status, GPS drift, camera image clarity, weather status, wind noise intensity, downhill risk value, external perception reliability, alarm level, and voice alarm content. The cycling status table is continuously updated according to a preset sampling period and serves as input data for the GPS trajectory processing module, downhill high-speed danger warning module, forward danger area identification module, environmental perception reliability assessment module, and safety priority arbitration module.

3. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The GPS trajectory processing module is used to convert GPS latitude and longitude into local planar coordinates with the starting point of cycling as the origin, and to calculate cycling distance, cycling speed and cycling direction angle based on the positional changes between adjacent sampling points; The local planar coordinates are used to provide a unified spatial coordinate basis for downhill judgment, judgment of dangerous areas ahead, and playback of cycling logs.

4. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The GPS trajectory processing module calculates the slope angle based on the elevation difference and horizontal distance between adjacent sampling points; When the riding speed is greater than a preset speed threshold, the slope angle is less than a negative slope threshold, and this state is maintained continuously for more than a preset number of samplings, the downhill high-speed danger warning module determines that the rider is in a high-speed downhill state.

5. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The forward danger zone identification module determines the forward risk search distance based on the riding speed, and identifies the danger zone within the forward risk search distance; The dangerous areas include mountain trails, urban alleyways, sharp bends, stairwells, water edges, bridge edges, roads near cliffs, road interruptions, and low-light areas. When both a high-speed downhill condition and a dangerous area exist simultaneously, the downhill high-speed danger warning module generates a downhill high-speed danger warning event.

6. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The downhill highway hazard warning module calculates the downhill risk value, which is obtained by weighting the speed risk factor, the slope risk factor, and the danger zone risk factor. When the downhill risk value is greater than or equal to the preset downhill risk threshold, the safety priority arbitration module sets the event as a high-priority safety event and controls the AI ​​voice alarm module and speaker module to issue a deceleration and risk avoidance alarm.

7. The intelligent cycling helmet computer control software according to claim 1, characterized in that: The environmental perception reliability assessment module calculates the external perception reliability based on network reliability, GPS positioning reliability, visual perception reliability, weather reliability, and wind noise reliability. When the reliability of external perception is lower than a preset reliability threshold, the environmental perception reliability assessment module determines that the smart cycling helmet is in a state of decreased external perception capability.

8. The intelligent cycling helmet computer control software according to claim 7, characterized in that: When the duration of the reduced external perception capability exceeds the preset failure judgment time, the restricted cycling return alarm module generates an environmental perception failure restricted cycling alarm event. The environmental perception failure limiting cycling alarm event has a higher priority than downhill high-speed danger warning events, turn signal events, navigation reminder events, intercom communication events, AI voice feedback events, camera recording events, and Bluetooth music playback events.

9. The intelligent cycling helmet computer control software according to claim 8, characterized in that: During the triggering of the environmental perception failure restriction riding alarm event, the safety priority arbitration module controls the Bluetooth music control module to pause music playback, controls the camera recording module to lock the current recording segment, controls the turn signal module to enter the fast flashing warning state, controls the intercom communication module to enter the low occupancy listening state, and controls the AI ​​voice alarm module to periodically issue a riding restriction return alarm. The restricted return trip alert is used to remind riders to abandon continuing to ride on unknown sections of road and return to the starting point along the original route or a route with public safety facilities.

10. The computer control software for an intelligent cycling helmet according to claim 1, characterized in that: The log recording and playback module stores GPS trajectory, cycling speed, elevation, slope angle, road type, dangerous area type, downhill risk value, external perception reliability, alarm level, voice alarm content, turn signal status, camera recording file index, Bluetooth music status, intercom communication status, and AI voice command records according to timestamps. After the ride ends, the log recording and playback module generates a riding safety log on the mobile terminal, which is used to replay, analyze and trace the downhill high-speed danger warning process, the environmental perception failure limiting riding alarm process, and the riding route.

Images