Intelligent remote control method and system for unmanned intelligent turntable

Abstract

The invention provides an intelligent remote control method and system for an unmanned intelligent turntable. According to the intelligent control method of the unmanned intelligent rotary table, joint movement characteristics of a biological target are captured through a detection device, and accurate tracking is achieved by dynamically regulating and controlling wave beams through a programmable metasurface. The system adopts a space-time diagram convolutional network to construct a motion topological graph, analyzes an abnormal behavior pattern of a target, and matches with a threat rule base to generate threat assessment. And based on the evaluation result, a control instruction is automatically generated to drive the turntable to execute locking tracking. According to the method, full-process automation from target detection, behavior recognition to intelligent tracking is realized, and the intelligent level and response efficiency of perimeter protection are remarkably improved. The method is used for intelligent remote control of the unmanned intelligent turntable. According to the invention, through intelligent beam regulation and control, dynamic topology behavior identification and threat evaluation, high-precision automatic tracking and threat response of the unmanned turntable to the biological target are realized.

Description

Technical Field

[0001] The present application relates to the field of intelligent remote control technology, and in particular to an intelligent remote control method and system for an unmanned intelligent turntable. Background Art

[0002] In the field of perimeter protection, accurate identification and tracking of concealed targets are key technical requirements. Traditional passive monitoring methods struggle to effectively detect biological targets with camouflage or low-observability characteristics (such as lurking individuals and wild animals). Existing systems often miss detections or misjudge targets, especially in complex environments (such as jungles, at night, or inclement weather). Therefore, an unmanned solution that can actively detect, intelligently identify biometrics, and enable autonomous tracking is urgently needed to improve the real-time performance of security systems and the detection rate of concealed targets.

[0003] Currently, some advanced systems utilize a fusion of millimeter-wave radar and deep learning. Millimeter-wave radar actively transmits signals to detect target micro-motion signatures, such as breathing and joint movement, and then uses convolutional neural networks to analyze reflected signals, enabling detection of biological targets in non-line-of-sight conditions. This approach addresses the limitations of optical sensors in concealed environments and can distinguish between living and non-living targets based on motion signatures.

[0004] However, this solution still has significant shortcomings: first, the fixed beam coverage of millimeter-wave radar is limited, making it difficult to dynamically optimize the detection direction, resulting in lags in tracking fast-moving or intermittent targets; second, the convolutional neural network model only extracts local spatiotemporal features and cannot model the topological associations of biological joint movements, such as limb coordination during walking, resulting in insufficient behavior recognition accuracy; third, there is a lack of closed-loop linkage with turntable control, and it relies on manual intervention to adjust the monitoring perspective, making it difficult to achieve truly autonomous response. Summary of the Invention

[0005] The present application provides an intelligent remote control method and system for an unmanned intelligent turntable, which is used to solve the problem of poor intelligent remote control effect in the prior art.

[0006] In a first aspect, the present application provides an intelligent remote control method for an unmanned intelligent turntable, comprising: When the detection device integrated in the unmanned intelligent turntable scans the perimeter area, it captures the reflected signal of the biological target and analyzes the characteristic data generated by the subtle movement of the biological joints from the reflected signal; generating a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; Responding to the beam steering instruction via a programmable metasurface, reconstructing the electromagnetic wave phase distribution of the programmable metasurface, and adjusting the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; Inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; Identifying a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, and matching the specific abnormal behavior pattern with a perimeter protection threat rule library to generate a threat level control vector; The trajectory prediction parameters of the biological target are calculated based on the threat level control vector, and a spatial angle control signal of the unmanned intelligent turntable is generated to drive the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

[0007] Optionally, the trajectory prediction parameters of the biological target are calculated based on the threat level control vector, and a spatial angle control signal of the unmanned intelligent turntable is generated to drive the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal, including: Determining the trajectory prediction time window length and the spatial trajectory expansion radius according to the threat type code and the threat intensity value in the threat level control vector; Taking the spatial coordinates of the biological target at the previous moment as the reference point, extending the trajectory prediction time window length along the movement direction to generate a main prediction path, and using the spatial trajectory extension radius as the scattering radius to generate a fan-shaped prediction area; Discretize the main predicted path into a sequence of spatial coordinate points, calculate the azimuth and pitch angles of each coordinate point relative to the unmanned intelligent turntable, and the steering angular velocity between adjacent coordinate points, and generate a spatial angle control signal including an azimuth angle sequence, a pitch angle sequence, and a steering angular velocity sequence; The spatial angle control signal is input into the servo controller of the unmanned intelligent turntable, and the azimuth and pitch angles of the turntable are continuously adjusted according to the steering angular velocity. When the deviation between the actual pointing direction of the turntable and the target coordinate point exceeds the tolerance angle, the adaptive compensation mechanism is triggered. The compensation angular acceleration is calculated with the actual deviation angle as input and the compensation angular acceleration is superimposed on the steering angular velocity to update the steering parameters.

[0008] Optionally, the spatial angle control signal is input into a servo controller of an unmanned intelligent turntable, and the azimuth and pitch angles of the turntable are continuously adjusted according to the steering angular velocity. When the deviation between the actual pointing direction of the turntable and the target coordinate point exceeds the tolerance angle, an adaptive compensation mechanism is triggered, and a compensation angular acceleration is calculated using the actual deviation angle as input, and the compensation angular acceleration is superimposed on the steering angular velocity to update the steering parameters, including: Inputting the spatial angle control signal to a servo controller, the servo controller driving a rotation mechanism of the turntable to continuously adjust the pointing azimuth angle and pointing pitch angle of the turntable according to the current value in the steering angular velocity sequence, so that the turntable faces the current target position indicated by the azimuth angle sequence and the pitch angle sequence; During the adjustment process, the actual azimuth angle and the actual pitch angle of the turntable are obtained, and a first deviation angle between the actual azimuth angle and the current value of the azimuth angle sequence and a second deviation angle between the actual pitch angle and the current value of the pitch angle sequence are calculated; comparing the first deviation angle and the second deviation angle with a preset tolerance angle, and activating an adaptive compensation mechanism when either the first deviation angle or the second deviation angle exceeds the tolerance angle; In the adaptive compensation mechanism, the first deviation angle and the second deviation angle are used as inputs to calculate a first compensation angular acceleration and a second compensation angular acceleration, respectively; Adding the first compensation angular acceleration to the azimuth component of the current value of the steering angular velocity sequence to form an updated azimuth steering angular velocity, and adding the second compensation angular acceleration to the pitch component of the current value of the steering angular velocity sequence to form an updated pitch steering angular velocity, and combining the updated azimuth steering angular velocity and pitch steering angular velocity to update the steering parameter as a whole; The overall updated steering parameters are used to replace corresponding values ​​in the original steering angular velocity sequence, and the continuous adjustment process of the turntable orientation is continued.

[0009] Optionally, the feature data acquired under the enhanced continuous tracking capability is input into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and the time series motion features in the feature data are mapped to the spatial topological nodes to form a motion topology map, including: Transmitting the feature data acquired under the enhanced continuous tracking capability to a data receiving interface of a spatiotemporal graph convolutional network, creating a spatial node array based on the natural connection structure of biological joints, wherein each spatial node corresponds to the physical location of a specific biological joint, and establishing connecting edges between anatomically adjacent nodes; Assigning the motion parameter time series of each biological joint in the feature data to the time series feature container of the corresponding spatial node; The spatial node array, the connection edges and the loaded temporal feature container are combined to generate a motion topology graph that integrates the spatial topology structure and the temporal motion features.

[0010] Optionally, identifying a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, and matching the specific abnormal behavior pattern with a perimeter protection threat rule library to generate a threat level control vector includes: Traversing the spatial nodes of the motion topology graph, extracting the fluctuation extreme value and fluctuation frequency of the azimuth change sequence for the time series feature container of each node, calculating the number of sudden accelerations and acceleration amplitude of the rate change sequence, and counting the abnormal oscillation period of the amplitude change sequence, and marking as an abnormal node any node that meets any of the conditions that the azimuth fluctuation frequency exceeds a first set threshold, the number of sudden accelerations exceeds a second set threshold per unit time, or the abnormal oscillation period is shorter than a third set threshold; An abnormal behavior pattern map is constructed based on the spatial distribution position of the abnormal nodes and their connection edge relationships, and the abnormal behavior pattern map is compared with the rule entries in the perimeter protection threat rule library one by one. When the similarity between the abnormal behavior pattern map and any rule entry exceeds the matching threshold, a threat level control vector containing the threat type code and threat intensity value is generated.

[0011] Optionally, when the detection device integrated in the unmanned intelligent turntable scans the perimeter area, it captures the reflected signal of the biological target and parses the characteristic data generated by the subtle movement of the biological joint from the reflected signal, including: Transmitting a continuous electromagnetic scanning signal to the peripheral area through the detection device, receiving a mixed signal reflected by the biological target, and separating a modulated signal component generated by periodic micro-displacement of the biological joint from the mixed signal; Performing time domain segmentation and frequency domain energy focusing on the modulated signal components, and extracting the oscillation frequency, amplitude fluctuation and phase offset of the signal waveform in each segmentation period; The oscillation frequency is mapped to the joint swing rate, the amplitude fluctuation is mapped to the joint displacement amplitude, and the phase offset is mapped to the joint movement direction angle, thereby generating feature data consisting of rate, amplitude, and direction angle.

[0012] Optionally, the programmable metasurface responds to the beam steering instruction, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target, including: Analyzing the target azimuth, target elevation, and spatial cone angle range from the beam steering instructions; Calculating, based on the target azimuth and the target elevation, a phase compensation value required to be generated by each control unit of the programmable metasurface, wherein the phase compensation value causes the electromagnetic waves emitted by each control unit to form a coherent superposition in the direction defined by the target azimuth and the target elevation, and adjusting the gradient distribution of the phase compensation value based on the spatial cone angle range so that the width of the coherent superposition beam is extended to the spatial cone angle range; The phase compensation value is written into the control unit array of the programmable metasurface, so that the transmission beam forms a dynamic tracking area centered on the target azimuth and target pitch angles and covering a spatial cone angle range. The dynamic tracking area continuously updates the phase distribution as the biological target moves.

[0013] Optionally, generating a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data includes: Calculating the movement vector of the biological target in three-dimensional space according to the directional angle changes of the continuous time stamps in the feature data, and extrapolating the movement path in the future time window based on the movement vector; Convert the position points on the motion path into spherical coordinates with the unmanned intelligent turntable as the origin, and generate a trajectory coordinate sequence including distance values, azimuth angles, and pitch angles; Determining the spatial cone angle range that the transmit beam needs to cover based on the maximum offset distance between adjacent coordinate points in the trajectory coordinate sequence; The azimuth, pitch angle and spatial cone angle range of the target tracking point are encapsulated as command parameters of the beam steering command.

[0014] In a second aspect, the present application provides an intelligent remote control system for an unmanned intelligent turntable, comprising: The first generation module captures the reflection signal of the biological target when the detection device integrated in the unmanned intelligent turntable scans the perimeter area, and parses the reflection signal to obtain characteristic data generated by the subtle movement of the biological joint; A second generating module generates a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; an enhancement module that responds to the beam steering instruction through a programmable metasurface, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; A formation module is provided for inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; a third generating module, which identifies a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, matches the specific abnormal behavior pattern with a perimeter protection threat rule library, and generates a threat level control vector; The driving module calculates the trajectory prediction parameters of the biological target based on the threat level control vector, generates a spatial angle control signal for the unmanned intelligent turntable, and drives the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

[0015] In a third aspect, an embodiment of the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an intelligent remote control method for an unmanned intelligent turntable as described in the first aspect above.

[0016] In an embodiment of the present application, when the detection device integrated with the unmanned intelligent turntable scans the peripheral area, it captures the reflection signal of the biological target and parses the characteristic data generated by the subtle movement of the biological joints from the reflection signal; based on the orientation information and motion trajectory of the biological target reflected by the characteristic data, a beam steering instruction is generated; the programmable metasurface responds to the beam steering instruction, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; the characteristic data obtained under the enhanced continuous tracking capability is input into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the connection relationship of the biological joints through the spatiotemporal graph convolutional network, and maps the time series motion features in the characteristic data to the spatial topological nodes to form a motion topology map; The specific abnormal behavior pattern of the biological target within the perimeter area is identified through the motion topology map, and the specific abnormal behavior pattern is matched with the perimeter protection threat rule library to generate a threat level control vector; based on the threat level control vector, the trajectory prediction parameters of the biological target are calculated, and the spatial angle control signal of the unmanned intelligent turntable is generated to drive the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

[0017] This application has the following beneficial effects: By analyzing biological joint micro-motion characteristics in reflected signals through a detection device, the system overcomes the limitations of traditional optical surveillance and achieves non-contact detection of concealed targets. A programmable metasurface reconstructs the beam phase distribution in real time, enabling intelligent focusing of the transmit beam and adaptive adjustment of the scanning range, improving the robustness of target tracking in complex environments. A spatiotemporal graph convolutional network maps joint motion characteristics into a dynamic topological graph, establishing a spatial-temporal correlation model of biological motion and significantly improving the biological plausibility of behavior recognition. By matching the motion topological graph with a threat rule library, the system achieves quantitative classification of abnormal behaviors and generates interpretable threat level control vectors. Dynamically predicting the target trajectory based on the threat level and generating high-precision spatial angle control signals, the system automates the entire "discovery-identification-lock" process for the turntable.

[0018] Furthermore, this application improves the trajectory prediction and tracking control mechanism. Dynamically determining the prediction time window and expansion radius based on the threat level, it generates a precise control signal containing a sequence of azimuth / pitch angles and steering angular velocity. Furthermore, it introduces a deviation-triggered adaptive compensation mechanism: when the actual pointing deviation exceeds the limit, the compensation angular acceleration is calculated in real time and the steering parameters are updated, forming a closed-loop control system of "prediction-execution-feedback-compensation." The servo controller achieves smooth tracking by continuously adjusting the steering angular velocity, and the compensation mechanism ensures tracking stability in dynamic environments.

[0019] This application solves the problems of trajectory prediction lag and mechanical response error accumulation of fast-moving targets, greatly improves tracking accuracy through adaptive compensation, and uses a fault-tolerant angle mechanism to avoid invalid jitter, achieving centimeter-level tracking stability in complex motion modes while reducing the energy consumption of the servo system.

[0020] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0022] Figure 1 A flow chart of an intelligent remote control method for an unmanned intelligent turntable provided by the present application is shown; Figure 2 A schematic diagram of the structure of an intelligent remote control system for an unmanned intelligent turntable provided by the present application is shown; Figure 3 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0024] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0025] Current perimeter protection systems, based on millimeter-wave radar and CNN-based stealth target detection, suffer from three key flaws: First, fixed beam patterns create detection blind spots, making it difficult to dynamically track fast-moving or intermittent biological targets; second, CNN models lack the ability to model the topological relationships of biological joint motion, resulting in insufficient behavioral recognition accuracy; and third, open-loop system response delays require manual intervention to adjust the monitoring perspective, making truly autonomous threat response impossible. These issues fundamentally stem from the existing technology's fragmented approach to beam steering, behavioral understanding, and executive control, as well as its lack of unified modeling of the spatiotemporal characteristics of biological motion.

[0026] To address these limitations, the present invention proposes an intelligent turntable control method based on a programmable metasurface and a spatiotemporal graph convolutional network. This method achieves a technological breakthrough by constructing a closed-loop system of "dynamic perception-intelligent analysis-autonomous response." First, the programmable metasurface is used to reconstruct the beam phase distribution in real time, enabling intelligent focusing of the transmit beam and adaptive adjustment of the scanning range, thus resolving the tracking blind spot problem of fixed beams. Second, the spatiotemporal graph convolutional network maps joint motion features into a dynamic topological graph, establishing a spatiotemporal correlation model of biological motion and improving the accuracy of behavioral recognition. Finally, a high-precision spatial angle signal is generated based on the threat level control vector, and an adaptive compensation mechanism is introduced to achieve full autonomy of the turntable process, from target detection to lock-on tracking. Through the collaborative innovation of dynamic beam optimization, motion topology modeling, and closed-loop control, this solution improves the detection rate of concealed targets and reduces response delays, thoroughly resolving the three core defects of existing technologies: discontinuous detection, inaccurate recognition, and untimely response.

[0027] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0028] Figure 1 The present invention provides a flowchart of an intelligent remote control method for an unmanned intelligent turntable, as shown in FIG. Figure 1 As shown, the method includes: 101. When the detection device integrated into the unmanned intelligent turntable scans the perimeter area, it captures the reflected signal of the biological target and analyzes the characteristic data generated by the subtle movement of the biological joints from the reflected signal; Optionally, in step 101, when the detection device integrated with the unmanned intelligent turntable scans the peripheral area, it captures the reflection signal of the biological target and parses the characteristic data generated by the subtle movement of the biological joint from the reflection signal, including: 1011. Transmitting a continuous electromagnetic scanning signal to the peripheral area through the detection device, receiving a mixed signal reflected by the biological target, and separating a modulated signal component generated by periodic micro-displacement of the biological joint from the mixed signal; 1012. Perform time domain segmentation and frequency domain energy focusing on the modulated signal components, and extract the oscillation frequency, amplitude fluctuation, and phase offset of the signal waveform within each segmentation period; 1013. Map the oscillation frequency to the joint swing rate, map the amplitude fluctuation to the joint displacement amplitude, and map the phase offset to the joint movement direction angle, to generate feature data consisting of rate, amplitude, and direction angle.

[0029] In the above scheme, the perimeter area refers to the boundary range monitored by the unmanned intelligent turntable, such as 50 meters outside the campus wall. Biological targets are moving objects with life characteristics, such as humans and animals. The reflected signal refers to the signal that bounces back after the electromagnetic wave emitted by the detection device encounters the organism. Feature data refers to the joint motion parameters extracted from the reflected signal, such as rate, amplitude, and direction angle, which are used to identify biological movement patterns. The frequency range of the scanning signal covers the frequency band sensitive to biological reflection.

[0030] In the embodiment of the present application, first, in step 1011, the detection device transmits a continuous electromagnetic scanning signal to the peripheral area, for example, the frequency is When the millimeter wave signal encounters a biological target, the periodic micro-displacement of the joint will cause the reflected signal to produce a Doppler frequency shift, forming a modulated signal component. The mixed signal s(t) obtained at the receiving end can be expressed as: in, is the static background reflection amplitude, is the amplitude of the joint reflection signal, is the Doppler shift frequency, which is determined by the joint movement speed Decide , is the speed of light; is the phase shift, is the environmental noise. Based on the mixed signal, the static background reflection signal is used as the reference noise. The filter coefficients are dynamically trained using the LMS algorithm. The environmental static interference is iteratively subtracted from the mixed signal. The calculation formula is: ,in is the filter coefficient vector, is the step size factor, is the reference noise vector. The mixed signal minus the environmental interference is then passed through a 1-5 Hz bandpass filter, such as the characteristic frequency band of human joint motion, to filter out residual high-frequency noise, such as mechanical vibration, and low-frequency interference, such as slow-moving vegetation, and finally output a pure biological joint modulation signal. The calculation formula is: in, , is the filter coefficient, , is the filter order.

[0031] Next, the separated modulated signal is segmented in the time domain by step 1012, for example, every 0.5 seconds is an analysis window, and the Hanning window function is applied in each window. Suppress spectrum leakage, The current sampling point number, the value range is , is the total number of sampling points in a single period, which is determined by the sampling rate and window length Determine, the calculation formula is For example, for a 50-sample signal, the sampling rate is , suppressing the truncation effect and generating a windowed signal , The original signal Amplitude of each sampling point. Point FFT to get the complex spectrum ,in is the frequency index, , is the Fourier basis function, corresponding to the physical frequency , is the sampling rate, which determines the maximum analysis frequency . Then calculate the amplitude spectrum , locate the maximum value index , then the extracted oscillation frequency is ,in is the cosine term energy, is the sinusoidal energy. Within the half-power bandwidth, that is Down to peak Frequency range, calculate the amplitude fluctuation ,in is the -3dB bandwidth boundary index, determined by interpolation; finally, the phase angle at the main frequency is calculated , calculate the phase difference between adjacent time periods based on the phase angle , representing the change in motion direction.

[0032] Finally, through step 1013, the oscillation frequency of the main frequency of the signal is directly corresponded to the number of flexion and extension of the joint per unit time. For example, a 2Hz oscillation frequency is detected, which means that the joint completes 2 complete swings per second, and is mapped to the swing rate. Similarly, the amplitude fluctuation is converted into physical displacement through a preset calibration curve. For example, a 0.3V voltage difference is calibrated in the laboratory to correspond to a 5cm flexion and extension amplitude of an adult knee joint. The calibration is based on the inverse square relationship between the electromagnetic wave reflection intensity and the target displacement distance. Similarly, the phase offset is converted into a spatial direction angle according to the kinematic geometry model. For example, a 0.6rad phase difference is mapped to a 60° right turn. The conversion principle is that the phase difference reflects the tangential velocity component of the joint relative to the radar line of sight, and the direction solution model is established in combination with the Doppler effect. Finally, a standardized feature data triplet is generated. .

[0033] For example, in the perimeter protection scenario of area A: Specifically, the unmanned turntable scans at 10 rpm. When the detection device detects biological target B approaching the fence, it captures the millimeter-wave signal reflected from its leg joints. Filtering isolates the 1.5Hz modulation component, which corresponds to the walking frequency. This signal is then divided into 0.5-second segments and analyzed to reveal the following: The amplitude fluctuates by 0.4V in the first segment (reflecting a 6cm displacement), while the phase shift in the second segment is 40° (reflecting a 60° right turn). This generates the characteristic data [1.5, 6, 60] and labels it as "walking right turn."

[0034] This step achieves a refined analysis of the joint movements of biological targets, converts the original electromagnetic signals into quantifiable behavioral feature data, provides high-precision input for subsequent target behavior identification, and significantly improves the ability to capture potential threatening actions in perimeter monitoring.

[0035] 102. Generate a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; Optionally, in step 102, generating a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data may specifically include: 1021. Calculate the movement vector of the biological target in three-dimensional space based on the directional angle changes of the consecutive time stamps in the feature data, and extrapolate the movement path in the future time window based on the movement vector; 1022. Convert the position points on the motion path into spherical coordinates with the unmanned intelligent turntable as the origin, and generate a trajectory coordinate sequence including distance values, azimuth angles, and pitch angles; 1023. Determine the spatial cone angle range that the transmit beam needs to cover based on the maximum offset distance between adjacent coordinate points in the trajectory coordinate sequence; 1024. Encapsulate the azimuth angle, pitch angle, and spatial cone angle range of the target tracking point as instruction parameters of the beam steering instruction.

[0036] In the above scheme, the orientation information refers to the direction angle (0°-360°) of the biological target relative to the unmanned intelligent turntable.

[0037] The motion trajectory refers to the spatial path of the target as it moves over time (such as turning from 30° north to 60° northeast).

[0038] Beam steering commands are the parameters that control the direction and coverage of the electromagnetic beam emitted by the turntable, including azimuth, pitch, and cone angles. Feature data refers to the real-time position and trajectory of the target.

[0039] In the embodiment of the present application, first, step 1021 is used to calculate the instantaneous speed according to the adjacent timestamps in the feature data, for example, the direction angle at t=1s , t=2s direction angle , the calculation formula is ,in, is the change in direction angle at continuous time stamps, for example , is the time interval; a three-dimensional motion model is established by combining historical trajectory points, where the horizontal displacement is , the vertical displacement is ,in is the target distance measured by the radar, is the current azimuth, calculated based on horizontal and vertical displacements The moving vector of , that is, the difference between the horizontal displacement and vertical displacement of the next second and the previous second is used as the moving vector. The future path is predicted based on the moving vector, and the formula is ,in is the target's current position coordinate, Forecast duration.

[0040] Secondly, at step 1022, the position points on the predicted future motion path, e.g. Convert to spherical coordinates: ,in is the pitch angle, is a rectangular coordinate system, and the calculation results are used to generate trajectory coordinates according to the time series: .

[0041] Next, in step 1023, the straight-line distance between adjacent spherical coordinate points is calculated according to the formula , is the distance between two adjacent points, is the azimuth difference, take all The maximum value , for example 2.5m, combined with the target speed Calculate beam coverage radius ,in The safety factor is 1.2, and the final calculation is the angle of the space cone ,in is the average distance of trajectory sequence.

[0042] Finally, step 1024 extracts the latest position point from the trajectory sequence as the tracking point, for example, the coordinates at t=5s , the key parameters of the target tracking point, azimuth, pitch angle and cone angle, are structured and encapsulated as control instructions.

[0043] For example, in perimeter security scenarios, unmanned intelligent turntables detect continuous feature data of biological targets (such as people): Specifically, first, based on the continuous feature data, the moving vector of the biological target is calculated with the direction angle of 30° at t=1 second and the direction angle of 45° at t=2 seconds: Seconds and distance increment (15m→18m), derive the horizontal velocity vector [2.5m / s east, 1.8m / s north], and extrapolate the future position (t=3 seconds: 22m at 58° east by north; t=4 seconds: 25m at 62° east by north). Next, convert the predicted rectangular coordinate position to the spherical coordinate system: t=3 seconds point (18.6m, 11.2m) → calculate the distance , azimuth , pitch angle (Target height difference 1.2m), t=4 seconds (21.8m, 13.1m) → spherical coordinates , generating the trajectory sequence [(22,58,4.2), (25,62,4.0)]. Then, the beam coverage is calculated based on the trajectory sequence: Find the maximum offset distance between adjacent points: t = 3 → 4 seconds displacement , adding a safety factor of 1.2 gives a beam coverage radius of 4.32m. Calculate the cone angle based on an average distance of 23.5m: Finally, select the latest tracking point , generate structured instructions.

[0044] This step enables the unmanned turntable to have the "prediction-tracking-adaptation" capabilities, achieving continuous locking of fast-moving targets in perimeter protection, effectively solving the problem of tracking loss caused by sudden turning or acceleration of the target in traditional systems, and greatly improving the reliability of security response.

[0045] 103. Responding to the beam steering instruction via the programmable metasurface, reconstructing the electromagnetic wave phase distribution of the programmable metasurface, and adjusting the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; Optionally, in step 103, the programmable metasurface responds to the beam steering instruction, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target, including: 1031. Analyze the target azimuth angle, target pitch angle, and spatial cone angle range from the beam steering instruction; 1032. Calculate, based on the target azimuth angle and the target elevation angle, a phase compensation value required to be generated by each control unit of the programmable metasurface, wherein the phase compensation value causes the electromagnetic waves emitted by each control unit to form a coherent superposition in the direction defined by the target azimuth angle and the target elevation angle, and adjust the gradient distribution of the phase compensation value based on the spatial cone angle range so that the width of the coherent superposition beam is extended to the spatial cone angle range; 1033. Write the phase compensation value into the control unit array of the programmable metasurface so that the transmit beam forms a dynamic tracking area centered on the target azimuth and target elevation and covering a spatial cone angle range, and the dynamic tracking area continuously updates the phase distribution as the biological target moves.

[0046] In the above solution, a programmable metasurface is a smart panel composed of controllable units that changes the beam direction by adjusting the unit phase. The phase compensation value is the phase offset required to adjust each unit to focus the beam in the target direction. The dynamic tracking area is the adaptive beam coverage area centered at the target location and with a cone angle as its width.

[0047] In this embodiment, the beam steering instruction is first decomposed according to its structure in step 1031. Bytes 0-3: 32-bit floating point numbers storing the target azimuth angle; bytes 4-7: 32-bit floating point numbers storing the target elevation angle; bytes 8-11: 32-bit floating point numbers storing the spatial cone angle range; and bytes 12-15: timestamp marking the time the instruction was generated. The target azimuth angle, target elevation angle, and spatial cone angle range are extracted from the instruction by field separation.

[0048] Secondly, in step 1032, a focused beam pointing to the target direction (azimuth angle θ, pitch angle φ) is formed according to the programmable metasurface and expanded to a specified cone angle. , calculate the basic focusing phase, according to the principle of electromagnetic wave propagation, each unit In the position ( , )The phase value to be compensated is ,in The compensation makes all the unit transmission waves achieve coherent superposition in the target direction; then the beam width is expanded according to the spatial cone angle range. , superimposed quadratic phase gradient , where the curvature coefficient Determined by a pre-stored mapping table, e.g. =21° =-700 rad / m², by adjusting The value controls the degree of beam divergence; the two are then combined to form the total phase And normalize to expand the width of the coherent superposition beam to interval; finally converted into a voltage signal and written into the metasurface unit.

[0049] Finally, step 1033 is used to parse the target azimuth, pitch, and cone angle parameters in the beam steering instruction and convert them into a phase distribution matrix. The phase compensation value is then written into the physical address of the metasurface control unit array, i.e., each unit receives an independent voltage drive signal, for example, 3.7V corresponds to a 30.52 radian phase offset, and the wavefront is reconstructed by changing the electromagnetic properties of the unit substrate. Within a millisecond response cycle, the array collaboratively generates a focused beam centered on the target azimuth and pitch angles, for example, 62° in azimuth and 4.0° in pitch, while expanding the beam width to the specified cone angle range. For example, a 21° cone angle covers the target motion prediction area; the system continuously monitors the target motion trajectory. When the feature data is updated, for example, the target suddenly turns right, causing the azimuth angle to change to 65°, the new phase distribution is immediately calculated according to the process of step 1032 and the control unit is rewritten to achieve real-time migration of the beam center and the cone coverage area; in interference scenarios, such as metal reflection in the sidelobe direction, the anti-interference mode is automatically triggered: the cone angle is reduced to 18° to increase the mainlobe gain, and the sidelobe suppression capability is enhanced to ensure stable tracking with a signal-to-noise ratio of >15dB under conditions of target acceleration, turning or occlusion.

[0050] For example, in perimeter protection zone A, the unmanned turntable detects target person C (1.75m tall) moving northeast at a speed of 1.8m / s, with an initial position 25m away from the turntable (azimuth angle 62°, pitch angle 4.0°). The system has generated beam steering instructions: target azimuth angle , target pitch angle , space cone angle .

[0051] Specifically, through the formula Compensate for the path difference to ensure that the beam accurately points to the target center; superimpose the secondary phase distribution , so that the beam width is expanded from the original 5° to 21°. The final phase value The signal is converted into a voltage signal, and the converted value (30.52 radians) is written to the control unit. The reconstructed beam completes dynamic tracking within 0.6 seconds. Initial state (t = 0 seconds): beam center locked at 62° / 4.0°, -3dB power point width at 21.2°, and sidelobe suppression < -20dB. Target turn (t = 0.5 seconds): Person C suddenly turns right and accelerates to 2 meters per second. The system extrapolates their position to an azimuth of 65°, immediately updating the command parameters (center 65°, cone angle 22°), recalculating and writing the new phase distribution within 18 milliseconds. Anti-interference response (t = 1.2 seconds): When a metal fence generates strong reflection interference in the beam sidelobe direction, the system automatically increases the phase gradient curvature coefficient, narrowing the beam width to 18° and strengthening sidelobe suppression to < -25dB, maintaining a target signal-to-noise ratio > 15dB.

[0052] This step ensures that the target is continuously tracked even when it is moving rapidly or obscured by obstacles by dynamically reconstructing a high-precision directional beam, significantly improving monitoring stability in complex environments.

[0053] 104. Inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; Optionally, in step 104, the feature data obtained under the enhanced continuous tracking capability is input into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and map the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph, including: 1041. Transmitting the feature data acquired under the enhanced continuous tracking capability to the data receiving interface of the spatiotemporal graph convolutional network, creating a spatial node array based on the natural connection structure of biological joints, wherein each spatial node corresponds to the physical location of a specific biological joint, and establishing connecting edges between anatomically adjacent nodes; 1042. Assign the motion parameter time series of each biological joint in the feature data to the time series feature container of the corresponding spatial node; 1043. Combine the spatial node array, the connection edge, and the loaded temporal feature container to generate a motion topology graph that integrates the spatial topology structure and the temporal motion feature.

[0054] In the above scheme, feature data refers to joint motion parameters extracted from the reflected signals, including time series of motion parameters for multiple biological joints, such as a knee swinging 1.5 times per second, an amplitude of 6 cm, and a direction of 60°. A spatiotemporal graph convolutional network is an artificial intelligence model that simultaneously analyzes the spatial positional relationships of joints, such as the connection between the knee and hip joints, and their temporal variations, such as the knee's amplitude changing from 6 cm to 8 cm. Spatial topological nodes represent biological joints, such as the knee and elbow, arranged according to their true physiological positions, for example, the knee node is below the hip node. Edges are the lines between nodes, representing anatomical connections (e.g., thigh = knee node - hip node). A temporal feature container is a "data bucket" attached to each node, storing the time-varying motion parameters of that joint. For example, a knee node stores the amplitude of the joint swinging for 5 consecutive seconds [6 cm, 7 cm, 8 cm...]. The motion topology graph is the resulting dynamic relationship graph, which includes joint positions, connections, and historical motion data, for example, showing how the knee and hip joints interact when a person lifts their leg. Motion parameter time series include velocity change sequences, amplitude change sequences, and angular change sequences.

[0055] In the embodiment of the present application, the enhanced feature dataset obtained in the continuous tracking phase in step 1041 is first transmitted to the standardized input interface of the spatiotemporal graph convolutional network via a high-speed data bus. The dataset is represented as a four-dimensional tensor ,in is the length of the time series, is the number of biological joints detected, The characteristic dimensions of each joint, such as speed, amplitude, and direction angle, are loaded. At the same time, the preset biological anatomical structure template library is loaded. For human targets, the predefined skeleton topology mapping function is used. , this function assigns a standard anatomical position to each joint type, where the node coordinates are based on the biological coordinate system, the origin is the center of the hip joint, and the positioning rule is calculated based on the geometric scale model, such as the knee joint position: in, The knee height offset constant stored in the anatomical database. Next, the system instantiates the spatial node array and creates a node collection as , assign node attributes as physical locations and spatial indexes , then according to Establish a set of connection edges, and verify the connection using the anatomical rule base, that is, the distance constraint ,in is the anatomical length of the limb segment, angle constraint ,in The anatomical reference vector is output as a structured spatial topology map based on the distance constraint and angle constraint. ,in is a node set, is the set of connected edges, is the position matrix.

[0056] Next, in step 1042, the spatial topology node array is constructed to strictly match anatomical identifiers to joint identities. For example, the "left knee" node only receives feature data labeled "LeftKnee." Triple data integrity checks are performed: timestamp continuity is checked to ensure uninterrupted data per second; parameter integrity is verified to confirm the completeness of the three-dimensional data of rate, amplitude, and azimuth angle; and physiological plausibility is verified, flagging abnormal data. For example, an alarm is issued if the knee swing rate exceeds the range of 0.5-3.5Hz. After successful matching, the system initializes a dedicated container for each joint node—a first-in, first-out cache with a fixed storage capacity (default is 30 seconds of data). The motion parameters are stored in the container in chronological order and automatically associated with spatial coordinate labels. For example, left knee data always carries the position [-0.15, -0.45, 0]). The numerical format is standardized, converting angles to standard values ​​between 0 and 360 degrees. Dynamic anomaly protection is implemented during the loading process: the average of adjacent values ​​is automatically interpolated for brief interruptions within 0.2 seconds, temporary data is written with a delay when electromagnetic interference characteristics are detected, and the original signal is rechecked when motion conflicts are found, such as when a joint simultaneously records opposite flexion and extension states. Finally, when loading is completed, the data pointer is reset to mark the latest state point, a time-motion intensity curve index is generated, and a container storage status report is output, such as "right elbow container is fully loaded in 28 / 30 seconds." This ensures that all joint motion histories, from the walking swing of the knee joint to the climbing and grasping of the wrist joint, are archived with millisecond-level accuracy in a spatiotemporal container matching the anatomical position, forming a standardized dynamic behavior database.

[0057] Finally, step 1043 calls the physical coordinate information of the spatial node array to build a basic skeleton, such as the three-dimensional positional relationship of the 15 joints of the human body, and establishes connection edges for adjacent nodes strictly according to anatomical rules, such as the rigid connection between the knee joint node and the hip joint node, and eliminates abnormal connections through distance threshold verification, such as edges that exceed 120% of the maximum length of the thigh bone; then the loaded time series feature container is accurately mounted to the corresponding spatial node, such as the left wrist joint container is bound to the left wrist node, and the time synchronization engine is executed to align the timestamps of all containers (microsecond accuracy), so that the static coordinates and dynamic motion parameters of each node are integrated into a four-dimensional state vector, that is, the three-dimensional spatial coordinates + the time dimension motion parameters ; Then, the real-time interaction strength of the connecting edges is calculated through a dynamic weight engine, that is, based on the joint angle difference and the relative velocity change rate, when the main motion joint is detected, such as the wrist during climbing, the relevant edge weight is automatically strengthened, for example, the wrist-elbow edge weight is increased to 0.9; finally, a standardized motion topology data structure is generated: it contains a node set, that is, a vector group with coordinates and time series, an edge set, that is, a valid anatomical connection with weights, and a time stack, that is, a topological snapshot of a continuous period. Topology self-optimization is triggered every 200 milliseconds, freezing redundant nodes with a displacement change rate lower than 0.01, hiding secondary connections with weights less than a threshold, and compressing the inactive joint point hierarchy, outputting a motion topology map with both anatomical accuracy and behavioral representation.

[0058] For example, in perimeter protection zone A, the unmanned turntable monitors person E approaching the fence: Specifically, the system creates a spatial topology consisting of 15 joint nodes (head, shoulder, elbow, knee, ankle, etc.), connecting them according to the human anatomy (e.g., left knee node → left hip node, right knee node → right hip node). Time series data is then assigned to the nodes: the left knee node container contains 3 seconds of continuous amplitude data (0.3cm, 5.1cm, 9.2cm) (slow leg lift). The right elbow node container contains azimuth angles (30°, 45°, 60°). Based on this, a motion topology is generated and analyzed: the left knee amplitude reaches 9.2cm (exceeding the 7cm threshold), while the left hip node azimuth angle changes by +20° (hip twist). Finally, the spatiotemporal graph output is recognized by the convolutional network as a "climbing preparatory motion," triggering an alarm.

[0059] This step integrates discrete joint motion data into a dynamic topological graph, enabling the system to understand the overall coordination of complex movements. For example, recognizing "running" requires high-frequency leg swings and synchronized arm coordination. Compared to traditional single-point analysis, this method significantly improves the accuracy of continuous threatening movements such as climbing, falling, and sneaking, while also reducing misjudgments caused by local data fluctuations, such as joint data anomalies caused by wind blowing clothing.

[0060] 105. Identify a specific abnormal behavior pattern of the biological target within the perimeter area using the motion topology map, match the specific abnormal behavior pattern with a perimeter protection threat rule library, and generate a threat level control vector; Optionally, in step 105, identifying a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, and matching the specific abnormal behavior pattern with a perimeter protection threat rule library to generate a threat level control vector may specifically include: 1051. Traverse the spatial nodes of the motion topology graph, extract the fluctuation extreme value and fluctuation frequency of the azimuth change sequence for the time series feature container of each node, calculate the number of sudden accelerations and acceleration amplitude of the rate change sequence, count the abnormal oscillation period of the amplitude change sequence, and mark as an abnormal node any node that meets any of the following conditions: the azimuth fluctuation frequency exceeds a first set threshold, the number of sudden accelerations exceeds a second set threshold per unit time, or the abnormal oscillation period is shorter than a third set threshold; 1052. Construct an abnormal behavior pattern map based on the spatial distribution position of the abnormal nodes and their connection edge relationships, and compare the abnormal behavior pattern map with the rule entries in the perimeter protection threat rule library one by one. When the similarity between the abnormal behavior pattern map and any rule entry exceeds a matching threshold, generate a threat level control vector including a threat type code and a threat intensity value.

[0061] In the above scheme, the dynamic motion topology graph refers to a data structure that describes the spatial position, connection relationships, and motion history of biological joints. Abnormal behavior patterns refer to actions that violate normal behavioral characteristics, such as rapid climbing and lurking. The threat rule library refers to a library of predefined threat behavior characteristics. For example, the "climbing" rule includes large upper limb movements and lower limb support. The threat level control vector refers to the output result. The abnormal behavior pattern map includes the motion transmission direction between adjacent abnormal nodes and the geometric distribution pattern of abnormal node groups. Rule entries include the preset illegal climbing node distribution pattern, the preset rapid forward movement transmission direction, and the preset lurking movement oscillation period characteristics. The threat type code corresponds to the matching rule entry, and the threat intensity value is calculated based on the number of abnormal nodes and the fluctuation amplitude.

[0062] In the embodiment of the present application, firstly, all spatial nodes of the motion topology graph are traversed through step 1051, and the three-level motion feature extraction and threshold determination are performed on the time series feature container of each node: first, the direction angle change sequence is extracted, and the number of times the absolute value of the angle difference between adjacent time points exceeds 60° is calculated as the fluctuation frequency, and the maximum angle jump value is recorded as the fluctuation extreme value, such as the wrist joint sequence middle arrive of The jump is the extreme value, and the formula is used to detect whether the fluctuation frequency exceeds the standard: in Times / second is the first set threshold, that is, if the mutation rate exceeds 5 times per second, it is considered abnormal. For time point The direction angle value, is the indicator function, which counts as 1 if the condition is met. Then calculate the sudden acceleration characteristics of the rate change sequence, and detect the mutation points where the relative rate of speed change exceeds 150% point by point: in, Times / second is the second set threshold, that is, the acceleration jump is greater than 3 times per second to trigger the mark. For time point The joint rate; finally, the oscillation periodicity of the amplitude change sequence is statistically analyzed, and the main frequency period value is obtained by fast Fourier transform: in, Times / second is the third set threshold, that is, a main period shorter than 0.1 seconds is considered abnormal oscillation. is the Fourier transform operator, The system automatically marks any node as an abnormal node and records its abnormal feature type with a spatial coordinate index. For example, the left knee joint is marked as an "acceleration-type abnormal node" because the velocity change sequence [1.1, 1.2, 4.7, 4.9] rad / s detects four velocity jumps greater than 150% within 0.3 seconds, exceeding the threshold of 3 times / second.

[0063] Finally, in step 1052, a weighted topological structure is established based on the spatial distribution position of the nodes and the relationship between the connecting edges. That is, abnormal nodes are used as vertices, such as the left wrist and right knee nodes, and edges are generated according to the anatomical connection rules, such as the wrist-elbow-shoulder chain. A weight value is assigned to each edge. The weight is calculated as the product of the sum of the node abnormal values ​​and the distance attenuation coefficient, where the abnormal value of the wrist node is 7.5 times the quantized value of the azimuth fluctuation frequency, the abnormal value of the knee node is 3.3 times the number of sudden accelerations, and the distance attenuation coefficient is the inverse of the node spacing; then the map is compared with the perimeter protection threat rule library one by one: the key topological features in the rule entries are extracted, such as the "climbing behavior" entry requires that the upper limb wrist / elbow node and the lower limb knee / ankle node are abnormal at the same time and the spatial distance is less than 1 meter, and the map similarity is calculated. ,in is the node set that matches the abnormal behavior pattern graph with the rule entry. For matching nodes The weight value in the anomaly map, A set of all required nodes specified for the rule entry, Node in the rule entry The preset benchmark weights, It is the spatial position overlap. For example, if the wrist-knee distance in the actual atlas is 0.8 meters < the rule threshold of 1 meter, the overlap = 100%; when the similarity exceeds the matching threshold (85% by default), a threat level control vector is generated based on the matching rule entry. The threat type code is taken from the rule ID (such as P2 for climbing), and the threat intensity value is calculated by normalizing the abnormal node weight. ,in For nodes The quantized value of the abnormal feature, for example, wrist (7.5) + knee (3.3) = 10.8 → normalized strength 7.2, finally outputs a vector such as [P2, 7.2].

[0064] For example, a suspicious target is found in perimeter protection zone D: Specifically, the angle mutation sequence [35°→152°→38°] (two mutations exceeding 90° within 0.5 seconds) with a fluctuation frequency of 6 Hz (>5 Hz threshold) is flagged as abnormal. The rate sequence [1.1, 1.2, 4.7, 4.9] rad / s (four accelerations exceeding 1.5 times within 0.3 seconds) with a burst acceleration of 4 times / second (>3 times threshold) is flagged as abnormal. The amplitude sequence oscillation period is 0.08 seconds (<0.1 second threshold) is flagged as abnormal. The upper limb (wrist) and lower limb (knee / ankle) nodes of the abnormal behavior are spatially linked. Then, the rule library is used to compare the "fence climbing" rule, which requires both upper limb mutation and lower limb acceleration. The node matching degree is: the wrist, knee, and ankle all meet the key points of the rule, and the similarity calculation is 92% (>85% threshold). This generates a threat vector. , Encode the "climbing behavior" type in the rule base, is the threat intensity value (based on the number of abnormal nodes 3× the intensity coefficient of movement .

[0065] This step uses joint-level anomaly detection in dynamic motion topology maps. The system can accurately identify hidden threat behaviors that are difficult to capture with traditional monitoring, and quantify the risk level based on biokinematic rules: when a spatial correlation pattern of high-frequency mutations in the upper limbs and sudden accelerations in the lower limbs is detected, it automatically matches high-risk behavior items such as "climbing" and "climbing over", and generates a threat vector containing behavior types and intensity levels, realizing intelligent judgment from micro-joint anomalies to macro-behavioral threats, significantly improving the early warning capabilities for complex threats such as camouflaged sneaking and rapid raids, while avoiding misjudgments caused by animal movements or swaying vegetation.

[0066] 106. Calculate the trajectory prediction parameters of the biological target based on the threat level control vector, and generate a spatial angle control signal for the unmanned intelligent turntable to drive the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

[0067] Optionally, calculating the trajectory prediction parameters of the biological target based on the threat level control vector in step 106, generating a spatial angle control signal for the unmanned intelligent turntable, and driving the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal may specifically include: 1061. Determine the trajectory prediction time window length and the spatial trajectory expansion radius according to the threat type code and the threat intensity value in the threat level control vector; 1062. Using the spatial coordinates of the biological target at the previous moment as a reference point, extend the trajectory prediction time window length along the motion direction to generate a main prediction path, and use the spatial trajectory extension radius as a scattering radius to generate a fan-shaped prediction area; 1063. Discretize the main predicted path into a sequence of spatial coordinate points, calculate the azimuth and pitch angles of each coordinate point relative to the unmanned intelligent turntable, and the steering angular velocity between adjacent coordinate points, and generate a spatial angle control signal including an azimuth angle sequence, a pitch angle sequence, and a steering angular velocity sequence; 1064. Input the spatial angle control signal into the servo controller of the unmanned intelligent turntable, continuously adjust the azimuth and pitch angles of the turntable according to the steering angular velocity, trigger the adaptive compensation mechanism when the deviation between the actual pointing direction of the turntable and the target coordinate point exceeds the tolerance angle, calculate the compensation angular acceleration with the actual deviation angle as input, and add the compensation angular acceleration to the steering angular velocity to update the steering parameters.

[0068] Among them, step 1064 "inputting the spatial angle control signal into the servo controller of the unmanned intelligent turntable, continuously adjusting the pointing azimuth and pitch angle of the turntable according to the steering angular velocity, triggering the adaptive compensation mechanism when the deviation between the actual pointing of the turntable and the target coordinate point exceeds the tolerance angle, calculating the compensation angular acceleration with the actual deviation angle as input and superimposing the compensation angular acceleration on the steering angular velocity to update the steering parameters" includes: inputting the spatial angle control signal into the servo controller, and the servo controller drives the rotating mechanism of the turntable to continuously adjust the pointing azimuth and pointing pitch angle of the turntable according to the current value in the steering angular velocity sequence, so that the turntable faces the current target position indicated by the azimuth sequence and the pitch angle sequence; during the adjustment process, obtaining the actual azimuth and actual pitch angle of the turntable, and calculating the first deviation angle between the actual azimuth and the current value of the azimuth sequence, and the difference between the actual pitch angle and the current value of the pitch angle sequence. a second deviation angle between the first and second deviation angles; comparing the sizes of the first and second deviation angles with a preset tolerance angle, and activating an adaptive compensation mechanism when either the first or second deviation angle exceeds the tolerance angle; in the adaptive compensation mechanism, taking the first and second deviation angles as inputs, respectively calculating a first compensation angular acceleration and a second compensation angular acceleration; adding the first compensation angular acceleration to the azimuth component of the current value of the steering angular velocity sequence to form an updated azimuth steering angular velocity, and adding the second compensation angular acceleration to the pitch component of the current value of the steering angular velocity sequence to form an updated pitch steering angular velocity, and combining the updated azimuth steering angular velocity and pitch steering angular velocity as an overall updated steering parameter; using the overall updated steering parameter to replace the corresponding value in the original steering angular velocity sequence, and continuing to execute the continuous adjustment process of the turntable pointing.

[0069] In the above scheme, the threat level control vector is a binary tuple consisting of a threat type code (e.g., P2 = climbing) and a threat intensity value (e.g., 7.2 / 10). Trajectory prediction parameters are the estimated parameters of the biological target's future motion, namely, the predicted duration and path radius. The spatial angle control signal is the command that controls the turntable's rotation, namely, the azimuth angle sequence, pitch angle sequence, and steering angular velocity. The adaptive compensation mechanism is a system that automatically corrects turntable tracking deviations.

[0070] In the embodiment of the present application, first, step 1061 is performed according to the threat type code in the threat level control vector. and Accurately calculate trajectory prediction parameters, the core calculation model is: Forecast time window length: in, is the type time coefficient, is the threat intensity value, The basic time window.

[0071] Space trajectory expansion radius: in is the type radius coefficient, Expand the radius based on the foundation.

[0072] Next, step 1062 uses the most recently captured biological target spatial coordinates, such as azimuth , pitch angle ,distance Meters are used as the reference point, and the direction of movement is analyzed based on the historical trajectory, through the first 5 coordinate points, such as the direction sequence , pitch sequence Fit the direction vector and determine that the target is moving northeast by east at a horizontal turning rate of 0.7 degrees / second and a pitch change rate of 0.06 degrees / second. Then, extend the trajectory prediction time window length along this direction of movement. For example, the destruction threat intensity is determined to be 10.99 seconds. The displacement increment is calculated based on the target's current moving speed of 2.3 meters / second to form the main predicted path. Generate an axial extension line of 25.28 meters (10.99×2.3). The coordinates of the path end are the azimuth. (65.2+0.7×10.99), distance 53.88 meters (28.6+25.28), pitch angle (3.8-0.06×10.99); Finally, with the space trajectory expansion radius of 4.86 meters as the scattering radius, a fan-shaped prediction area with an angle of 60 degrees is expanded on both sides of the main prediction path. For example, the left and right boundaries of the fan-shaped area are respectively at the azimuth and The coverage area spans 8.3 meters from east to west and 5.1 meters from north to south. This area completely accommodates the possible deviation of the target's turning path, such as the target suddenly turning to the direction It is still covered at a distance of 52 meters, while preventing the expansion of invalid areas and constraining the sector radius not to exceed the terrain boundary.

[0073] Then, in step 1063, the total number of discrete points is first determined based on the prediction window length and the time step, and the position of each point is generated by linear interpolation from the starting coordinate to the end coordinate of the path, where the coordinate calculation formula of the point is: in, is the total number of discrete points, As the reference azimuth, then calculate the azimuth of each point relative to the turntable and pitch angle , is the turntable height, and then the steering angular velocity is calculated by the central difference method , is the time step, and the final generated sequence contains azimuth angles , pitch angle sequence , steering angular velocity sequence Synchronous control signal.

[0074] Finally, after the spatial angle control signal is input into the servo controller of the unmanned intelligent turntable through step 1064, the current value in the steering angular velocity sequence, i.e., the direction angle component and the pitch angle component , drive the rotating mechanism to continuously adjust the turntable direction; collect the actual azimuth angle of the turntable in real time during the movement and actual pitch angle , respectively with the target sequence value Calculate the deviation angle ; When any deviation angle exceeds the preset tolerance angle When , the adaptive compensation mechanism is triggered: first, the compensation angular acceleration is calculated by the proportional differential control formula: , The proportional gain Enhanced deviation correction, differential gain Suppressive mutation interference, The deviation change rate represents the disturbance intensity; then the compensation acceleration is added to the original steering angular velocity: The control cycle Ensure real-time updates, the physical process of converting angular acceleration into velocity increments; ultimately the updated steering parameters Instead of the original sequence value, the servo controller immediately drives the turntable to perform the correction movement, forming a "sensing-calculation-compensation-execution" closed-loop control. For example, when the crosswind causes the azimuth angle to deviate by 1.2° (the rate of change is 3° / s), the compensation acceleration is generated. The steering speed is increased from 5.0° / s to 0.16° / s, the deviation is suppressed to 0.2° within 0.3 seconds, and the locking accuracy is maintained at ±0.3° in azimuth and ±0.1° in pitch.

[0075] For example, after a biological target is discovered in perimeter protection zone F (threat type code S2-Armed Assault, threat intensity 9.3): Specifically, based on threat type S2 (surprise), the prediction window length coefficient is set to 2.0, and the intensity of 9.3 corresponds to a radius coefficient of 2.2. The prediction window length is calculated as 3 seconds (base value) × 2.0 = 6 seconds, and the spatial trajectory extension radius is calculated as 1.5 meters × 2.2 = 3.3 meters. Setting baseline parameters: Surprise threats require longer-term warnings, so the window length and radius are doubled. Then, using the previous target coordinates (azimuth 48.5°, distance 22 meters) as the reference point, the main path is extended for 6 seconds along the direction of motion (northeast 58°). At the target speed of 3 m / s, the path is extended 18 meters to the coordinate point (40.2 m, 58°). A fan-shaped prediction area is generated: with the main path as the centerline, a scattering area of ​​±3.3 meters is extended to both sides, i.e., the left boundary: azimuth , distance 40m, right boundary: direction , distance 40m, when the target turns sharply. Sudden turn , the scattering radius ensures that it is within the prediction area; then the main path is discretized into 12 points (0.5 second intervals), and the coordinates of the 7th point (32m, )→Azimuth= , target height 1.7 meters → pitch angle = , calculate the turning angular velocity of point 7→8: azimuth = , 0.5 second interval → angular velocity 1.4° / s, Δ pitch angle = →Angular velocity / s, output signal: azimuth sequence , pitch angle sequence , angular velocity series Finally, at t=3.5 seconds, the strong crosswind causes the actual azimuth (59.8°) to deviate from the target point. The deviation of 0.7° is greater than the tolerance angle of 0.5°, which triggers compensation, i.e., the proportional compensation term: Kp , differential compensation term: Kd , compensated angular acceleration = , original steering angular velocity / s →Updated to The turntable corrected 0.65° in 0.25 seconds and restored the locked target coordinate point.

[0076] This step uses a threat-driven prediction model. The system automatically extends the prediction window and expands the tracking range when the target engages in high-risk behaviors such as climbing or running. The coordinated control of the coordinate sequence and angular velocity signal enables smooth steering of the turntable. When strong winds or sudden changes in direction of the target cause tracking deviations, the compensation mechanism dynamically corrects the angle within 0.2 seconds to ensure continuous locking of high-speed moving targets, significantly reducing the risk of target loss in complex environments.

[0077] The following is a complete example of steps 101 to 106: At 2:15 a.m. in the industrial park perimeter protection system, an unmanned intelligent turntable scanned the 50-meter protection area outside the wall at a speed of 10 revolutions per minute and detected a suspicious biological target (about 1.75 meters tall) moving covertly in the shadow of vegetation in area C.

[0078] First, a turntable transmits 77GHz millimeter waves to capture the target's leg joint reflection signals. After filtering out static vegetation interference using the LMS algorithm, the 1.8Hz biomodulation signal is isolated. Time-frequency analysis is performed within a 0.5-second analysis window. During the first period, an amplitude fluctuation of 0.45V (mapped to a 6.8cm displacement using a calibration curve) and a phase difference of 0.7rad (calculated as a 50° right turn using the geometric model) are detected. During the second period, a sudden high-frequency oscillation of 2.5Hz (greater than the 1.5Hz threshold for normal walking) is captured. The resulting feature data triplet is [1.8, 6.8, 50], [2.5, 0.3, 55].

[0079] Secondly, based on the continuous frames of feature data (azimuth ) Calculate the moving vector: horizontal speed 2.8m / s (east by north), pitch change rate -0.05° / s. Predict the trajectory in the next 5 seconds: reference point (azimuth , distance 25m) extends to (direction Calculate the beam cone angle: the maximum offset distance between adjacent points is 3.2m × safety factor 1.2 = 3.84m, the average distance is 32m → cone angle Programmable metasurface beam reconstruction: Phase compensation value ,wavelength =3.9mm, the control unit writes 3.8V voltage to point the beam center to 55° azimuth / 4° pitch, and the width is expanded to 13.7° cone angle to cover the predicted path.

[0080] Next, a spatiotemporal graph convolutional network constructed a motion topology: the spatial nodes were 15 joints (focusing on the left knee node coordinates [-0.15, -0.45, 0]); the temporal container was the left knee-loaded leg-lift data [0.4cm, 5.2cm, 8.1cm]. Anomaly detection features included a sudden acceleration rate of 4 times / second (>3 times the threshold) at the left knee node and a sudden change rate of 7Hz (>5Hz threshold) at the right wrist orientation angle. The generated anomaly map showed a knee-wrist spatial distance of 1.1m and an edge weight of 0.78. The matching threat rule base showed a similarity of 91% with the "lurking climb" rule (>85% threshold). The output threat vector was [L3, 7.8] (L3 = lurking threat, intensity 7.8).

[0081] Finally, the prediction parameters are calculated based on the threat vector: the lurking window length coefficient is 1.8, which leads to T = 3s × 1.8 = 5.4 seconds; the intensity is 7.8, which leads to a radius factor of 1.52, which leads to R = 1.5m × 1.52 = 2.28m. The discretized main path is divided into 11 points (time step 0.5s). The coordinates of the sixth point (azimuth 61.2°, distance 34.5m) are: Azimuth ; Pitch angle ; Angular velocity .

[0082] In case of strong crosswind, the actual bearing deviates by 1.5° (>0.5° tolerance angle compensation): , update the steering speed , 0.28 seconds to correct the deviation, the deviation is reduced to 0.2°, and the turntable continues to lock the target direction.

[0083] Figure 2 The present invention provides a schematic diagram of the structure of an intelligent remote control system for an unmanned intelligent turntable, as shown in FIG. Figure 2 As shown, the system includes: The first generating module 21 captures the reflection signal of the biological target when the detection device integrated in the unmanned intelligent turntable scans the peripheral area, and parses the reflection signal to obtain characteristic data generated by the subtle movement of the biological joint; A second generating module 22 generates a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; an enhancement module 23, which responds to the beam steering instruction through a programmable metasurface, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; Forming module 24, inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; A third generating module 25 identifies a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, matches the specific abnormal behavior pattern with a perimeter protection threat rule library, and generates a threat level control vector; The driving module 26 calculates the trajectory prediction parameters of the biological target based on the threat level control vector, generates a spatial angle control signal for the unmanned intelligent turntable, and drives the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

[0084] Figure 2 The intelligent remote control system for unmanned intelligent turntable can execute Figure 1 The implementation principle and technical effects of the intelligent remote control method for an unmanned intelligent turntable described in the illustrated embodiment will not be elaborated on here. The specific manner in which each module and unit performs operations in the intelligent remote control system for an unmanned intelligent turntable in the above embodiment has been described in detail in the embodiments of the method and will not be elaborated on here.

[0085] In one possible design, Figure 2 The intelligent remote control system for the unmanned intelligent turntable of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0086] The processing component 32 is used for the above Figure 1 The embodiment of the invention provides an intelligent remote control method for an unmanned intelligent turntable.

[0087] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0088] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0089] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0090] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0091] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0092] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0093] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0094] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0095] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent remote control method for an unmanned intelligent turntable, characterized in that: include: When the detection device integrated in the unmanned intelligent turntable scans the perimeter area, it captures the reflected signal of the biological target and analyzes the characteristic data generated by the subtle movement of the biological joints from the reflected signal; generating a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; Responding to the beam steering instruction via a programmable metasurface, reconstructing the electromagnetic wave phase distribution of the programmable metasurface, and adjusting the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; Inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; Identifying a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, and matching the specific abnormal behavior pattern with a perimeter protection threat rule library to generate a threat level control vector; The trajectory prediction parameters of the biological target are calculated based on the threat level control vector, and a spatial angle control signal of the unmanned intelligent turntable is generated to drive the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

2. The method according to claim 1, characterized in that Calculating the trajectory prediction parameters of the biological target based on the threat level control vector, generating a spatial angle control signal for the unmanned intelligent turntable, and driving the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal, including: Determining the trajectory prediction time window length and the spatial trajectory expansion radius according to the threat type code and the threat intensity value in the threat level control vector; Taking the spatial coordinates of the biological target at the previous moment as the reference point, extending the trajectory prediction time window length along the movement direction to generate a main prediction path, and using the spatial trajectory extension radius as the scattering radius to generate a fan-shaped prediction area; Discretize the main predicted path into a sequence of spatial coordinate points, calculate the azimuth and pitch angles of each coordinate point relative to the unmanned intelligent turntable, and the steering angular velocity between adjacent coordinate points, and generate a spatial angle control signal including an azimuth angle sequence, a pitch angle sequence, and a steering angular velocity sequence; The spatial angle control signal is input into the servo controller of the unmanned intelligent turntable, and the azimuth and pitch angles of the turntable are continuously adjusted according to the steering angular velocity. When the deviation between the actual pointing direction of the turntable and the target coordinate point exceeds the tolerance angle, the adaptive compensation mechanism is triggered. The compensation angular acceleration is calculated with the actual deviation angle as input and the compensation angular acceleration is superimposed on the steering angular velocity to update the steering parameters.

3. The method according to claim 2, characterized in that The spatial angle control signal is input into the servo controller of the unmanned intelligent turntable, and the azimuth and pitch angles of the turntable are continuously adjusted according to the steering angular velocity. When the deviation between the actual direction of the turntable and the target coordinate point exceeds the tolerance angle, the adaptive compensation mechanism is triggered. The actual deviation angle is used as input to calculate the compensation angular acceleration and the compensation angular acceleration is added to the steering angular velocity to update the steering parameters, including: Inputting the spatial angle control signal to a servo controller, the servo controller driving a rotation mechanism of the turntable to continuously adjust the pointing azimuth angle and pointing pitch angle of the turntable according to the current value in the steering angular velocity sequence, so that the turntable faces the current target position indicated by the azimuth angle sequence and the pitch angle sequence; During the adjustment process, the actual azimuth angle and the actual pitch angle of the turntable are obtained, and a first deviation angle between the actual azimuth angle and the current value of the azimuth angle sequence and a second deviation angle between the actual pitch angle and the current value of the pitch angle sequence are calculated; comparing the first deviation angle and the second deviation angle with a preset tolerance angle, and activating an adaptive compensation mechanism when either the first deviation angle or the second deviation angle exceeds the tolerance angle; In the adaptive compensation mechanism, the first deviation angle and the second deviation angle are used as inputs to calculate a first compensation angular acceleration and a second compensation angular acceleration, respectively; Adding the first compensation angular acceleration to the azimuth component of the current value of the steering angular velocity sequence to form an updated azimuth steering angular velocity, and adding the second compensation angular acceleration to the pitch component of the current value of the steering angular velocity sequence to form an updated pitch steering angular velocity, and combining the updated azimuth steering angular velocity and pitch steering angular velocity to update the steering parameter as a whole; The overall updated steering parameters are used to replace corresponding values ​​in the original steering angular velocity sequence, and the continuous adjustment process of the turntable orientation is continued.

4. The method according to claim 1, wherein Inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network to construct spatial topological nodes based on the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph, including: Transmitting the feature data acquired under the enhanced continuous tracking capability to a data receiving interface of a spatiotemporal graph convolutional network, creating a spatial node array based on the natural connection structure of biological joints, wherein each spatial node corresponds to the physical location of a specific biological joint, and establishing connecting edges between anatomically adjacent nodes; Assigning the motion parameter time series of each biological joint in the feature data to the time series feature container of the corresponding spatial node; The spatial node array, the connection edges and the loaded temporal feature container are combined to generate a motion topology graph that integrates the spatial topology structure and the temporal motion features.

5. The method according to claim 1, wherein Identifying a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, matching the specific abnormal behavior pattern with a perimeter protection threat rule library, and generating a threat level control vector, including: Traversing the spatial nodes of the motion topology graph, extracting the fluctuation extreme value and fluctuation frequency of the azimuth change sequence for the time series feature container of each node, calculating the number of sudden accelerations and acceleration amplitude of the rate change sequence, and counting the abnormal oscillation period of the amplitude change sequence, and marking as an abnormal node any node that meets any of the conditions that the azimuth fluctuation frequency exceeds a first set threshold, the number of sudden accelerations exceeds a second set threshold per unit time, or the abnormal oscillation period is shorter than a third set threshold; An abnormal behavior pattern map is constructed based on the spatial distribution position of the abnormal nodes and their connection edge relationships, and the abnormal behavior pattern map is compared with the rule entries in the perimeter protection threat rule library one by one. When the similarity between the abnormal behavior pattern map and any rule entry exceeds the matching threshold, a threat level control vector containing the threat type code and threat intensity value is generated.

6. The method according to claim 1, characterized in that When the detection device integrated into the unmanned intelligent turntable scans the perimeter area, it captures the reflected signal of the biological target and analyzes the characteristic data generated by the subtle movement of the biological joints from the reflected signal, including: Transmitting a continuous electromagnetic scanning signal to the peripheral area through the detection device, receiving a mixed signal reflected by the biological target, and separating a modulated signal component generated by periodic micro-displacement of the biological joint from the mixed signal; Performing time domain segmentation and frequency domain energy focusing on the modulated signal components, and extracting the oscillation frequency, amplitude fluctuation and phase offset of the signal waveform in each segmentation period; The oscillation frequency is mapped to the joint swing rate, the amplitude fluctuation is mapped to the joint displacement amplitude, and the phase offset is mapped to the joint movement direction angle, thereby generating feature data consisting of rate, amplitude, and direction angle.

7. The method according to claim 1, characterized in that Responding to the beam steering instruction via a programmable metasurface, reconstructing the electromagnetic wave phase distribution of the programmable metasurface, and adjusting the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target, including: Analyzing the target azimuth, target elevation, and spatial cone angle range from the beam steering instructions; Calculating, based on the target azimuth and the target elevation, a phase compensation value required to be generated by each control unit of the programmable metasurface, wherein the phase compensation value causes the electromagnetic waves emitted by each control unit to form a coherent superposition in the direction defined by the target azimuth and the target elevation, and adjusting the gradient distribution of the phase compensation value based on the spatial cone angle range so that the width of the coherent superposition beam is extended to the spatial cone angle range; The phase compensation value is written into the control unit array of the programmable metasurface, so that the transmission beam forms a dynamic tracking area centered on the target azimuth and target pitch angles and covering a spatial cone angle range. The dynamic tracking area continuously updates the phase distribution as the biological target moves.

8. The method according to claim 1, characterized in that Generating a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the characteristic data includes: Calculating the movement vector of the biological target in three-dimensional space according to the directional angle changes of the continuous time stamps in the feature data, and extrapolating the movement path in the future time window based on the movement vector; Convert the position points on the motion path into spherical coordinates with the unmanned intelligent turntable as the origin, and generate a trajectory coordinate sequence including distance values, azimuth angles, and pitch angles; Determining the spatial cone angle range that the transmit beam needs to cover based on the maximum offset distance between adjacent coordinate points in the trajectory coordinate sequence; The azimuth, pitch angle and spatial cone angle range of the target tracking point are encapsulated as command parameters of the beam steering command.

9. An intelligent remote control system for an unmanned intelligent turntable, characterized in that: include: The first generation module captures the reflection signal of the biological target when the detection device integrated in the unmanned intelligent turntable scans the perimeter area, and parses the reflection signal to obtain characteristic data generated by the subtle movement of the biological joint; A second generating module generates a beam steering instruction based on the position information and motion trajectory of the biological target reflected by the feature data; an enhancement module that responds to the beam steering instruction through a programmable metasurface, reconstructs the electromagnetic wave phase distribution of the programmable metasurface, and adjusts the direction and coverage of the transmission beam of the detection device to enhance the continuous tracking capability of the biological target; A formation module is provided for inputting the feature data obtained under the enhanced continuous tracking capability into a pre-constructed spatiotemporal graph convolutional network, so as to construct spatial topological nodes according to the biological joint connection relationship through the spatiotemporal graph convolutional network, and mapping the time series motion features in the feature data to the spatial topological nodes to form a motion topology graph; a third generating module, which identifies a specific abnormal behavior pattern of the biological target within the perimeter area through the motion topology map, matches the specific abnormal behavior pattern with a perimeter protection threat rule library, and generates a threat level control vector; The driving module calculates the trajectory prediction parameters of the biological target based on the threat level control vector, generates a spatial angle control signal for the unmanned intelligent turntable, and drives the unmanned intelligent turntable to perform automatic locking and adaptive tracking actions according to the spatial angle control signal.

10. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the intelligent remote control method for an unmanned intelligent turntable as described in any one of claims 1 to 7.

Images