Suspension type dynamic optimization active noise reduction method and robot

Abstract

The invention provides a suspension type dynamic optimization active noise reduction method. The suspension type dynamic optimization active noise reduction method specifically comprises the following steps of S1, performing initialization; s2, data acquisition; s3, noise reduction point planning; s4, performing initial noise reduction; s5, optimizing the effect; and S6, performing dynamic adjustment. The invention further provides a suspension type dynamic optimization active noise reduction robot which comprises an air bag suspension system, a drive control system, a noise reduction execution system, a sensing positioning system and a central processing system. The lower portion of the air bag suspension system is fixedly connected with the bearing cabin, and the driving control system, the noise reduction execution system, the sensing positioning system and the central processing system are integrated in the bearing cabin. According to the method, a noise reduction point dynamic optimization mechanism surrounding the user can be established, and the optimal noise reduction point location is screened in real time in combination with noise source distribution and the position of the user; and through precise reverse sound wave generation and directional emission, efficient noise cancellation of a specific area around the user is realized, and the noise reduction range and precision are improved.

Description

Technical Field

[0001] This invention relates to the field of noise reduction technology, specifically to a suspended dynamic optimization active noise reduction method and robot. Background Technology

[0002] Currently, active noise cancellation technology is mainly used in fixed scenarios (such as home theater sound arrays) or portable wearable devices (such as noise-canceling headphones), which have significant technical limitations: fixed devices cannot adjust the noise cancellation range as the user moves and can only cover a specific area; wearable devices are limited by the way they are worn, and the noise cancellation range is limited to the area around the head and cannot cope with complex environments with multiple sound sources.

[0003] In the field of levitation equipment, existing airbag levitation robots (such as CN114870231A) can achieve unpowered levitation and basic movement, but their functions are concentrated on goods transportation and environmental monitoring, without integrating active noise reduction capabilities. Some noise-reducing levitation devices (such as CN218525674U) only achieve passive noise reduction through physical sound insulation of airbags, and cannot cancel noise through reverse sound waves. At the same time, the noise reduction points of existing active noise reduction systems are mostly preset fixed positions, lacking an optimization mechanism based on the user's real-time position and the dynamic changes of noise sources, making it difficult to adapt to the noise reduction needs in dynamic scenarios.

[0004] In summary, existing technologies suffer from three major pain points: "disconnect between the suspended carrier and active noise reduction function", "noise reduction point cannot dynamically adapt to environmental changes", and "noise reduction range is not synchronized with user movement". There is an urgent need for an integrated solution that combines three-dimensional levitation and movement capabilities, dynamic noise reduction point optimization function and precise reverse sound wave emission capability. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a suspended dynamic optimization active noise reduction method and robot. It can establish a dynamic optimization mechanism for noise reduction points around the user, combining the distribution of noise sources with the user's location to select the best noise reduction point in real time. Furthermore, through precise reverse sound wave generation and directional emission, it can achieve efficient noise cancellation in specific areas around the user, thereby improving the noise reduction range and accuracy.

[0006] To achieve the above technical solution, the present invention provides a suspended dynamic optimization active noise reduction method, which specifically includes the following steps:

[0007] S1. Initialization: The airbag suspension system drives the robot to levitate to a height of 1.2-1.6m, the perception and positioning system establishes the user coordinate system, and the noise reduction execution system performs a self-test.

[0008] S2, Data Acquisition: The noise acquisition unit obtains noise source parameters, and the user tracking module updates the user's head coordinates;

[0009] S3, Noise Reduction Point Planning: Generate a set of candidate noise reduction points with a radius of 0.4-0.8m centered on the user, and calculate the potential value of each point;

[0010] S4. Initial noise reduction: Drive the robot to move to the noise reduction point with the highest potential value and emit a reverse sound wave;

[0011] S5. Effect Optimization: The error microphone detects the noise reduction effect, and if it does not meet the standard, it will perform a second optimization.

[0012] S6. Dynamic Adjustment: If a user displacement ≥ 0.3m or a change in noise is detected, return to step S2 to replan.

[0013] Preferably, the specific process of step S1 includes:

[0014] S11. Inflate the outer main airbag to the preset pressure, and the robot suspends at an initial height of 1.2-1.6m to adapt to the user's sitting / standing posture;

[0015] S12. The perception and positioning system is activated, and Bluetooth positioning and visual recognition are jointly calibrated to establish a spatial coordinate system with the user's initial position as the origin.

[0016] S13. The noise reduction system performs a self-test, and the microphone array and speaker array are initialized to ensure that the signal path is normal.

[0017] Preferably, in step S2, after the noise acquisition unit obtains the noise source parameters, it calculates the noise source parameters using spherical near-field acoustic holography: azimuth angle, elevation angle, dominant frequency, and sound pressure level; after the user tracking module updates the user's head coordinates, it records the user's movement trend.

[0018] Preferably, the specific process of step S3 includes:

[0019] S31. The central processing system generates a set of candidate noise reduction points based on the "noise source-user relative position": A spherical region with a radius of 0.4-0.8m is formed, centered on the user's head, and divided into a grid with a step size of 0.1m, resulting in ≤500 candidate points (P1, P2, ..., P...). n );

[0020] S32. Calculate the "noise reduction potential value" (value 0-100) for each candidate point. The calculation model is: potential value = 60% × (1 - sound wave attenuation coefficient) + 30% × (1 - environmental interference coefficient) + 10% × (1 - moving path length / maximum path length).

[0021] Among them, the sound wave attenuation coefficient is calculated based on the distance from the noise source / user. For every 0.1m increase in distance, the attenuation coefficient increases by 0.05. The environmental interference coefficient is calculated based on the degree of obstruction by surrounding obstacles. For every 10% increase in the proportion of obstructed area, the interference coefficient increases by 0.1.

[0022] Preferably, step S4 specifically includes:

[0023] S41. The central processing system selects the top 3 candidate points with the highest potential value and generates the optimal movement path through the motion path planning algorithm.

[0024] S42. The drive control system drives the robot to move to the initial noise reduction point (P_max) with the highest potential value, with a movement error ≤0.05m;

[0025] S43. The reverse acoustic wave generation unit generates a reverse acoustic wave signal based on real-time noise data using the FxLMS algorithm. The directional emission unit focuses the emission and targets the user's head area.

[0026] Preferably, the specific process of step S5 is as follows:

[0027] S51. The "error microphone" in the noise acquisition unit detects the sound pressure level after noise reduction. If the sound pressure level reduction is ≥20dB, it meets the standard and the current noise reduction point is maintained; if the reduction is <20dB, it does not meet the standard and triggers secondary optimization.

[0028] S52. Move the remaining candidate points sequentially from high to low according to the potential value, with a step size of 0.1m and a moving time of ≤0.5s / point. Stay at each candidate point for 0.3s, repeat the reverse sound wave emission and effect detection until a qualified noise reduction point is found.

[0029] Preferably, the specific process in step S6 is as follows:

[0030] S61. Real-time monitoring trigger conditions: If the user displacement is ≥0.3m, or the main frequency of the noise source fluctuates by ±30Hz or the sound pressure level changes by ±5dB, return to step S2, re-collect data and plan noise reduction points.

[0031] S62. When there is no triggering condition, the noise reduction potential value is updated every 5 seconds, and the robot position is fine-tuned by ≤0.05m to maintain the best noise reduction effect.

[0032] This invention also provides a suspended dynamic optimization active noise reduction robot for implementing the above-mentioned noise reduction method, specifically including: an airbag suspension system, a drive control system, a noise reduction execution system, a perception and positioning system, and a central processing system; the airbag suspension system is fixedly connected to a carrier chamber below, and the drive control system, noise reduction execution system, perception and positioning system, and central processing system are integrated in the carrier chamber; the airbag suspension system has a double-layer sealed structure, the outer main airbag is filled with helium to provide basic buoyancy, and the inner adjusting airbag is connected to a miniature electric air pump and a pressure sensor, adjusting lifting and levitation by inflation and deflation; the drive control system includes... Six sets of hexagonally distributed miniature vector thrusters and attitude sensors achieve horizontal movement and attitude correction through differential control; the noise reduction execution system includes a spherical microphone array, an improved FxLMS algorithm reverse acoustic wave generation unit, and a directional speaker array to achieve noise acquisition, reverse acoustic wave generation, and focused emission; the perception and positioning system includes a Bluetooth-AoA+RGB-D vision-integrated user tracking module and a ring-shaped ultrasonic obstacle avoidance module; the central processing system receives data from each module, uses a noise reduction point optimization algorithm and a path planning algorithm to output control commands to other systems, and achieves dynamic noise reduction.

[0033] Preferably, the spherical microphone array is a spherical structure with a diameter of 50mm composed of 8 MEMS microphones, which reconstructs a three-dimensional sound field based on spherical near-field acoustic holography technology, with a noise source localization error of ≤3° and a spectrum analysis range of 20-2000Hz.

[0034] Preferably, the directional loudspeaker array is a square array composed of four piezoelectric directional loudspeakers, and the direction of the sound waves is controlled by phased array technology. The diameter of the focusing area is ≤0.2m, and the frequency response range is 50-1500Hz.

[0035] The beneficial effects of the suspended dynamic optimization active noise reduction method and robot provided by this invention are as follows:

[0036] (1) The floating dynamic optimization active noise reduction method provided by the present invention can establish a dynamic optimization mechanism for noise reduction points around the user, and combine the noise source distribution and the user's location to screen the best noise reduction point in real time; and achieve efficient noise cancellation in a specific area around the user through precise reverse sound wave generation and directional emission, thereby improving the noise reduction range and accuracy.

[0037] (2) Significantly improved noise reduction accuracy: The spherical microphone array realizes three-dimensional sound field reconstruction, the directional speaker array focuses the reverse sound wave, and with the dynamic optimization algorithm, the noise in the target area around the user is reduced by 20-30dB, which is 30% better than the traditional portable noise-canceling headphones (noise reduction of 15-20dB);

[0038] (3) High intelligence and autonomy: It integrates multimodal user tracking and environmental obstacle avoidance, and can automatically respond to user movement and noise changes without manual intervention, dynamically adjust the noise reduction point, and improve user experience satisfaction by 90% compared with manually adjusted noise reduction devices.

[0039] (4) Energy consumption and stability optimization: The double-layer airbag design reduces the energy consumption of suspension (compared to pure propeller suspension robots, the battery life is increased by 50%, up to 3-4 hours), and the attitude sensor and PID control ensure suspension stability (tilt angle error ≤1°), avoiding the impact of shaking on the noise reduction effect;

[0040] (5) Breakthrough in spatial adaptability: Through the combination of airbag suspension and vector thruster, three-dimensional spatial movement (lifting + horizontal + turning) is achieved, freeing it from the limitations of fixed installation and adapting it to various scenarios such as home, office, and coffee shop. The applicable range is increased by more than 80% compared with traditional fixed noise reduction devices. Attached Figure Description

[0041] Figure 1 This is a flowchart of the noise reduction method in this invention.

[0042] Figure 2 A schematic diagram of the optimal noise reduction area.

[0043] Figure 3 This is a schematic diagram of the noise reduction robot in this invention.

[0044] In the diagram: 1. Airbag suspension system; 2. Carrier cabin; 3. Drive control system; 4. Noise reduction execution system; 5. Sensing and positioning system; 6. Central processing system; 7. Reverse acoustic wave generation unit; 8. RGB-D camera; 9. Ultrasonic sensor. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0046] Example 1: A floating dynamic optimization active noise reduction method.

[0047] Reference Figure 1 and Figure 2 As shown, a floating dynamic optimization active noise reduction method specifically includes the following steps:

[0048] S1. System initialization (time ≤ 3s)

[0049] S11. The outer main airbag is inflated to the preset pressure (0.12MPa), and the robot floats to the initial height (1.2-1.6m, adaptable to the user's sitting / standing posture);

[0050] S12. The perception and positioning system is activated, and Bluetooth positioning and visual recognition are jointly calibrated to establish a spatial coordinate system (X / Y / Z axes) with the user's initial position as the origin.

[0051] S13. The noise reduction system performs a self-test, and the microphone array and speaker array are initialized to ensure that the signal path is normal.

[0052] S2. Noise and location data acquisition (sampling period 0.1s)

[0053] S21. The noise acquisition unit collects ambient noise and calculates the noise source parameters using spherical near-field acoustic holography: azimuth (0-360°), elevation (-30°—30°), dominant frequency (100-600Hz low-frequency noise), and sound pressure level (65-80dB).

[0054] S22. The user tracking module updates the user's head coordinates (X1, Y1, Z1) in real time and records the user's movement trend (if the displacement is ≥0.2m for 3 consecutive cycles, it is determined to be a movement state).

[0055] S3, Noise Reduction Point Set Programming and Potential Value Calculation

[0056] S31. The central processing system generates a set of candidate noise reduction points based on the "noise source-user relative position": A spherical region with a radius of 0.4-0.8m is formed, centered on the user's head, and divided into a grid with a step size of 0.1m, resulting in ≤500 candidate points (P1, P2, ..., P...). n );

[0057] S32. Calculate the "noise reduction potential value" (value 0-100) for each candidate point. The calculation model is: potential value = 60% × (1 - sound wave attenuation coefficient) + 30% × (1 - environmental interference coefficient) + 10% × (1 - moving path length / maximum path length).

[0058] Among them, the sound wave attenuation coefficient is calculated based on the distance from the noise source / user (the attenuation coefficient increases by 0.05 for every 0.1m increase in distance); the environmental interference coefficient is calculated based on the degree of obstruction by surrounding obstacles (the interference coefficient increases by 0.1 for every 10% increase in the proportion of obstruction area).

[0059] S4. Initial noise reduction point movement and reverse sound wave emission

[0060] S41. The central processing system selects the top 3 candidate points with the highest potential value and generates the optimal movement path (avoiding obstacles and with the shortest path length) through the motion path planning algorithm.

[0061] S42. The drive control system drives the robot to move to the initial noise reduction point (P_max) with the highest potential value, with a movement error ≤0.05m;

[0062] S43. The reverse acoustic wave generation unit generates a reverse acoustic wave signal based on real-time noise data using the FxLMS algorithm. The directional emission unit focuses the emission and targets the user's head area.

[0063] S5, Noise Reduction Effect Detection and Secondary Optimization

[0064] S51. The "error microphones" (two microphones, located 0.1m to the side of the user's head) in the noise acquisition unit detect the sound pressure level after noise reduction. If the sound pressure level reduction is ≥20dB (meets the standard), the current noise reduction point is maintained; if the reduction is <20dB (does not meet the standard), a secondary optimization is triggered.

[0065] S52. Move to the remaining candidate points in descending order of potential value (step size 0.1m, moving time ≤ 0.5s / point);

[0066] S53. Hold each candidate point for 0.3s, repeat the reverse sound wave emission and effect detection until a satisfactory noise reduction point is found (maximum 5 iterations).

[0067] S6. Dynamic Adjustment and Iterative Optimization

[0068] S61, if the noise level changes by ±5dB, return to step S2, re-acquire data and plan the noise reduction point;

[0069] S62. When there is no triggering condition, the noise reduction potential value is updated every 5 seconds, and the robot position is finely adjusted (≤0.05m) to maintain the best noise reduction effect.

[0070] The main algorithms involved in this invention are as follows:

[0071] (1) Noise reduction point optimization algorithm:

[0072] Implemented in Python, the algorithm follows a process of "grid generation - potential value calculation - optimal selection" and uses the NumPy library for matrix operations, with an optimization time of ≤0.3s.

[0073] In the potential value calculation, the sound wave attenuation coefficient is based on the "inverse square law of distance", and the environmental interference coefficient is calculated by recognizing the area ratio of obstacles through the camera.

[0074] (2) Motion path planning algorithm:

[0075] Based on the A* algorithm improvement, an "obstacle avoidance cost" (the closer the obstacle is, the higher the cost) is introduced to generate the shortest path. Path tracking adopts PID control, and the position error is ≤0.05m.

[0076] (3) Improved FxLMS algorithm:

[0077] Based on the C language and ported to the STM32H750 processor, a variable step size factor (step size range 0.01-0.1) is introduced, which improves the convergence speed by 30% compared with the traditional FxLMS, ​​and the reverse acoustic wave generation delay is ≤0.05s.

[0078] The floating dynamic optimization active noise reduction method provided by this invention can establish a dynamic optimization mechanism for noise reduction points around the user, and select the best noise reduction point in real time by combining the distribution of noise sources and the user's location; and achieve efficient noise cancellation in specific areas around the user through precise reverse sound wave generation and directional emission, thereby improving the noise reduction range and accuracy.

[0079] Example 2: A suspended dynamic optimization active noise reduction robot.

[0080] Reference Figure 3 As shown, a suspended dynamic optimization active noise reduction robot is used to implement the noise reduction method in Embodiment 1, specifically including: an airbag suspension system 1, a carrier cabin 2, a drive control system 3, a noise reduction execution system 4, a perception and positioning system 5, and a central processing system 6; the carrier cabin 2 is fixedly connected below the airbag suspension system 1, and the drive control system 3, the noise reduction execution system 4, the perception and positioning system 5, and the central processing system 6 are integrated in the carrier cabin 1.

[0081] Airbag Suspension System 1: Employs a double-layer sealed structure consisting of an outer main airbag and an inner adjusting airbag. The outer main airbag uses a polyimide-rubber composite membrane, filled with helium (purity ≥99.9%) to provide basic buoyancy. Its volume is designed based on the robot's total mass (e.g., 0.4 m³ for a total mass of 500g). 3 The inner regulating airbag connects a miniature electric air pump (model: DP-100, flow rate 100mL / min) and a pressure sensor (accuracy ±0.1kPa). By inflating and deflating the airbag, the overall density is adjusted to achieve a lifting speed of 0.1-0.5m / s and a fixed-point suspension accuracy of ±0.05m. The outer surface of the airbag is wrapped with 0.5mm thick flexible sound insulation cotton (noise reduction coefficient ≥0.8) to suppress interference noise generated by airbag vibration.

[0082] Drive control system 3: It consists of 6 sets of miniature vector thrusters (30mm in diameter, rated thrust 5N), attitude sensors (MPU9250, including a three-axis gyroscope / accelerometer / magnetometer) and drive circuits; the 6 sets of thrusters are distributed in a regular hexagonal pattern at the bottom of the carrier hull 2, and achieve a horizontal (X / Y axis) moving speed of 0.2-0.8m / s and 360° turning through differential control; the attitude sensors collect tilt angle (range ±15°) and angular velocity (range ±2000° / s) data in real time, and feed them back to the central processing system 6, which corrects the attitude through PID algorithm to ensure suspension stability (tilt angle error ≤1°).

[0083] Noise reduction execution system 4: includes a noise acquisition unit, a reverse sound wave generation unit 7, and a directional emission unit.

[0084] Noise acquisition unit: It adopts a spherical array (50mm in diameter) composed of 8 MEMS microphones (model: SGM3770, signal-to-noise ratio 72dB), and realizes 360° three-dimensional sound field reconstruction based on spherical near-field acoustic holography technology. The noise source positioning error is ≤3° and the spectrum analysis range is 20-2000Hz.

[0085] Reverse acoustic wave generation unit 7: Equipped with an STM32H750 processor, it runs an improved FxLMS algorithm (introducing a variable step size factor, which improves the convergence speed by 30%) to perform real-time FFT transformation on the noise signal (sampling rate 48kHz) and generate a reverse acoustic wave signal with a phase difference of 180° and an amplitude error of ≤5%.

[0086] Directional transmission unit: It consists of an array of 4 piezoelectric directional loudspeakers (frequency response range 50-1500Hz, directivity ≥15dB), which control the direction of sound wave propagation through phased array technology. The diameter of the focusing area is ≤0.2m, ensuring that the reverse sound wave accurately acts on the user's surroundings.

[0087] Perception and positioning system 5: includes a user tracking module and an environmental obstacle avoidance module.

[0088] User tracking module: integrates Bluetooth 5.1 AoA positioning (positioning accuracy ±0.1m) and RGB-D camera 8 (model: D435i) visual recognition to obtain the user's head coordinates in real time (update frequency 10Hz).

[0089] Environmental obstacle avoidance module: It uses 8 ultrasonic sensors 9 (detection distance 0.02-4m, response time ≤10ms) arranged in a ring to detect surrounding obstacles and trigger obstacle avoidance commands (minimum safe distance 0.3m).

[0090] Central Processing System 6: It adopts an ARM Cortex-A72 processor (1.5GHz) and has built-in "noise reduction point optimization algorithm" and "motion path planning algorithm". It receives user / environment data from the perception and positioning system and sound field data from the noise acquisition unit, and outputs control commands to the drive control system (movement path) and noise reduction execution system (reverse sound wave parameters) to realize multi-module collaborative work.

[0091] In this embodiment, the hardware selection and assembly details are as follows:

[0092] (1) Airbag Suspension System 1:

[0093] Outer main airbag: Polyimide-rubber composite membrane (thickness 0.1mm, tensile strength ≥50MPa), heat-sealed, volume 0.4m³. 3 After being filled with helium, the buoyancy is ≥5.5N (which offsets the weight of the robot's total mass of 500g);

[0094] Inner regulating airbag: PVC membrane (0.08mm thick), volume 0.05m³ 3 Connect the DP-100 miniature air pump to the MS5803 pressure sensor. The air pump is powered by 5V and has an operating current of ≤200mA.

[0095] (2) Drive control system 3:

[0096] Miniature vector thruster: Brushless Motor BL1020 (KV value 3000) is selected, equipped with a 50mm propeller, rated voltage 11.1V, maximum thrust 5N; 6 groups of thrusters are fixed by aluminum alloy brackets, spaced 50mm apart, and distributed in a regular hexagonal pattern;

[0097] Attitude sensor: MPU9250 module, I2C communication interface, 100Hz sampling rate, mounted in the center of the carrier chamber via a shock-absorbing pad to reduce thruster vibration interference.

[0098] (3) Noise Reduction Execution System 4:

[0099] Microphone array: 8 SGM3770 MEMS microphones, fixed by 3D printed spherical brackets (50mm in diameter), with uniform spacing. The output signal is converted by TIADS1263 ADC (24-bit precision) and then transmitted to the processor.

[0100] Directional loudspeakers: 4 PS100 piezoelectric loudspeakers with a frequency response of 50-1500Hz and a sensitivity of 92dB. They are integrated via PCB board, spaced 30mm apart, and arranged in a square pattern.

[0101] (4) Sensing and Positioning System 5:

[0102] RGB-D webcam: Intel RealSense D435i, USB 3.0 interface, 87°×58° field of view, depth accuracy ±0.05m;

[0103] Ultrasonic sensor: HC-SR04, 8 sensors are arranged in a ring on the side wall of the carrier chamber, with a spacing of 45°, and a detection frequency of 40kHz.

[0104] (5) Central Processing System 6:

[0105] Processor: Raspberry Pi 4B (ARM Cortex-A72, 1.5GHz), with a 16GB microSD card, running Linux;

[0106] Power supply: 11.1V 5000mAh lithium battery pack (discharge rate 10C), converted to 5V / 12V via DC-DC module to power each module, with a battery life of 3.5 hours (suspended + moving + noise reduction mode).

[0107] Test environment: Office setting

[0108] Environmental noise: air conditioner operating noise (main frequency 250Hz, sound pressure level 72dB) + noise from people talking (main frequency 500Hz, sound pressure level 68dB);

[0109] User status: Initially sitting (coordinates (0,0,1.3m)), later got up and moved to (1.5m,0,1.6m).

[0110] Test results:

[0111] Initial noise reduction point optimization: The optimal point with a potential value of 92 was selected (coordinates (0.6m, 0, 1.3m)), and the movement time was 0.8s;

[0112] Noise reduction effect: The sound pressure level in the user's head area decreased from 72dB to 45dB, a reduction of 27dB, achieving the expected goal;

[0113] Dynamic adjustment: After the user moves 1.5m, the system triggers a re-optimization and finds a new optimal point (coordinates (2.1m, 0, 1.6m)) within 500ms, maintaining the sound pressure level at 48dB (reduced by 20dB);

[0114] Obstacle avoidance test: The desk was detected (0.2m away), the path was automatically adjusted, and the path was detoured 0.5m to avoid the obstacle without any collision.

[0115] This suspended, dynamic optimization, and active noise reduction robot has the following advantages:

[0116] (1) Breakthrough in spatial adaptability: Through the combination of airbag suspension and vector thruster, three-dimensional spatial movement (lifting + horizontal + turning) is achieved, breaking free from the limitations of fixed installation, adapting to various scenarios such as home, office, and coffee shop, and improving the applicable range by more than 80% compared with traditional fixed noise reduction devices;

[0117] (2) Significantly improved noise reduction accuracy: The spherical microphone array realizes three-dimensional sound field reconstruction, the directional speaker array focuses the reverse sound wave, and with the dynamic optimization algorithm, the noise in the target area around the user is reduced by 20-30dB, which is 30% better than the traditional portable noise-canceling headphones (noise reduction of 15-20dB);

[0118] (3) High intelligence and autonomy: It integrates multimodal user tracking and environmental obstacle avoidance, and can automatically respond to user movement and noise changes without manual intervention, dynamically adjust the noise reduction point, and improve user experience satisfaction by 90% compared with manually adjusted noise reduction devices.

[0119] (4) Energy consumption and stability optimization: The double-layer airbag design reduces the energy consumption of suspension (compared to pure propeller suspension robots, the battery life is increased by 50%, up to 3-4 hours), and the attitude sensor and PID control ensure suspension stability (tilt angle error ≤1°), avoiding the noise reduction effect caused by shaking.

[0120] The above description is only a preferred embodiment of the present invention, but the present invention should not be limited to the content disclosed in the embodiments and drawings. Therefore, any equivalent or modified embodiments made without departing from the spirit of the present invention shall fall within the protection scope of the present invention.

Claims

1. A suspended dynamic optimization active noise reduction method, characterized in that... Includes the following steps: S1. Initialization: The airbag suspension system drives the robot to levitate to a height of 1.2-1.6m, the perception and positioning system establishes the user coordinate system, and the noise reduction execution system performs a self-test. S2, Data Acquisition: The noise acquisition unit obtains noise source parameters, and the user tracking module updates the user's head coordinates; S3, Noise Reduction Point Planning: Generate a set of candidate noise reduction points with a radius of 0.4-0.8m centered on the user, and calculate the potential value of each point; S4. Initial noise reduction: Drive the robot to move to the noise reduction point with the highest potential value and emit a reverse sound wave; S5. Effect Optimization: The error microphone detects the noise reduction effect, and if it does not meet the standard, it will perform a second optimization. S6. Dynamic Adjustment: If a user displacement ≥ 0.3m or a change in noise is detected, return to step S2 to replan.

2. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, The specific process of step S1 includes: S11. Inflate the outer main airbag to the preset pressure, and the robot suspends at an initial height of 1.2-1.6m to adapt to the user's sitting / standing posture; S12. The perception and positioning system is activated, and Bluetooth positioning and visual recognition are jointly calibrated to establish a spatial coordinate system with the user's initial position as the origin. S13. The noise reduction system performs a self-test, and the microphone array and speaker array are initialized to ensure that the signal path is normal.

3. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, In step S2, after the noise acquisition unit obtains the noise source parameters, it calculates the noise source parameters, namely azimuth, elevation, main frequency and sound pressure level, using spherical near-field acoustic holography. After the user tracking module updates the user's head coordinates, it records the user's movement trend.

4. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, The specific process of step S3 includes: S31. The central processing system generates a set of candidate noise reduction points based on the "noise source-user relative position": A spherical region with a radius of 0.4-0.8m is formed, centered on the user's head, and divided into a grid with a step size of 0.1m, resulting in ≤500 candidate points (P1, P2, ..., P...). n ); S32. Calculate the "noise reduction potential value" (value 0-100) for each candidate point. The calculation model is: potential value = 60% × (1 - sound wave attenuation coefficient) + 30% × (1 - environmental interference coefficient) + 10% × (1 - moving path length / maximum path length). Among them, the sound wave attenuation coefficient is calculated based on the distance from the noise source / user. For every 0.1m increase in distance, the attenuation coefficient increases by 0.

05. The environmental interference coefficient is calculated based on the degree of obstruction by surrounding obstacles. For every 10% increase in the proportion of obstructed area, the interference coefficient increases by 0.

1.

5. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, Step S4 specifically includes: S41. The central processing system selects the top 3 candidate points with the highest potential value and generates the optimal movement path through the motion path planning algorithm. S42. The drive control system drives the robot to move to the initial noise reduction point (P_max) with the highest potential value, with a movement error ≤0.05m; S43. The reverse acoustic wave generation unit generates a reverse acoustic wave signal based on real-time noise data using the FxLMS algorithm. The directional emission unit focuses the emission and targets the user's head area.

6. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, The specific process of step S5 is as follows: S51. The "error microphone" in the noise acquisition unit detects the sound pressure level after noise reduction. If the sound pressure level reduction is ≥20dB, it meets the standard and the current noise reduction point is maintained; if the reduction is <20dB, it does not meet the standard and triggers secondary optimization. S52. Move the remaining candidate points sequentially from high to low according to the potential value, with a step size of 0.1m and a moving time of ≤0.5s / point. Stay at each candidate point for 0.3s, repeat the reverse sound wave emission and effect detection until a qualified noise reduction point is found.

7. The suspended dynamic optimization active noise reduction method according to claim 1, characterized in that, The specific process in step S6 is as follows: S61. Real-time monitoring trigger conditions: If the user displacement is ≥0.3m, or the main frequency of the noise source fluctuates by ±30Hz or the sound pressure level changes by ±5dB, return to step S2, re-collect data and plan noise reduction points. S62. When there is no triggering condition, the noise reduction potential value is updated every 5 seconds, and the robot position is fine-tuned by ≤0.05m to maintain the best noise reduction effect.

8. A suspended dynamic optimization active noise reduction robot, characterized in that: To implement any one of the noise reduction methods as described in claims 1-7, the system specifically includes: an airbag suspension system, a drive control system, a noise reduction execution system, a sensing and positioning system, and a central processing system; a carrier chamber is fixedly connected below the airbag suspension system, and the drive control system, noise reduction execution system, sensing and positioning system, and central processing system are integrated within the carrier chamber; the airbag suspension system has a double-layer sealed structure, with the outer main airbag filled with helium to provide basic buoyancy, and the inner adjusting airbag connected to a miniature electric air pump and a pressure sensor, adjusting lifting and levitation through inflation and deflation; the drive control system includes 6 groups of regularly hexagonally distributed... The system comprises a miniature vector thruster and attitude sensor, which achieves horizontal movement and attitude correction through differential control; the noise reduction execution system includes a spherical microphone array, an improved FxLMS algorithm reverse acoustic wave generation unit, and a directional speaker array, which realizes noise acquisition, reverse acoustic wave generation, and focused emission; the perception and positioning system includes a Bluetooth-AoA+RGB-D vision-integrated user tracking module and a ring-distributed ultrasonic obstacle avoidance module; the central processing system receives data from each module, and outputs control commands to other systems through a noise reduction point optimization algorithm and a path planning algorithm to achieve dynamic noise reduction.

9. The suspended dynamic optimization active noise reduction robot according to claim 8, characterized in that, The spherical microphone array is a spherical structure with a diameter of 50mm composed of 8 MEMS microphones. It reconstructs a three-dimensional sound field based on spherical near-field acoustic holography technology, with a noise source localization error of ≤3° and a spectrum analysis range of 20-2000Hz.

10. The suspended dynamic optimization active noise reduction robot according to claim 8, characterized in that, The directional loudspeaker array is a square array composed of four piezoelectric directional loudspeakers. The direction of the sound waves is controlled by phased array technology, the diameter of the focusing area is ≤0.2m, and the frequency response range is 50-1500Hz.

Images