Intelligent guiding method and system for high-precision positioning robot based on millimeter waves

Abstract

The invention relates to the technical field of robot guidance, and discloses a high-precision positioning robot intelligent guidance method and system based on millimeter waves, and the method comprises the steps: carrying out the time domain noise suppression processing and environment static background component signal removal processing of an environment echo signal; signal enhancement is carried out by using self-adaptive time sequence weighting and an energy focusing mechanism; decomposing the enhanced dynamic signal by using a multi-scale empirical mode decomposition algorithm to obtain a low-frequency disturbance component and a high-frequency human body dynamic component, and performing disturbance suppression processing based on disturbance coherence degree; generating a human body space position coordinate as a guide target by using the human body space positioning model; and the robot executes human body intelligent following and obstacle avoidance navigation according to the guide target and the environment information. Through multi-antenna millimeter wave signal enhancement, disturbance suppression and multi-scale decomposition, stable extraction and high-precision spatial positioning of human body echoes are realized, and intelligent following and dynamic obstacle avoidance guidance of a robot to a human body are completed based on a positioning result.

Description

Technical Field

[0001] This invention relates to the fields of industrial vision and computer vision, and more particularly to the field of robot guidance, specifically a method and system for intelligent robot guidance based on millimeter-wave high-precision positioning. Background Technology

[0002] With the widespread deployment and continuous application of service robots, inspection robots, and companion robots in complex indoor and outdoor environments such as shopping malls, hospitals, parks, communities, and homes, higher demands are placed on the robots' autonomous perception capabilities, spatial positioning accuracy, and stability of the guidance process when performing autonomous service, inspection and monitoring, and human-robot interaction tasks. Robots not only need to accurately acquire the target's location, but also need to achieve stable and continuous target perception and motion guidance under conditions of continuous target movement, complex environmental structures, and diverse interference factors, thereby ensuring the reliability and safety of task execution.

[0003] Millimeter waves, due to their high frequency and strong spatial resolution, can achieve centimeter-level positioning accuracy. Therefore, millimeter-wave indoor positioning serves as a core support for technologies such as smart homes, the Internet of Things, and robot navigation, and has received widespread attention in recent years. However, various low-frequency dynamic disturbances in the home environment (such as swaying curtains, pet activity, and air conditioning airflow) generate signal reflection characteristics similar to human movement. These interference signals cannot be effectively distinguished by traditional static modeling or time filtering methods, leading to positioning errors or false trajectories. Therefore, how to decouple human activity from environmental interference in complex home environments and process and separate the dynamic components in millimeter-wave signals in real time has become a crucial problem that urgently needs to be solved to improve the robustness and accuracy of home positioning systems.

[0004] Existing robot guidance methods mostly rely on satellite positioning, inertial navigation, or visual perception for path tracking and control. For example, patent CN109900273B discloses an outdoor mobile robot guidance method and system. This method calculates the robot's current longitude, latitude, and heading angle, combines this with a pre-planned set of coordinate points for the target guidance path, constructs a guidance line function, and calculates lateral and angular deviations. These deviations are then used as feedback input to a digital PID controller to drive a servo motor, enabling the robot to move along the guidance route. This method can simultaneously correct angular and lateral errors, making it suitable for outdoor environments with significant variations in road width. While it has some practicality in path tracking control, it lacks fine-grained differentiation of environmental echo signals and is prone to positioning drift in complex electromagnetic and dynamic environments. It struggles to support high-precision, stable human guidance, and the positioning accuracy of guidance methods based on latitude, longitude, and heading angles significantly decreases indoors, in semi-enclosed spaces, or in scenarios where GNSS signals are blocked.

[0005] To address this issue, this invention proposes a method and system for intelligent guidance of robots based on millimeter-wave high-precision positioning, promoting the widespread application of millimeter-wave indoor positioning technology in smart homes and robot navigation. Summary of the Invention

[0006] This invention provides a method and system for intelligent guidance of robots based on millimeter-wave high-precision positioning. Addressing the problem that millimeter-wave radar is susceptible to multipath reflection, environmental noise, and static object interference in complex indoor environments, leading to the submersion of human echo signals and difficulty in accurately extracting dynamic reflection components, step S1 employs temporal noise suppression and static background component removal to effectively separate environmental dynamic reflection components, providing high signal-to-noise ratio environmental dynamic reflection component signals for subsequent human perception. Step S2 introduces an adaptive temporal weighting and energy focusing mechanism to focus on enhancing reflection components more likely to originate from human movement and suppressing non-human-related signals, thereby improving the separability of effective human echoes in both the temporal and energy dimensions. Step S3... The steps involve multi-scale empirical mode decomposition combined with a perturbation coherence suppression mechanism to adaptively suppress low-frequency perturbations, effectively reducing the impact of robot motion, ground reflection, and gradually changing environmental factors on the positioning results. Then, in step S4, by fusing the perturbation-suppressed low-frequency components with high-frequency human dynamic components, stable human dynamic reflection data is constructed. Combined with the human spatial positioning model, high-precision and continuous estimation of human spatial coordinates is achieved, significantly improving the accuracy and robustness of millimeter-wave positioning. By using high-precision human spatial coordinates as guidance targets and combining them with environmental perception information for intelligent path planning and dynamic obstacle avoidance, the robot can achieve smooth and reliable human following and intelligent guidance in complex environments.

[0007] To achieve the above objectives, the present invention provides an intelligent guidance method for a high-precision positioning robot based on millimeter waves, comprising the following steps: S1: The robot collects environmental echo signals through a multi-antenna millimeter-wave radar, performs time-domain noise suppression processing and removes static background component signals from the environmental echo signals to obtain dynamic reflection component signals of the environment. S2: Based on the environmental dynamic reflection component signal, the signal is enhanced by adaptive time-series weighting and energy focusing mechanism to obtain the enhanced dynamic signal; S3: The enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components. The low-frequency disturbance components are then subjected to disturbance suppression processing based on disturbance coherence to obtain the disturbance-suppressed low-frequency disturbance components. S4: Based on the low-frequency disturbance component and high-frequency human body dynamic component after disturbance suppression, construct stable dynamic reflection data of the human body, and use the human body spatial positioning model to convert the stable dynamic reflection data of the human body into human body spatial position coordinates. S5: Using the spatial coordinates of the human body as a guidance target, the robot combines the guidance target and environmental information to perform intelligent human body following and obstacle avoidance navigation.

[0008] As a further improvement of the present invention: Further, step S1, which involves performing time-domain noise suppression processing and removing static background components from the environmental echo signal, includes: S11: The robot periodically transmits millimeter-wave signals to the environment and simultaneously receives environmental echo signals. The environmental echo signals are in the form of signal sets, including echo signals received by multiple antennas. The environmental echo signals are processed by a time-domain noise suppression method that combines sliding time-domain smoothing and adaptive weighted filtering to obtain time-domain noise-suppressed environmental echo signals. S12: Based on the environmental echo signal after time-domain noise suppression, the signal mean of the environmental echo signal after time-domain noise suppression is calculated according to the preset background modeling time length to construct the environmental static background component signal. S13: Perform a differential operation between the time-domain noise-suppressed environmental echo signal and the environmental static background component signal to obtain the environmental dynamic reflection component signal. The environmental dynamic reflection component signal is in the form of a signal set, which includes the dynamic reflection component signal of the echo signals received by multiple antennas after time-domain noise suppression and background removal processing.

[0009] Furthermore, the signal enhancement in step S2 utilizes adaptive timing weighting and energy focusing mechanisms, including: S21: Sequentially extract the dynamic reflection component signals from the environmental dynamic reflection component signals; S22: Based on the dynamic reflection component signal, construct a time-domain smoothing window, calculate the signal stability of the signal value in the dynamic reflection component signal within the time-domain smoothing window, and convert the signal stability into an adaptive time-series weight of the signal value. Based on the adaptive time-series weight, perform adaptive time-series weighting on the signal value in the dynamic reflection component signal to obtain the time-series weighted dynamic reflection component signal. S23: Divide the time-weighted dynamic reflection component signal into several signal frames of equal length and non-overlapping, calculate the energy proportion of any signal value in the signal frame, and use the energy proportion as the energy focusing weight of the signal value. Based on the energy focusing weight, weight the signal value to obtain the dynamic reflection component enhancement signal corresponding to the time-weighted dynamic reflection component signal. S24: Construct all dynamic reflection component enhancement signals into a set form as the enhanced dynamic signal.

[0010] Further, in step S3, the enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components, including: S31: Construct multi-scale timescale parameters and generate local analysis windows for arbitrary timescale parameters; S32: Extract the enhanced dynamic reflection component signal from the enhanced dynamic signal, and perform empirical mode decomposition on the enhanced dynamic reflection component signal according to the local analysis window of the time scale parameter to obtain multiple sets of modal components of the enhanced dynamic reflection component signal. Specifically, the empirical mode decomposition process based on the local analysis window is as follows: the local analysis window is slid across the enhanced signal of the dynamic reflection component, local extreme points within the local analysis window are extracted, and a local upper envelope curve and a local lower envelope curve are constructed based on all local extreme points in the enhanced signal of the dynamic reflection component, and the mean curve of the two is calculated. The enhanced signal of the dynamic reflection component is subtracted from the mean curve to obtain the first group of modal components, and the obtained current modal component is taken as the component to be decomposed. The local analysis window is slid across the component to be decomposed, and the next group of modal components is extracted repeatedly until the number of extracted modal components reaches the preset number of components H, thus obtaining H groups of modal components under the time scale parameters associated with the local analysis window. S33: Calculate the zero-crossing rate and instantaneous energy change rate of each group of modal components in the dynamic reflection component enhancement signal, construct a joint frequency discrimination index for the modal components based on the zero-crossing rate and instantaneous energy change rate, and generate an adaptive frequency threshold for the dynamic reflection component enhancement signal. Modal components with a joint frequency discrimination index lower than the adaptive frequency threshold are marked as low-frequency disturbance components, and modal components with a joint frequency discrimination index not lower than the adaptive frequency threshold are marked as high-frequency human dynamic components. The dynamic reflection component enhances the signal. The h-th modal component is , H represents the preset number of modal components. The formula for generating the joint frequency discrimination index is: ; ; in, Represents modal components The joint frequency discrimination index Represents modal components Zero crossover rate, Represents modal components The instantaneous rate of change of energy, Represents modal components The nth component value in Indicates the selection of a set The maximum value in, Represents the balance coefficient, enhancing the dynamic reflection component signal. This represents the enhanced dynamic reflection component of the echo signal received by the m-th antenna. M represents the number of antennas, and N represents the length of the modal component.

[0011] Further, step S3, which involves performing perturbation suppression processing on the low-frequency perturbation component based on perturbation coherence to obtain the perturbation-suppressed low-frequency perturbation component, also includes: S34: Calculate the perturbation coherence between the low-frequency perturbation component and the preset perturbation template component; S35: Convert the disturbance coherence into a disturbance suppression coefficient for low-frequency disturbance components; Specifically, the conversion formula for converting the perturbation coherence degree into the perturbation suppression coefficient is as follows: ; in, This indicates the degree of perturbation coherence. This represents the preset perturbation coherence threshold. This represents the perturbation suppression coefficient corresponding to the perturbation coherence. S36: Based on the disturbance suppression coefficient, calculate the product of the disturbance suppression coefficient and the low-frequency disturbance component to obtain the low-frequency disturbance component after disturbance suppression.

[0012] Furthermore, the construction of stable dynamic reflex data of the human body in step S4 includes: S41: Obtain multiple sets of low-frequency disturbance components and high-frequency human body dynamic components after disturbance suppression obtained from the decomposition of the enhanced signal of dynamic reflection component. S42: Construct a low-frequency compensation coefficient as the compensation coefficient for the low-frequency disturbance component after disturbance suppression, and perform weighted summation on multiple sets of low-frequency disturbance components after disturbance suppression and high-frequency human dynamic components to obtain the human stable dynamic reflection signal corresponding to the enhanced signal of the dynamic reflection component. S43: Obtain the reflection phase of the antenna associated with the enhanced dynamic reflection component signal, and splice the reflection phase with the human body stable dynamic reflection signal corresponding to the enhanced dynamic reflection component signal to obtain the human body stable dynamic reflection signal vector. S44: Construct stable dynamic reflection data of the human body from all stable dynamic reflection signal vectors.

[0013] Furthermore, step S4, which uses a human spatial positioning model to convert the stable dynamic reflection data of the human body into human spatial position coordinates, also includes: S45: The human spatial positioning model receives the human stable dynamic reflection data and splits the human stable dynamic reflection data into multiple sets of human stable dynamic reflection signal vectors, wherein each set of human stable dynamic reflection signal vectors corresponds to the echo signal received by an antenna. S46: Perform frequency analysis on the human body stable dynamic reflection signal in the human body stable dynamic reflection signal vector to obtain the delay peak value of the human body stable dynamic reflection signal, and convert the delay peak value into the radial distance of the human body spatial position coordinates relative to the antenna. S47: Estimate the azimuth angle of the human body's spatial position coordinates relative to the antennas based on the spacing between the antennas, the wavelength of the transmitted millimeter wave signal, and the phase difference between the antennas; S48: Based on the azimuth and radial distance of the human body spatial position coordinates relative to the antenna, fuse the radial distance and azimuth of multiple antennas, and convert the fused radial distance and azimuth into human body spatial position coordinates.

[0014] Furthermore, in step S5, the robot combines the guided target and environmental information to perform intelligent human following and obstacle avoidance navigation, including: S51: The robot acquires information about the surrounding environment through visual sensors, identifies obstacles, passable areas and dynamically changing areas in the environment, and constructs a corresponding environmental grid map. Impassable areas are marked in the environmental grid map, and the robot determines the search space for feasible paths within the current planning range in combination with the guidance target. S52: Use path planning methods to determine the optimal travel path from the robot's current position to the guided target; S53: During the robot's journey along the planned path, it continuously monitors environmental information and changes in the position of the guided target. When a new obstacle is detected or the guided target deviates, it dynamically replans or locally corrects the current path until it approaches the guided target. S54: By setting a safe distance threshold between the guided target and the robot, the robot automatically reduces its speed when the distance between the robot and the guided target is lower than the safe distance threshold.

[0015] This invention also proposes an intelligent guidance system for a robot based on millimeter-wave high-precision positioning. The intelligent guidance system for a robot based on millimeter-wave high-precision positioning includes a millimeter-wave radar sensing device, a signal acquisition and synchronization device, a signal processing device, a control unit, and a movement execution and obstacle avoidance device, so as to realize the intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described above.

[0016] Compared with existing technologies, this invention proposes an intelligent guidance method and system for robots based on millimeter-wave high-precision positioning. This technology has the following beneficial effects: First, by introducing an adaptive temporal weighting mechanism based on signal stability, this invention can effectively distinguish between continuous human motion reflections and random noise or transient interference, enabling stable human dynamic reflection components to be self-enhanced in the temporal dimension. Simultaneously, this invention combines spatial energy focusing processing based on the energy proportion of signal frames, further concentrating human reflection energy in the local temporal domain, suppressing the interference of low-energy stray components on the overall signal, thereby significantly improving the prominence and consistency of human motion characteristics in the dynamic reflection components, providing higher-quality enhanced dynamic signals for subsequent dynamic signal separation and human localization.

[0017] Meanwhile, this invention constructs multi-scale time-scale parameters and employs empirical mode decomposition with local analysis window constraints to adaptively decompose the enhanced signal of dynamic reflection components at different time resolutions, effectively alleviating the problems of traditional empirical mode decomposition methods being susceptible to endpoint effects and scale aliasing. Furthermore, this invention constructs a joint frequency discrimination index based on zero crossover rate and instantaneous energy change rate, and adaptively generates frequency thresholds by combining the statistical characteristics of modal components, achieving accurate differentiation between low-frequency disturbance components caused by low-frequency environmental disturbances and high-frequency human dynamic components, avoiding the instability caused by manually set thresholds. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an intelligent guidance method for a high-precision positioning robot based on millimeter waves, provided in one embodiment of the present invention. Figure 2 A structural diagram of a robot system provided in an embodiment of the present invention; Figure 3 This is an experimental drawing provided for an embodiment of the present invention.

[0019] The meanings of the reference numerals in the figure are as follows: 101. Millimeter-wave radar sensing device; 102. Signal acquisition and synchronization device; 103. Signal processing device; 104. Control unit; 105. Motion execution and obstacle avoidance device; 106. Signal indicator light. Detailed Implementation

[0020] The realization of the objectives, functional characteristics, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0021] This invention provides an intelligent guidance method for robots based on millimeter-wave high-precision positioning. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this invention: a server, a terminal, etc. In other words, the intelligent guidance method for robots based on millimeter-wave high-precision positioning can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0022] Reference Figure 1 As shown, Embodiment 1 of the present invention is as follows: A method for intelligent guidance of robots based on millimeter-wave high-precision positioning includes the following steps: S1: The robot collects environmental echo signals through a multi-antenna millimeter-wave radar, performs time-domain noise suppression processing and removes static background components from the environmental echo signals to obtain dynamic reflection components.

[0023] Specifically, the environmental echo signal undergoes time-domain noise suppression processing and environmental static background component signal removal processing, including: S11: The robot periodically transmits millimeter-wave signals to the environment and simultaneously receives environmental echo signals. The environmental echo signals are in the form of signal sets, including echo signals received by multiple antennas. The environmental echo signals are processed by a time-domain noise suppression method that combines sliding time-domain smoothing and adaptive weighted filtering to obtain time-domain noise-suppressed environmental echo signals. Specifically, an example of an environmental echo signal is: ; in, Indicates the ambient echo signal. This represents the echo signal received by the m-th antenna, where M represents the number of antennas. Indicates echo signal N discrete signal values ​​in the data. Indicates echo signal The nth discrete signal value in the data. Indicates echo signal The first in A discrete signal value Indicates the length of the echo signal. ; By sequentially performing time-domain noise suppression processing on the echo signals in the environmental echo signals, the echo signals after time-domain noise suppression are reconstructed into a signal set form, which serves as the time-domain noise-suppressed environmental echo signal. The formula for time-domain noise suppression is: ; ; ; in, Indicates echo signal The corresponding echo signal after time-domain noise suppression, In order to be The temporal noise suppression results are shown, where L represents the half-width of the temporal smoothing window (L is set to 3). Indicates echo signal Mid-time smoothing window position Adaptive filter weights, This represents an exponential function with the natural constant as its base. This represents the time-distance attenuation factor, used to control the attenuation rate of weights that are far from the current discrete signal value. This represents a dynamic inhibition regulation factor, used to adjust the degree of suppression of abnormal fluctuations. Represents discrete signal values The fluctuations between them, set The values ​​are 0.4 and 0.2 respectively. S12: Based on the environmental echo signal after time-domain noise suppression, the signal mean of the environmental echo signal after time-domain noise suppression is calculated according to the preset background modeling time length to construct the environmental static background component signal. Optionally, the preset background modeling time length is 10; specifically, the echo signal after time-domain noise suppression The formula for constructing the static background component of the environment is: ; in, This represents the echo signal after time-domain noise suppression. The static background component of the environment, This indicates the preset background modeling time length; The environmental static background component of the echo signal after all time-domain noise suppression is taken as the environmental static background component signal; S13: Perform a differential operation between the time-domain noise-suppressed environmental echo signal and the environmental static background component signal to obtain the environmental dynamic reflection component signal. The environmental dynamic reflection component signal is in the form of a signal set, which includes the dynamic reflection component signal of the echo signals received by multiple antennas after time-domain noise suppression and background removal processing.

[0024] Specifically, the environmental dynamic reflection component signal is in the form of a signal set. By sequentially performing differential operations on the time-domain noise-suppressed echo signal and the associated environmental static background component, the dynamic reflection component signal corresponding to the time-domain noise-suppressed echo signal is obtained. All dynamic reflection component signals are then constructed into a signal set as the environmental dynamic reflection component signal, specifically the time-domain noise-suppressed echo signal. With the associated environmental static background component The formula for difference operations is: ; ; in, This represents the echo signal after time-domain noise suppression. The corresponding dynamic reflection component signal, They represent respectively With the static background component of the environment The result of the difference operation, the Corresponding dynamic reflection component signal N signal values ​​in the data.

[0025] It should be noted that this invention effectively improves the reliability of millimeter-wave radar in sensing human dynamic reflection signals in complex environments by introducing an environmental echo signal preprocessing mechanism that combines adaptive temporal noise suppression with static environmental background modeling. Specifically, the temporal noise suppression processing employs an adaptive weighted filtering method based on the joint constraints of time distance and amplitude fluctuation. This adaptively reduces the weight of echo components that are weakly correlated with the current moment or exhibit abnormal fluctuations, thereby suppressing random noise and transient interference without damaging the real reflection structure of the human body, significantly improving the temporal smoothness and stability of the echo signal. Furthermore, by performing mean statistics on the echo signal after temporal noise suppression within a preset background modeling time, a static background component is constructed, enabling adaptive modeling and elimination of static reflections from walls, ground, and fixed facilities. This avoids the problem of insufficient environmental adaptability caused by manual threshold settings. Subsequently, by performing differential operations between the noise-suppressed echo signal and the static background component, the dynamic reflection component caused by human movement can be effectively separated, significantly enhancing the distinguishability and prominence of human reflections in multipath echoes. This provides a high signal-to-noise ratio and high-stability input data foundation for subsequent human dynamic feature extraction, spatial positioning, and intelligent following navigation, thereby improving the robot's perception accuracy and guidance robustness in complex indoor environments.

[0026] S2: Based on the environmental dynamic reflection component signal, the signal is enhanced using adaptive time-series weighting and energy focusing mechanisms to obtain the enhanced dynamic signal.

[0027] Specifically, signal enhancement is achieved using adaptive timing weighting and energy focusing mechanisms, including: S21: Sequentially extract the dynamic reflection component signals from the environmental dynamic reflection component signals; S22: Based on the dynamic reflection component signal, construct a time-domain smoothing window, calculate the signal stability of the signal value in the dynamic reflection component signal within the time-domain smoothing window, and convert the signal stability into an adaptive time-series weight of the signal value. Based on the adaptive time-series weight, perform adaptive time-series weighting on the signal value in the dynamic reflection component signal to obtain the time-series weighted dynamic reflection component signal. The higher the signal stability, the more likely the signal value associated with the signal stability comes from human motion reflection; the lower the signal stability, the more likely the signal value associated with the signal stability comes from noise or transient interference. Specifically, the dynamic reflection component signal Medium signal value The formula for calculating signal stability within the time-domain smoothing window is: ; in, Indicates signal value Signal stability within the time-domain smoothing window, Represents dynamic reflection component signal The first in The signal value, where L represents the half-width of the time-domain smoothing window (set L to 3). S23: Divide the time-weighted dynamic reflection component signal into several signal frames of equal length and non-overlapping, calculate the energy proportion of any signal value in the signal frame, and use the energy proportion as the energy focusing weight of the signal value. Based on the energy focusing weight, weight the signal value to obtain the dynamic reflection component enhancement signal corresponding to the time-weighted dynamic reflection component signal. Specifically, the energy of the signal value is the square of the signal value, and the energy ratio of the signal value represents the ratio of the energy of the signal value to the total energy of the signal values ​​in the signal frame in which the signal value is located; S24: Construct all dynamic reflection component enhancement signals into a set form as the enhanced dynamic signal.

[0028] S3: The enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components. The low-frequency disturbance components are then subjected to disturbance suppression processing based on disturbance coherence to obtain the disturbance-suppressed low-frequency disturbance components.

[0029] Specifically, the enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components, including: S31: Construct multi-scale timescale parameters and generate local analysis windows for arbitrary timescale parameters; Specifically, the formula for constructing the multi-scale time scale parameters is as follows: ; in, This represents the time scale parameter at the e-th scale. Indicates the number of scales. This represents the preset minimum time scale parameter, which is set. =1, Represents the scale increment factor, set It is 1.2; The time scale parameter The corresponding local analysis window has a window length of 2. ; S32: Extract the enhanced dynamic reflection component signal from the enhanced dynamic signal, and perform empirical mode decomposition on the enhanced dynamic reflection component signal according to the local analysis window of the time scale parameter to obtain multiple sets of modal components of the enhanced dynamic reflection component signal. Specifically, the empirical mode decomposition process based on the local analysis window is as follows: the local analysis window is slid across the enhanced dynamic reflection component signal to extract local extreme points within the local analysis window. Based on all local extreme points in the enhanced dynamic reflection component signal, a local upper envelope curve and a local lower envelope curve are constructed, and their mean curves are calculated. The enhanced dynamic reflection component signal is subtracted from the mean curve to obtain the first group of modal components. The obtained current modal component is used as the component to be decomposed. The local analysis window is slid across the component to be decomposed, and the next group of modal components is extracted repeatedly until the number of extracted modal components reaches the preset number of components H (set to 10), thus obtaining H groups of modal components under the time scale parameters associated with the local analysis window. S33: Calculate the zero-crossing rate and instantaneous energy change rate of each group of modal components in the dynamic reflection component enhancement signal. Based on the zero-crossing rate and instantaneous energy change rate, construct a joint frequency discrimination index for the modal components and generate an adaptive frequency threshold for the dynamic reflection component enhancement signal. Mark modal components with a joint frequency discrimination index lower than the adaptive frequency threshold as low-frequency disturbance components and modal components with a joint frequency discrimination index not lower than the adaptive frequency threshold as high-frequency human dynamic components.

[0030] In one embodiment of the present invention, the zero-crossing rate of the modal component represents the ratio between the number of sign inversions of adjacent component values ​​in the modal component and the (length-1) of the modal component, where the length of the modal component is also N. The dynamic reflection component enhances the signal. The h-th modal component is , Modal components The formula for generating the joint frequency discrimination index is: ; ; in, Represents modal components The joint frequency discrimination index Represents modal components Zero crossover rate, Represents modal components The instantaneous rate of change of energy, Represents modal components The nth component value in Indicates the selection of a set The maximum value in, Represents the balance coefficient, set It is 0.6; Specifically, the modal components have the same length as the echo signal; In another embodiment of the present invention, the dynamic reflection component enhancement signal is obtained. Mean value of joint frequency discrimination index of H group modal components and the standard deviation of the joint frequency discrimination index The dynamic reflection component enhancement signal is constructed. Adaptive frequency thresholds for all modal components: ,in This represents the threshold adjustment factor, set. It is 0.2.

[0031] Step S3, which involves performing perturbation suppression processing on the low-frequency perturbation component based on perturbation coherence to obtain the perturbation-suppressed low-frequency perturbation component, further includes: S34: Calculate the perturbation coherence between the low-frequency perturbation component and the preset perturbation template component; As an embodiment of the present invention, multiple sets of low-frequency disturbance components are extracted during the initial stage of robot operation or the static stage of human body, and the mean of the extracted multiple sets of low-frequency disturbance components is calculated as a disturbance template component. The disturbance coherence between the low-frequency disturbance component and the preset disturbance template component is the cosine similarity between the low-frequency disturbance component and the preset disturbance template component. S35: Convert the disturbance coherence into a disturbance suppression coefficient for low-frequency disturbance components; Specifically, the conversion formula for converting the perturbation coherence degree into the perturbation suppression coefficient is as follows: ; in, This indicates the degree of perturbation coherence. This indicates the preset perturbation coherence threshold, set. It is 0.2. This represents the perturbation suppression coefficient corresponding to the perturbation coherence. S36: Based on the disturbance suppression coefficient, calculate the product of the disturbance suppression coefficient and the low-frequency disturbance component to obtain the low-frequency disturbance component after disturbance suppression.

[0032] It should be noted that this invention introduces a perturbation template component and calculates the perturbation coherence between the low-frequency perturbation component and the template component to achieve adaptive characterization and quantitative discrimination of environmental background perturbation features; furthermore, the perturbation coherence is mapped to a perturbation suppression coefficient to specifically suppress low-frequency perturbation components that are highly similar to the perturbation template component, effectively reducing background interference caused by robot motion or environmental changes, while retaining stable characteristics unrelated to actual human motion.

[0033] S4: Based on the low-frequency disturbance component and the high-frequency human body dynamic component after disturbance suppression, construct stable dynamic reflection data of the human body, and use the human body spatial positioning model to convert the stable dynamic reflection data of the human body into human body spatial position coordinates.

[0034] Specifically, constructing stable dynamic reflex data of the human body includes: S41: Obtain multiple sets of low-frequency disturbance components and high-frequency human body dynamic components after disturbance suppression obtained from the decomposition of the enhanced signal of dynamic reflection component. S42: Construct a low-frequency compensation coefficient as the compensation coefficient for the low-frequency disturbance component after disturbance suppression, and perform weighted summation on multiple sets of low-frequency disturbance components after disturbance suppression and high-frequency human dynamic components to obtain the human stable dynamic reflection signal corresponding to the enhanced signal of the dynamic reflection component. Optionally, the low-frequency compensation coefficient can be set to 0.2; It should be noted that the low-frequency disturbance components after multiple disturbance suppression correspond to the decomposition results at different time scales. The persistence of the components at each scale on the time axis is different. By introducing a low-frequency compensation coefficient, these components are weighted and accumulated, which is equivalent to focusing on retaining the high-frequency human dynamic components that maintain a consistent trend of change at multiple time scales. This further suppresses the low-frequency disturbance components after disturbance suppression that are discontinuous or abrupt in time, and thus obtains the stable dynamic reflection signal of the human body after time consistency screening.

[0035] S43: Obtain the reflection phase of the antenna associated with the enhanced dynamic reflection component signal, and splice the reflection phase with the human body stable dynamic reflection signal corresponding to the enhanced dynamic reflection component signal to obtain the human body stable dynamic reflection signal vector. As an optional embodiment of the present invention, a complex baseband signal is obtained by IQ demodulating the echo signal received by the antenna; within the effective signal range corresponding to the main reflection of the human body, the complex baseband signal is averaged in the complex domain, and the reflection phase of the antenna is determined by the arctangent function based on the amplitude of the averaged complex signal, so as to avoid the influence of phase jump and noise interference on the phase estimation result. S44: Construct stable dynamic reflection data of the human body from all stable dynamic reflection signal vectors.

[0036] Step S4, which uses a human spatial positioning model to convert the stable dynamic reflection data of the human body into human spatial position coordinates, also includes: S45: The human spatial positioning model receives the human stable dynamic reflection data and splits the human stable dynamic reflection data into multiple sets of human stable dynamic reflection signal vectors, wherein each set of human stable dynamic reflection signal vectors corresponds to the echo signal received by an antenna. S46: Perform frequency analysis on the human body stable dynamic reflection signal in the human body stable dynamic reflection signal vector to obtain the delay peak value of the human body stable dynamic reflection signal, and convert the delay peak value into the radial distance of the human body spatial position coordinates relative to the antenna. As an embodiment of the present invention, the human body stable dynamic reflection signal corresponding to the m-th antenna peak latency The calculation formula is: ; in, This indicates a time delay, the value of which ranges from 2 to 20 nanoseconds. This indicates the frequency modulation bandwidth of the millimeter-wave signal emitted by the robot. This indicates the transmission duration of the millimeter-wave signal emitted by the robot within a single cycle. Represents the imaginary unit. , Indicates stable dynamic reflection signals of the human body The nth signal value, Indicates the selection that makes To reach the maximum As output Represents an exponential function with the natural constant as its base; peak latency Converted to radial distance of human body spatial position coordinates relative to the antenna : ; in, Indicates the speed of electromagnetic wave propagation; S47: Estimate the azimuth angle of the human body's spatial position coordinates relative to the antennas based on the spacing between the antennas, the wavelength of the transmitted millimeter wave signal, and the phase difference between the antennas; As another embodiment of the present invention, the human body spatial position coordinates are relative to the m-th root and the... The formula for estimating the azimuth angle of a single antenna is: ; in, Represents the spatial position coordinates of the human body relative to the m-th root and the m-th root. The azimuth angle of the antenna. This indicates the wavelength of the emitted millimeter-wave signal. Represents the m-th root and the m-th root. The spacing between the antennas These represent the m-th antenna and the m-th antenna, respectively. The phase of the root antenna, Let represent the arcsine function, where , ; S48: Based on the azimuth and radial distance of the human body spatial position coordinates relative to the antenna, fuse the radial distance and azimuth of multiple antennas, and convert the fused radial distance and azimuth into human body spatial position coordinates.

[0037] Specifically, the conversion formula for the spatial position coordinates of the human body is as follows: ; in, Indicates the radial distance after fusion. Indicates the azimuth angle after merging. The coordinates of the human body's spatial position in the robot coordinate system are calculated by taking the average radial distance of the M antennas. And it is obtained by calculating the average azimuth angle of any two antennas. M represents the number of antennas. In the robot coordinate system, the dot is located at the robot's geometric center, the X-axis points in the robot's forward direction, and the Y-axis points to the robot's left side.

[0038] It should be noted that this invention vectorizes and decomposes stable dynamic reflection data of the human body according to antenna channels, enabling independent modeling and parallel processing of multi-antenna echo signals, effectively reducing the impact of multipath superposition on spatial positioning accuracy. Specifically, this invention extracts delay peaks based on frequency analysis and completes radial distance estimation within a limited nanosecond time delay range, establishing a strict correspondence between human body distance calculation and millimeter-wave frequency modulation parameters, improving the stability and physical consistency of distance calculation. Furthermore, this invention combines multi-antenna phase difference information with antenna array geometry for azimuth angle estimation, achieving precise perception of human body spatial orientation; by fusing the radial distance and azimuth angle of multiple antennas and mapping them uniformly to the robot coordinate system, continuous, smooth, and spatially consistent human body position coordinates are obtained, significantly improving the accuracy, robustness, and real-time performance of human body positioning in complex environments, providing reliable guidance targets for robot autonomous obstacle avoidance, following, and human-machine interaction.

[0039] S5: Using the spatial coordinates of the human body as a guidance target, the robot combines the guidance target and environmental information to perform intelligent human body following and obstacle avoidance navigation.

[0040] Specifically, the robot combines the guidance target and environmental information to perform intelligent human following and obstacle avoidance navigation, including: S51: The robot acquires information about the surrounding environment through visual sensors, identifies obstacles, passable areas and dynamically changing areas in the environment, and constructs a corresponding environmental grid map. Impassable areas are marked in the environmental grid map, and the robot determines the search space for feasible paths within the current planning range in combination with the guidance target. S52: Use path planning methods to determine the optimal travel path from the robot's current position to the guided target; Optionally, the path planning method may employ a path planning method based on the A* algorithm or the Dijkstra algorithm, and comprehensively consider path length, turning smoothness, and safe distance from obstacles during the path planning process to generate a navigation path that meets the following requirements; S53: During the robot's journey along the planned path, it continuously monitors environmental information and changes in the position of the guided target. When a new obstacle is detected or the guided target deviates, it dynamically replans or locally corrects the current path until it approaches the guided target. S54: By setting a safe distance threshold between the guided target and the robot, the robot automatically reduces its speed when the distance between the robot and the guided target is lower than the safe distance threshold.

[0041] Optionally, the safe distance threshold can be set to 5 meters.

[0042] It should be noted that this invention constructs an environmental grid map by integrating visual perception and guidance target information, achieving accurate modeling of passable areas and obstacles, providing reliable environmental constraints for path planning. Then, a path planning method is used to generate the optimal travel path within a limited search space. Furthermore, the planning process comprehensively considers path length, turning smoothness, and safety distance, effectively improving the continuity and safety of robot movement. Further, by continuously monitoring environmental changes and the position of the guidance target during travel, dynamic replanning and local correction of the path are achieved, enhancing the robot's adaptability in complex and dynamic environments. Simultaneously, this invention incorporates a safety distance threshold for adaptive speed control, enabling the robot to smoothly decelerate when approaching the guidance target, avoiding sudden braking and collision risks, thereby improving the stability and overall navigation reliability of the robot during intelligent guidance and following.

[0043] Example 2: As an embodiment of the present invention, this embodiment provides an intelligent guidance system for a robot based on millimeter-wave high-precision positioning. The robot is a human perception and intelligent guidance robot based on millimeter-wave radar, as shown below. Figure 2 The robot system structure diagram shown includes a millimeter-wave radar sensing device 101, a signal acquisition and synchronization device 102, a signal processing device 103, a control unit 104, a movement execution and obstacle avoidance device 105, and a signal indicator 106. The millimeter-wave radar sensing device 101, signal acquisition and synchronization device 102, signal processing device 103, control unit 104, and movement execution and obstacle avoidance device 105 are encapsulated into a millimeter-wave high-precision positioning robot intelligent guidance system to realize the millimeter-wave high-precision positioning robot intelligent guidance method as described in Embodiment 1.

[0044] The millimeter-wave radar sensing device 101 includes an array of multiple transmitting antennas and multiple receiving antennas, used to transmit millimeter-wave signals to the environment and receive multi-channel environmental echo data reflected from the guided target and the environment. The signal acquisition and synchronization device 102 is used to perform high-speed analog-to-digital conversion, timestamp alignment and frame synchronization processing on the multi-channel environmental echo data, and output environmental echo signals with consistent timing. The signal processing device 103 is used to perform signal processing on the environmental echo signal as in steps S1 to S3 to obtain the low-frequency disturbance component after disturbance suppression and the high-frequency human body dynamic component. The control unit 104 is used to construct stable dynamic reflection data of the human body, convert the stable dynamic reflection data of the human body into human body spatial position coordinates using a human body spatial positioning model, and use the human body spatial position coordinates as a guidance target. The mobile execution and obstacle avoidance device 105 includes a vision sensor for sensing environmental information and generating navigation commands based on the guidance target to control the robot to perform movement, following and obstacle avoidance actions. The signal indicator light 106 is used to indicate the current operating status of the robot. When the signal indicator light is red, it means that the robot is currently stationary. When the signal indicator light is green, it means that the robot is currently running. Example

[0045] For reference Figure 3 The experimental diagram shown below, sub Figure 1 The distribution of the frequency domain energy spectrum at different distances and frequencies can be observed. It can be seen that there is a clear low-frequency disturbance region below 1Hz, while the region between 1Hz and 5Hz is the concentrated region of human dynamic reflection energy. The darker the color of the heat map, the stronger the energy. son Figure 2 The results of multi-scale empirical mode decomposition of the enhanced reflection signal are shown. Mode 1 corresponds to the high-frequency component (high-frequency human dynamic component), and Mode 2 is the low-frequency disturbance component. The residual term has a slow variation trend, and the separation effect of different frequency components in time can be intuitively observed. son Figure 3 The spectrum changes before and after low-frequency disturbance suppression are shown. After suppression, the energy of the signal near 0.5Hz is significantly reduced, while the energy in the human dynamic frequency band (about 2Hz to 4Hz) remains strong, indicating that the signal enhancement algorithm and the disturbance suppression algorithm can effectively eliminate low-frequency interference in the environment. Subfigure 4 shows a comparison between the stable dynamic reflection signal of the human body after disturbance suppression and after time consistency screening. The screened stable dynamic reflection signal of the human body is smoother and more stable in time, eliminating noise fluctuations in discontinuous segments and retaining the stable dynamic reflection characteristics generated by human movement.

[0046] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in terms of the scope of the patent invention.

[0047] It should be noted that the sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0048] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0049] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for intelligent guidance of a robot based on millimeter-wave high-precision positioning, characterized in that, The method includes: S1: The robot collects environmental echo signals through a multi-antenna millimeter-wave radar, performs time-domain noise suppression processing and removes static background component signals from the environmental echo signals to obtain dynamic reflection component signals of the environment. S2: Based on the environmental dynamic reflection component signal, the signal is enhanced by adaptive time-series weighting and energy focusing mechanism to obtain the enhanced dynamic signal; S3: The enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components. The low-frequency disturbance components are then subjected to disturbance suppression processing based on disturbance coherence to obtain the disturbance-suppressed low-frequency disturbance components. S4: Based on the low-frequency disturbance component and high-frequency human body dynamic component after disturbance suppression, construct stable dynamic reflection data of the human body, and use the human body spatial positioning model to convert the stable dynamic reflection data of the human body into human body spatial position coordinates. S5: Using the spatial coordinates of the human body as a guidance target, the robot combines the guidance target and environmental information to perform intelligent human body following and obstacle avoidance navigation.

2. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 1, characterized in that, Step S1 involves performing time-domain noise suppression processing and removing static background components from the environmental echo signal, including: S11: The robot periodically transmits millimeter-wave signals to the environment and simultaneously receives environmental echo signals. The environmental echo signals are in the form of signal sets, including echo signals received by multiple antennas. The environmental echo signals are processed by a time-domain noise suppression method that combines sliding time-domain smoothing and adaptive weighted filtering to obtain time-domain noise-suppressed environmental echo signals. S12: Based on the environmental echo signal after time-domain noise suppression, the signal mean of the environmental echo signal after time-domain noise suppression is calculated according to the preset background modeling time length to construct the environmental static background component signal. S13: Perform a differential operation between the time-domain noise-suppressed environmental echo signal and the environmental static background component signal to obtain the environmental dynamic reflection component signal. The environmental dynamic reflection component signal is in the form of a signal set, which includes the dynamic reflection component signal of the echo signals received by multiple antennas after time-domain noise suppression and background removal processing.

3. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 1, characterized in that, The S2 step utilizes adaptive timing weighting and energy focusing mechanisms to enhance the signal, including: S21: Sequentially extract the dynamic reflection component signals from the environmental dynamic reflection component signals; S22: Based on the dynamic reflection component signal, construct a time-domain smoothing window, calculate the signal stability of the signal value in the dynamic reflection component signal within the time-domain smoothing window, and convert the signal stability into an adaptive time-series weight of the signal value. Based on the adaptive time-series weight, perform adaptive time-series weighting on the signal value in the dynamic reflection component signal to obtain the time-series weighted dynamic reflection component signal. S23: Divide the time-weighted dynamic reflection component signal into several signal frames of equal length and non-overlapping, calculate the energy proportion of any signal value in the signal frame, and use the energy proportion as the energy focusing weight of the signal value. Based on the energy focusing weight, weight the signal value to obtain the dynamic reflection component enhancement signal corresponding to the time-weighted dynamic reflection component signal. S24: Construct all dynamic reflection component enhancement signals into a set form as the enhanced dynamic signal.

4. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 1, characterized in that, In step S3, the enhanced dynamic signal is decomposed using a multi-scale empirical mode decomposition algorithm to obtain low-frequency disturbance components and high-frequency human body dynamic components, including: S31: Construct multi-scale timescale parameters and generate local analysis windows for arbitrary timescale parameters; S32: Extract the enhanced dynamic reflection component signal from the enhanced dynamic signal, and perform empirical mode decomposition on the enhanced dynamic reflection component signal according to the local analysis window of the time scale parameter to obtain multiple sets of modal components of the enhanced dynamic reflection component signal. S33: Calculate the zero-crossing rate and instantaneous energy change rate of each group of modal components in the dynamic reflection component enhancement signal, construct a joint frequency discrimination index for the modal components based on the zero-crossing rate and instantaneous energy change rate, and generate an adaptive frequency threshold for the dynamic reflection component enhancement signal. Modal components with a joint frequency discrimination index lower than the adaptive frequency threshold are marked as low-frequency disturbance components, and modal components with a joint frequency discrimination index not lower than the adaptive frequency threshold are marked as high-frequency human dynamic components. The dynamic reflection component enhances the signal. The h-th modal component is , H represents the preset number of modal components. The formula for generating the joint frequency discrimination index is: ； ； in, Represents modal components The joint frequency discrimination index Represents modal components zero crossover rate Represents modal components The instantaneous rate of change of energy, Represents modal components The nth component value in Indicates the selection of a set The maximum value in, Represents the balance coefficient, enhancing the dynamic reflection component signal. This represents the enhanced dynamic reflection component of the echo signal received by the m-th antenna. M represents the number of antennas, and N represents the length of the modal component.

5. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 4, characterized in that, Step S3, which involves performing perturbation suppression processing on the low-frequency perturbation component based on perturbation coherence to obtain the perturbation-suppressed low-frequency perturbation component, further includes: S34: Calculate the perturbation coherence between the low-frequency perturbation component and the preset perturbation template component; S35: Convert the disturbance coherence into a disturbance suppression coefficient for low-frequency disturbance components; S36: Based on the disturbance suppression coefficient, calculate the product of the disturbance suppression coefficient and the low-frequency disturbance component to obtain the low-frequency disturbance component after disturbance suppression.

6. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 1, characterized in that, The S4 step of constructing stable dynamic human body reflex data includes: S41: Obtain multiple sets of low-frequency disturbance components and high-frequency human body dynamic components after disturbance suppression obtained from the decomposition of the enhanced signal of dynamic reflection component. S42: Construct a low-frequency compensation coefficient as the compensation coefficient for the low-frequency disturbance component after disturbance suppression, and perform weighted summation on multiple sets of low-frequency disturbance components after disturbance suppression and high-frequency human dynamic components to obtain the human stable dynamic reflection signal corresponding to the enhanced signal of the dynamic reflection component. S43: Obtain the reflection phase of the antenna associated with the enhanced dynamic reflection component signal, and splice the reflection phase with the human body stable dynamic reflection signal corresponding to the enhanced dynamic reflection component signal to obtain the human body stable dynamic reflection signal vector. S44: Construct stable dynamic reflection data of the human body from all stable dynamic reflection signal vectors.

7. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 6, characterized in that, Step S4, which uses a human spatial positioning model to convert the stable dynamic reflection data of the human body into human spatial position coordinates, also includes: S45: The human spatial positioning model receives the human stable dynamic reflection data and splits the human stable dynamic reflection data into multiple sets of human stable dynamic reflection signal vectors, wherein each set of human stable dynamic reflection signal vectors corresponds to the echo signal received by an antenna. S46: Perform frequency analysis on the human body stable dynamic reflection signal in the human body stable dynamic reflection signal vector to obtain the delay peak value of the human body stable dynamic reflection signal, and convert the delay peak value into the radial distance of the human body spatial position coordinates relative to the antenna. S47: Estimate the azimuth angle of the human body's spatial position coordinates relative to the antennas based on the spacing between the antennas, the wavelength of the emitted millimeter wave signal, and the phase difference between the antennas; S48: Based on the azimuth and radial distance of the human body spatial position coordinates relative to the antenna, fuse the radial distance and azimuth of multiple antennas, and convert the fused radial distance and azimuth into human body spatial position coordinates.

8. The intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in claim 1, characterized in that, The S5 step involves the robot combining the guided target and environmental information to perform intelligent human following and obstacle avoidance navigation, including: S51: The robot acquires information about the surrounding environment through visual sensors, identifies obstacles, passable areas and dynamically changing areas in the environment, and constructs a corresponding environmental grid map. Impassable areas are marked in the environmental grid map, and the robot determines the search space for feasible paths within the current planning range in combination with the guidance target. S52: Use path planning methods to determine the optimal travel path from the robot's current position to the guided target; S53: During the robot's journey along the planned path, it continuously monitors environmental information and changes in the position of the guided target. When a new obstacle is detected or the guided target deviates, it dynamically replans or locally corrects the current path until it approaches the guided target. S54: By setting a safe distance threshold between the guided target and the robot, the robot automatically reduces its speed when the distance between the robot and the guided target is lower than the safe distance threshold.

9. A high-precision positioning robot intelligent guidance system based on millimeter waves, characterized in that, The intelligent guidance system for a robot based on millimeter-wave high-precision positioning includes a millimeter-wave radar sensing device, a signal acquisition and synchronization device, a signal processing device, a control unit, and a movement execution and obstacle avoidance device, to realize the intelligent guidance method for a robot based on millimeter-wave high-precision positioning as described in any one of claims 1-8.

Images