Baby carriage speed limit automatic measuring device and method based on multi-mode sensing

Abstract

The invention provides a baby carriage speed limit automatic measuring device and method based on multi-mode sensing. The device comprises a multi-mode speed measuring module, an intelligent control unit, an automatic braking system and a self-power-supply module. The multi-mode speed measurement module, the automatic braking system and the self-power-supply module are all electrically connected with the intelligent control unit, the multi-mode speed measurement module comprises a Hall sensor group and a photoelectric encoder, the automatic braking system comprises an electromagnetic brake and a mechanical interlocking device, the self-power-supply module comprises an axle generator and a super capacitor group, and the super capacitor group is electrically connected with the axle generator. And the intelligent control unit is used for executing data fusion processing and grading braking logic. A self-adaptive multi-stage braking mechanism is provided for the baby carriage, flexible braking protection is achieved through electromagnetic braking when the baby carriage speed is high, and when the baby carriage speed is high, speed reduction is forced to reach the safety level through mechanical interlocking, so that the baby carriage is further balanced between use safety and flexibility.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent stroller safety, in particular to a stroller speed limit automatic determination device and method based on multi-modal sensing. BACKGROUND

[0002] In recent years, the popularity of children's riding toy cars (strollers) has significantly increased, accompanied by the increasing prominence of children's riding safety issues. Currently, the safety speed limit of strollers mainly relies on the manual operation of physical brake devices by guardians. However, the existing technical solutions have significant limitations:

[0003] Artificial dependence and response lag: The traditional speed limit mechanism (such as manual brake) completely relies on guardians to discover overspeed in time and take action. In the case of guardians being distracted or not responding in time, there are safety hazards, and it cannot provide proactive and immediate protection.

[0004] Insufficient sensing reliability: Single sensor speed measurement solutions (such as using only an optical encoder) are easily affected by environmental interference, affecting accuracy and usability. For example, the optical encoder's grating pulse signal is easily lost when there is strong light or dirt obstruction; and a single Hall sensor may also be inaccurate in a strong magnetic field environment. Single sensor failure points are concentrated, and the system has poor robustness.

[0005] Single function: Existing solutions mostly focus on the sound and light alarm function after overspeed, lack effective automatic intervention ability. It cannot actively reduce the speed or stop the car when the guardian cannot respond immediately, and generally lacks the ability to link with remote monitoring (such as mobile phone App), limiting the timeliness and comprehensiveness of safety monitoring.

[0006] Energy sustainability challenges: Solutions that rely on external power sources or disposable / rechargeable batteries have limited use scenarios (such as no power source outdoors), require frequent replacement / charging, or battery waste pollution, increasing maintenance costs and reducing user experience. SUMMARY

[0007] Therefore, the present application provides a stroller speed limit automatic determination device and method based on multi-modal sensing to overcome the above-mentioned deficiencies in the prior art.

[0008] To achieve the above-mentioned purposes, the technical solutions adopted by the present application are as follows: The application discloses a kind of based on multi-modal sensor's stroller speed limit automation measuring device, comprising: multi-modal speed measurement module, intelligent control unit (MCU), automatic brake system and self-powered module;The multi-modal speed measurement module, automatic brake system and self-powered module are all with the intelligent control unit electric connection, the multi-modal speed measurement module includes Hall sensor group and photoelectric encoder, the automatic brake system includes electromagnetic brake and mechanical interlock device, the self-powered module includes wheel axle generator and super capacitor group, the Hall sensor group is used to collect magnetic pulse signal and carries out magneto-electric speed measurement, the photoelectric encoder is used to collect grating pulse signal and carries out photoelectric speed measurement, the intelligent control unit is used to carry out data fusion processing and hierarchical braking logic;The hierarchical braking logic includes when the stroller speed exceeds first threshold value, sound-light warning is carried out, when the speed exceeds second threshold value, the electromagnetic brake is triggered by PWM signal to execute brake, when the speed exceeds third threshold value, the mechanical interlock device is triggered by mechanical interlock trigger signal to execute brake, the first threshold value, second threshold value and third threshold value are in order from small to large;The wheel axle generator is reversely charged to super capacitor during braking process;The self-powered module is used to provide electrical energy for each module.

[0009] Further, the Hall sensor group includes two Hall sensors and two permanent magnets, the Hall sensor is installed on the frame, and the permanent magnet is symmetrically installed on the wheel rim.

[0010] Further, the device further includes a remote monitoring module for real-time display of vehicle speed and multi-stage alarm push; the remote monitoring module is connected with the intelligent control unit in communication by Bluetooth.

[0011] Further, the intelligent control unit reserves CAN bus expansion interface to support access to more sensors.

[0012] Further, the self-powered module further includes a solar patch.

[0013] Further, the device further includes a multi-modal sensor electrically connected with the intelligent control unit, the multi-modal sensor includes an accelerometer, a gyroscope, a GPS and a camera, the multi-modal sensor is used to obtain the original acceleration of the stroller, the road inclination angle, the road position coordinates and the road image, and the intelligent control unit is further used to execute a dynamic speed limit strategy according to the information obtained by the multi-modal sensor, wherein the dynamic speed limit strategy includes dividing the road terrain into flat, steep slope and rugged, and setting the first threshold value as a default speed v0 when the road terrain is flat, setting the first threshold value as v0×70% when the road terrain is steep, and setting the first threshold value as v0×50% when the road terrain is rugged.

[0014] A stroller speed limit automatic measurement method based on multi-modal sensing, for the stroller speed limit automatic measurement device based on multi-modal sensing, the method execution subject is an intelligent control unit, the method comprises: S1: collect the magnetic pulse signal of the Hall sensor group and the grating pulse signal of the photoelectric encoder; S2: adopt Kalman filtering algorithm to optimally estimate and fuse the magnetic pulse signal and the grating pulse signal, suppress transient measurement noise, and improve the position and speed measurement accuracy of the stroller; S3: calculate the real-time speed of the stroller according to the fused signal; S4: execute a staged braking logic according to the real-time speed of the stroller; the staged braking logic includes performing an audible and light warning when the speed of the stroller exceeds a first threshold, triggering the electromagnetic brake to execute braking through a PWM signal when the speed exceeds a second threshold, and triggering the mechanical interlocking device to execute braking through a mechanical interlocking trigger signal when the speed exceeds a third threshold, the first threshold, the second threshold and the third threshold are in turn from small to large.

[0015] Further, the first threshold is 10% more than the default speed v0, the second threshold is 20% more than the default speed v0, and the third threshold is 30% more than the default speed v0.

[0016] Further, the size of the first threshold in step S4 is determined according to a dynamic speed limit strategy, and the specific method is: S4.1: data acquisition: collect real-time data obtained by multi-modal sensors, the real-time data including original acceleration of the stroller, road inclination angle, road position coordinates and road image, the multi-modal sensors including accelerometer, gyroscope, GPS and camera; S4.2: preprocessing: denoising, calibration and time synchronization of the real-time data; S4.3: feature extraction: calculate the slope angle, road bump degree and road type of the current road according to the preprocessed data, the road bump degree is analyzed according to the vibration intensity of the stroller, and the road type includes asphalt, gravel and steps; S4.4: data fusion: calculate the comprehensive terrain risk by using a weighted fusion strategy, and the specific calculation formula is: Comprehensive terrain risk = w1·slope + w2·vibration intensity + w3·image classification confidence; S4.5: classify the current road terrain according to the comprehensive terrain risk, the slope angle of the current road, the road bump degree and the road type; S4.6: dynamically adjust the speed limit according to the terrain classification result, specifically: when the road surface terrain is flat, set the first threshold value as the default speed v0, when the road surface terrain is steep, set the first threshold value as v0x70%, when the road surface terrain is rugged, set the first threshold value as v0x50%.

[0017] S4.7: real-time monitoring of whether the adjusted speed limit meets the safety conditions, if an abnormality is detected, triggering the mechanical interlocking device to execute braking through the mechanical interlocking signal.

[0018] Further, the step 4.3 specifically includes: Calculate the slope angle of the current road surface through the gyroscope and GPS elevation data, analyze the road bumping degree based on the accelerometer frequency spectrum, and use the CNN (Convolutional Neural Networks) model to identify the road surface type in the camera picture.

[0019] Compared with the prior art, the present application constructs a set of stroller speed limit automation system integrating multi-modal perception, intelligent hierarchical braking decision, and energy recovery self-power supply, which has specific beneficial effects including: 1. Multi-modal fusion speed measurement: fusion of Hall sensor group (magneto speed measurement) and photoelectric encoder (photoelectric speed measurement): two sensors with different principles are redundant and complementary to each other, which significantly improves the environmental adaptability and robustness of the speed measurement system. The Hall sensor is not affected by light, and the photoelectric encoder can still work when the photoelectric encoder is disturbed by strong light or is contaminated; the photoelectric encoder provides backup when the Hall sensor is disturbed by strong magnetic field. The intelligent control unit processes the dual-channel signals through data fusion algorithms (such as weighted average, state machine or Kalman filter), outputs more accurate and reliable real-time vehicle speed, and fundamentally solves the problem of single sensor failure.

[0020] 2. Adaptive intelligent speed limit decision and automatic braking based on rules and states: design multi-level decision logic, not only based on accurate speed measurement for overspeed judgment, but also combined with riding state (such as acceleration, inclination), preset safety threshold (such as age / pattern corresponding speed limit, dangerous acceleration threshold) and environmental factors, intelligently decide whether to intervene and the intervention intensity (such as warning, hierarchical braking), realize the leap from passive warning to active safety control. The system can automatically trigger the micro electro-mechanical braking actuator (such as electromagnetic brake, micro motor driven friction plate) when detecting overspeed, abnormal acceleration, sudden stop or potential overturning risk, without the need for manual intervention by the guardian, significantly reducing the risk of accidents caused by negligence.

[0021] 3. Energy self-sustaining power supply system: integrated wheel axle generator and super capacitor group: use the kinetic energy of wheel axle rotation when the stroller is running to drive the micro generator to generate electricity; braking energy recovery mechanism: during electromagnetic braking and mechanical braking, the reverse torque or residual kinetic energy of the wheel axle drives the generator to rotate in reverse to generate electricity, recovering part of the braking energy and storing it in the super capacitor group. The super capacitor has the characteristics of fast charge and discharge and long cycle life, perfectly matching the needs of frequent braking and energy recovery. This system realizes energy self-cycling, significantly reduces or eliminates the dependence on external power supply, solves the problem of energy sustainability, and improves product convenience and environmental protection. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0023] Figure 1 The structure block diagram of a stroller speed limit automatic measurement device based on multi-modal sensing according to the present application.

[0024] Figure 2 The flow chart of a stroller speed limit automatic measurement method based on multi-modal sensing according to the present application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments.

[0026] As Figure 1As shown, a stroller speed limit automatic measuring device based on multi-modal sensing includes: a multi-modal speed measuring module 110, an intelligent control unit 120, an automatic braking system 130, and a self-power module 140; the multi-modal speed measuring module, the automatic braking system, and the self-power module are electrically connected with the intelligent control unit, the multi-modal speed measuring module 110 includes a Hall sensor group and a photoelectric encoder, the automatic braking system 130 includes an electromagnetic brake and a mechanical interlocking device, the self-power module 140 includes an axle generator and a super capacitor group, the Hall sensor group is used to collect magnetic pulse signals for magneto-electric speed measurement, the photoelectric encoder is used to collect grating pulse signals for photoelectric speed measurement, the intelligent control unit 120 is used to perform data fusion processing and hierarchical braking logic; the hierarchical braking logic includes: when the stroller speed exceeds a first threshold value, an audible and visual warning is performed; when the speed exceeds a second threshold value, the electromagnetic brake is triggered by a PWM signal to perform braking; when the speed exceeds a third threshold value, the mechanical interlocking device is triggered by a mechanical interlocking trigger signal to perform braking; the first threshold value, the second threshold value, and the third threshold value are in order from small to large; the axle generator is reversely charged to the super capacitor group during braking; and the self-power module 140 is used to provide power for each module.

[0027] When performing hierarchical braking, the intelligent control unit MCU determines that the current speed exceeds the first threshold value, and executes a first-level alarm (audible and visual alarm); when the MCU determines that the current speed exceeds the second threshold value, a second-level brake (electromagnetic brake, such as TB-1265 electromagnetic brake) is executed, the MCU outputs a PWM signal to drive the electromagnetic brake coil to generate an electromagnetic force and push the brake lever, at this time the friction plate is pressed against the wheel hub to achieve stepless braking, the response time of the electromagnetic brake is less than 0.5 seconds, and the braking force adjustment range is 0-50 N·m (controlled by the PWM signal duty cycle); when the MCU determines that the current speed exceeds the third threshold value, a third-level emergency brake (mechanical interlocking forced brake) is executed, at this time the electromagnetic brake fails (the MCU detects an abnormal current), the MCU outputs an emergency signal to trigger the mechanical interlocking device to release the spring pre-tightening force, the buckle is embedded in the gap between the wheel spokes, physically blocks rotation, and makes the stroller forced to slow down to a safe threshold. After the mechanical interlocking is triggered, the electromagnetic brake power is automatically cut off to prevent the braking force from being superimposed to cause the wheel to be locked.

[0028] By adopting a dual-redundancy braking design (electromagnetic brake as the main system and mechanical interlocking as the fault protection), multi-level braking protection is provided for the stroller, so that when the speed is high, flexible braking protection is obtained through the electromagnetic brake, and when the speed is very high, the speed is forced to slow down to a safe level through the mechanical interlocking, so that the stroller further balances between safety and flexibility in use.

[0029] Through reverse charging of the axle generator during braking, kinetic energy is recovered and stored in the super capacitor to realize energy recovery.

[0030] Specifically, the mechanical interlocking device triggers the spring damping device when overspeed (exceeding the third threshold value) to assist in deceleration, and the reset mode is to manually pull the reset rod to release the buckle lock. The photoelectric encoder calculates the rotating speed by high and low level changes. The Hall sensor is an A44E switch type Hall element. In actual operation, the Hall sensor group can be involved in an anti-electromagnetic interference mode, and the rotating speed of the stroller is calculated by the magnetic pulse frequency; the photoelectric encoder detects the rotating speed of the stroller through the grating pulse signal, and performs high-precision compensation. The intelligent control unit 120 adopts a low-power single-chip microcomputer ATmega16L to process data, integrates the Hall magnetic and photoelectric signals through Kalman filtering, and eliminates single-point errors. In addition, the intelligent control unit 120 drives the electromagnetic brake or the mechanical interlocking device to execute braking through the output of a hierarchical braking instruction (PWM signal). The wheel shaft generator converts mechanical energy into electrical energy during braking, and its rotating speed is 15 km / h, and the output power is generally greater than 5W. The super capacitor group can realize rapid charging and discharging, and has strong low-temperature adaptability.

[0031] Further, the Hall sensor group includes two Hall sensors and two permanent magnets, the Hall sensors are installed on the frame, and the permanent magnets are symmetrically installed on the wheel rims; the photoelectric encoder includes a punched code disc and a photosensitive resistor group, the punched code disc is installed on the stroller wheel shaft, and the photosensitive resistor group is installed on the frame.

[0032] By adopting the dual-redundancy design of the Hall sensor group and the photoelectric encoder, the system can still operate normally when any sensor fails, and the anti-interference ability of the device is improved.

[0033] Further, the device further includes a remote monitoring module for real-time display of the vehicle speed and push of multi-level alarms; the remote monitoring module is in communication connection with the intelligent control unit through Bluetooth.

[0034] Specifically, the remote monitoring module can be a smart phone APP, and a low-power Bluetooth 5.0 module can be integrated in the device to transmit the vehicle speed and multi-level alarm information to the guardian's smart terminal (mobile phone APP) in real time. The smart phone APP displays the vehicle speed in real time, and performs sound and light alarm or mobile phone notification when receiving multi-level alarm information.

[0035] Further, the intelligent control unit reserves a CAN bus expansion interface to support access to more sensors.

[0036] Further, the self-powered module further includes a solar patch.

[0037] The solar patch assists in power supply on rainy days, and the power supply power is 2W / m².

[0038] Embodiment 1: Children's bicycle application Two symmetrical permanent magnets are installed on each of the front and rear wheel rims, and a Hall sensor is fixed on the frame; An optical encoder is additionally installed on the front wheel shaft, with a code disc aperture of 0.8 cm and a photosensitive resistor sampling frequency of 1 kHz; The intelligent control unit receives two-way signals, and if the speed is greater than 8 km / h (preset threshold), the electromagnetic brake is triggered, and an alarm message is sent to the parent's mobile phone through Bluetooth.

[0039] Embodiment 2: Baby stroller application A pure optical encoding scheme is adopted (to avoid metal interference), with a code disc radius of 15 cm; The brake system is changed to a non-contact eddy current brake to avoid mechanical wear; The self-powered module integrates a solar patch to supplement power supply in rainy and cloudy environments.

[0040] Further, the device further comprises a multi-modal sensor electrically connected to the intelligent control unit, the multi-modal sensor comprising an accelerometer, a gyroscope, a GPS, and a camera, the multi-modal sensor being configured to acquire original acceleration of the child's vehicle, road inclination angle, road position coordinates, and road images, and the intelligent control unit is further configured to execute a dynamic speed limit strategy according to the information acquired by the multi-modal sensor, the dynamic speed limit strategy comprising classifying road terrain into flat, steep, and rugged, and setting the first threshold value as a default speed v0 (8 km / h) when the road terrain is flat, setting the first threshold value as v0 x 70% when the road terrain is steep, and setting the first threshold value as v0 x 50% when the road terrain is rugged.

[0041] In specific implementation, the default speed v0 can be 5 km / h or 8 km / h. The default speed v0 can be flexibly set according to the age or riding mode of the rider, which is not limited in the present application.

[0042] As shown in Figure 2 A multi-modal sensor-based child's vehicle speed limit automatic determination method for the foregoing multi-modal sensor-based child's vehicle speed limit automatic determination device, the method is executed by an intelligent control unit, and the method comprises: S1: Collecting magnetic pulse signals of the Hall sensor group and grating pulse signals of the optical encoder; S2: Using Kalman filtering algorithm to optimally estimate and fuse the magnetic pulse signals and grating pulse signals, to suppress instantaneous measurement noise and improve the measurement accuracy of the child's vehicle position and speed; S3: Calculating the real-time vehicle speed of the child's vehicle according to the fused signals; S4: Execute graded braking logic according to the real-time speed of the baby carriage; the graded braking logic includes sound and light warning when the speed of the baby carriage exceeds a first threshold, triggering the electromagnetic brake to perform braking through a PWM signal when the speed exceeds a second threshold, and triggering the mechanical interlocking device to perform braking through a mechanical interlocking trigger signal when the speed exceeds a third threshold, and the first threshold, the second threshold and the third threshold are arranged from small to large.

[0043] During specific implementation, the data collected by the Hall sensor group and the photoelectric encoder are sent to the intelligent control unit MCU after analog-to-digital conversion. The intelligent control unit MCU analyzes the current speed of the stroller and compares it with the threshold database to generate a graded braking instruction. The graded braking instruction drives the electromagnetic brake through an electrical signal and physically triggers the mechanical interlocking device.

[0044] Furthermore, the first threshold is 10% higher than the default speed v0, the second threshold is 20% higher than the default speed v0, and the third threshold is 30% higher than the default speed v0.

[0045] Furthermore, the size of the first threshold in step S4 is determined according to a dynamic speed limit strategy, specifically as follows: S4.1: Data Acquisition: Collecting real-time data obtained by multimodal sensors, including the stroller's original acceleration, road surface inclination angle, road surface position coordinates, and road surface image. The multimodal sensors include an accelerometer, a gyroscope, a GPS, and a camera. S4.2: Preprocessing: performing denoising, calibration and time synchronization on the real-time data; Specifically, denoising is performed by a low-pass filter, calibration is performed by eliminating zero bias, and time synchronization is performed by aligning the timestamps of multi-source data.

[0046] S4.3: Feature extraction: Calculating the slope angle, road bumpiness, and road type of the current road surface based on the preprocessed data, wherein the road bumpiness is analyzed based on the vibration intensity of the stroller, and the road types include asphalt, gravel, and steps; S4.4: Data fusion: A weighted fusion strategy is used to calculate the comprehensive terrain risk. The specific calculation formula is: Comprehensive terrain risk = w1·slope + w2·vibration intensity + w3·image classification confidence; Specifically, the weights can be set to w1 = 0.35, w2 = 0.35, and w3 = 0.3. The specific setting values ​​of the weights can be determined according to actual conditions and are not limited in this application.

[0047] S4.5: classifying the current road surface terrain according to the comprehensive terrain risk, the slope angle of the current road surface, the road surface bumping degree and the road surface type, the current road surface terrain classification including flat, steep slope and rough; For example, the terrain classification rules are as follows: Flat: slope < 5° and shock intensity < 2g and road surface type is asphalt and comprehensive terrain risk < 1.

[0048] Steep slope: slope ≥ 10° or GPS elevation rate of change exceeds threshold.

[0049] Rough: shock intensity ≥ 3g or road surface type is asphalt, gravel or step or comprehensive terrain risk ≥ 4.5.

[0050] The above terrain classification cases are only exemplary, and different adjustments can be made according to actual conditions in specific implementation, which are not limited in the present application.

[0051] S4.6: dynamically adjusting the speed limit according to the terrain classification result, specifically: when the road surface terrain is flat, the first threshold is set as the default speed v0, when the road surface terrain is steep slope, the first threshold is set as v0x70% (safe downhill speed), and when the road surface terrain is rough, the first threshold is set as v0x50% (rollover prevention limit).

[0052] S4.7: real-time monitoring whether the adjusted speed limit meets the safety condition, if an abnormality (such as motor response delay) is detected, triggering the mechanical interlocking device to perform braking through the mechanical interlocking signal.

[0053] Further, the step 4.3 specifically includes: The slope angle of the current road surface is calculated by the gyroscope and GPS elevation data, the road surface bumping degree is analyzed based on the accelerometer frequency spectrum, and the road surface type in the camera picture is identified using the CNN (Convolutional Neural Networks) model.

[0054] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A device for automatically measuring the speed limit of a baby carriage based on multimodal sensing, characterized in that: include: Multimodal speed measurement module, intelligent control unit, automatic braking system and self-powered module; the multimodal speed measurement module, automatic braking system and self-powered module are all electrically connected to the intelligent control unit, the multimodal speed measurement module includes a Hall sensor group and a photoelectric encoder, the automatic braking system includes an electromagnetic brake and a mechanical interlocking device, the self-powered module includes an axle generator and a supercapacitor group, the Hall sensor group is used to collect magnetic pulse signals for magnetoelectric speed measurement, the photoelectric encoder is used to collect grating pulse signals for photoelectric speed measurement, and the intelligent control unit is used to perform data fusion processing and graded braking logic; the graded braking logic includes sound and light warnings when the speed of the baby carriage exceeds a first threshold, triggering the electromagnetic brake to perform braking through a PWM signal when the speed exceeds a second threshold, and triggering the mechanical interlocking device to perform braking through a mechanical interlocking trigger signal when the speed exceeds a third threshold, the first threshold, the second threshold and the third threshold are sequentially arranged from small to large; the axle generator is reversely charged to the supercapacitor during the braking process; the self-powered module is used to provide power to each module.

2. The automatic speed limit measuring device for a baby carriage based on multimodal sensing according to claim 1, characterized in that: The Hall sensor group includes two Hall sensors and two permanent magnets. The Hall sensors are mounted on the vehicle frame, and the two permanent magnets are symmetrically mounted on the wheel rims. The photoelectric encoder includes a punched code disk and a photoresistor group. The punched code disk is mounted on the wheel axle, and the photoresistor group is mounted on the vehicle frame.

3. The multi-modal sensing-based automatic speed limit measuring device for a baby carriage according to claim 2, characterized in that: The device further comprises: a remote monitoring module for displaying the vehicle speed in real time and pushing multi-level alarms; the remote monitoring module is connected to the intelligent control unit via Bluetooth.

4. The multi-modal sensing-based automatic speed limit measuring device for a baby carriage according to claim 3, characterized in that: The intelligent control unit reserves a CAN bus expansion interface to support the access of more sensors.

5. The multi-modal sensing-based automatic speed limit measuring device for a baby carriage according to claim 4, characterized in that: The self-powered module further comprises a solar patch.

6. The multi-modal sensing-based automatic speed limit measuring device for a baby carriage according to claim 5, characterized in that: The device also includes a multimodal sensor electrically connected to the intelligent control unit, the multimodal sensor including an accelerometer, a gyroscope, a GPS, and a camera. The multimodal sensor is used to obtain the original acceleration of the baby carriage, the road surface inclination angle, the road surface position coordinates, and the road surface image. The intelligent control unit is also used to execute a dynamic speed limit strategy based on the information obtained by the multimodal sensor. The dynamic speed limit strategy includes dividing the road surface terrain into flat, steep slope, and rugged. When the road surface terrain is flat, the first threshold is set to the default speed v0; when the road surface terrain is a steep slope, the first threshold is set to v0×70%; when the road surface terrain is rugged, the first threshold is set to v0×50%.

7. A method for automatically determining a baby stroller's speed limit based on multimodal sensing, used in the device for automatically determining a baby stroller's speed limit based on multimodal sensing according to any one of claims 1 to 6, wherein the method is executed by an intelligent control unit, and is characterized in that: The method comprises: S1: collects the magnetic pulse signal of the Hall sensor group and the grating pulse signal of the photoelectric encoder; S2: Using the Kalman filter algorithm to optimally estimate and fuse the magnetic pulse signal and the grating pulse signal, thereby suppressing instantaneous measurement noise and improving the measurement accuracy of the baby carriage position and speed; S3: Calculate the real-time speed of the baby carriage based on the fused signal; S4: Execute graded braking logic according to the real-time speed of the baby carriage; the graded braking logic includes sound and light warning when the speed of the baby carriage exceeds a first threshold, triggering the electromagnetic brake to perform braking through a PWM signal when the speed exceeds a second threshold, and triggering the mechanical interlocking device to perform braking through a mechanical interlocking trigger signal when the speed exceeds a third threshold, and the first threshold, the second threshold and the third threshold are arranged from small to large.

8. The method for automatically determining the speed limit of a baby carriage based on multimodal sensing according to claim 7, wherein: The first threshold is 10% higher than the default speed v0, the second threshold is 20% higher than the default speed v0, and the third threshold is 30% higher than the default speed v0.

9. The method for automatically determining the speed limit of a baby carriage based on multimodal sensing according to claim 8, characterized in that: The size of the first threshold in step S4 is determined according to the dynamic speed limit strategy, and the specific method is: S4.1: Data Acquisition: Collecting real-time data obtained by multimodal sensors, including the stroller's original acceleration, road surface inclination angle, road surface position coordinates, and road surface image. The multimodal sensors include an accelerometer, a gyroscope, a GPS, and a camera. S4.2： Preprocessing: performing denoising, calibration and time synchronization on the real-time data; S4.3: Feature extraction: Calculating the slope angle, road bumpiness, and road type of the current road surface based on the preprocessed data, wherein the road bumpiness is analyzed based on the vibration intensity of the stroller, and the road types include asphalt, gravel, and steps; S4.4: Data fusion: A weighted fusion strategy is used to calculate the comprehensive terrain risk. The specific calculation formula is: Comprehensive terrain risk = w1·slope + w2·vibration intensity + w3·image classification confidence; S4.5: Classify the current road surface terrain based on the comprehensive terrain risk, the slope angle of the current road surface, the degree of road surface bumpiness, and the road surface type; S4.6: Dynamically adjust the speed limit based on the terrain classification result, specifically: when the road terrain is flat, set the first threshold to the default speed v0; when the road terrain is steep, set the first threshold to v0 × 70%; when the road terrain is rough, set the first threshold to v0 × 50%; S4.7: Monitor in real time whether the adjusted speed limit meets safety conditions. If an abnormality is detected, trigger the mechanical interlock device to perform braking through a mechanical interlock trigger signal.

10. The method for automatically determining the speed limit of a baby carriage based on multimodal sensing according to claim 9, wherein: The step 4.3 specifically includes: The slope angle of the current road surface is calculated using the gyroscope and GPS elevation data, the degree of road bumps is analyzed based on the accelerometer spectrum, and the CNN model is used to identify the road type in the camera image.

Images