Residential space perception configuration method, apparatus, device, and medium

Abstract

The application discloses a kind of living space perception configuration method, device, equipment and medium, it is related to radar detection technical field, the method includes: according to living space data, movable space component data and radar perception node parameter establishes space perception model;Control movable space component switches state along opposite direction, and the displacement of component and radar echo are synchronously collected, obtain bidirectional state switching data;Determine state sensitive observation unit, non-sensitive observation unit and affected node in combination with space perception model and bidirectional state switching data;Establish parameter transfer relationship, determine anchor node and establish track constraint relationship, generate state switching configuration scheme;According to current component displacement, update affected node, associate human target echo, obtain perception configuration result.The application can distinguish component motion echo and human target echo when movable space component switches state, and keep the continuous adaptation of radar perception parameter and human target track.

Description

Technical Field

[0001] This application relates to the field of radar detection technology, and in particular to methods, devices, equipment and media for sensing and configuring living spaces. Background Technology

[0002] With the increasing application of radar sensing nodes in living spaces, the position and movement of people can be sensed through radar echoes. Moving space components such as electric doors, sliding partitions, and curtains will change the radar wave obstruction and reflection paths at different positions, requiring radar sensing configurations to adapt to these changes in component status.

[0003] Currently, background echo parameters, observation gates, and detection thresholds are typically set based on fixed spatial structures and radar sensing node parameters, and human targets are detected based on echo data collected by radar sensing nodes. When the space environment changes, background echoes can be re-acquired and corresponding sensing parameters adjusted.

[0004] However, movable spatial components generate motion echoes and continuously alter radar wave propagation conditions during state transitions. Existing methods struggle to distinguish between echo changes caused by component motion and human target echoes. Furthermore, recalibrating the entire system can lead to untimely updates of radar sensing parameters and interruptions in human target tracks. Therefore, distinguishing between component motion echoes and human target echoes during state transitions, while maintaining continuous adaptation between radar sensing parameters and human target tracks, is a pressing issue that needs to be addressed. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, equipment and medium for sensing configuration in living space, which aims to solve the technical problem of how to distinguish between the motion echo of movable space components and the echo of human targets when movable space components switch states, and to maintain the continuous adaptation of radar sensing parameters and human target tracks.

[0006] To achieve the above objectives, this application proposes a method for sensing and configuring living spaces, the method comprising: A spatial perception model is established based on living space data, movable space component data, and radar sensing node parameters. Control the movable spatial components to switch states in opposite directions, and simultaneously collect component displacement and radar echo to obtain bidirectional state switching data; Based on the spatial perception model and the bidirectional state switching data, determine the state-sensitive observation unit, the non-sensitive observation unit, and the affected nodes; Based on the state-sensitive observation unit, a parameter transfer relationship is established that changes with the displacement of the component. Based on the non-sensitive observation unit, the anchoring node is determined, a track constraint relationship is established between the anchoring node and the affected node, and a state switching configuration scheme is generated. When the movable spatial component switches states, the affected node is updated according to the current component displacement and the parameter transfer relationship, and the human target echo is associated according to the state switching configuration scheme and the trajectory constraint relationship to obtain the perception configuration result.

[0007] Furthermore, to achieve the above objectives, this application also proposes a living space sensing configuration device, the device comprising: The model building module is used to build a spatial perception model based on living space data, movable space component data and radar sensing node parameters. The switching acquisition module is used to control the movable spatial components to switch states in opposite directions, synchronously acquiring component displacement and radar echoes to obtain bidirectional state switching data; The unit determination module is used to determine the state-sensitive observation units, non-sensitive observation units, and affected nodes based on the spatial perception model and the bidirectional state switching data. The relationship generation module is used to establish parameter transfer relationships that change with the displacement of the component based on the state-sensitive observation unit, determine the anchoring node based on the non-sensitive observation unit, establish the track constraint relationship between the anchoring node and the affected node, and generate a state switching configuration scheme. The operation configuration module is used to update the affected nodes according to the current component displacement and the parameter transfer relationship when the movable space component switches states, and associate the human target echo according to the state switching configuration scheme and the track constraint relationship to obtain the perception configuration result.

[0008] In addition, to achieve the above objectives, this application also proposes a living space sensing configuration device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the living space sensing configuration method as described above.

[0009] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the living space sensing configuration method described above.

[0010] One or more technical solutions proposed in this application have at least the following technical effects: A spatial perception model is established based on residential space data, movable spatial component data, and radar sensing node parameters. This model is used to correlate spatial structure, component status, and radar sensing range, providing spatial basis for judging the impact of component movement on radar echoes. Movable spatial components are controlled to switch states in opposite directions while simultaneously collecting component displacement and radar echoes, forming bidirectional state-switching data that can be compared at the same component position. This reduces the impact of accidental echo changes during unidirectional movement on the judgment results. By combining the spatial perception model and bidirectional state-switching data, state-sensitive observation units, non-sensitive observation units, and affected nodes are identified, thereby distinguishing observations affected by changes in component status. The system identifies a relatively stable observation area and determines the nodes whose sensing parameters need updating. Based on state-sensitive observation units, it establishes parameter transfer relationships that change with component displacement. Based on non-sensitive observation units, it determines anchor nodes, establishes track constraint relationships between anchor nodes and affected nodes, and generates a state switching configuration scheme. This allows affected nodes to call corresponding parameters according to component positions and uses relatively stable node tracks to constrain the association range of human target echoes. When a movable spatial component switches states, the system updates affected nodes based on the current component displacement and associates human target echoes according to the state switching configuration scheme and track constraint relationships to obtain the sensing configuration result. This application can distinguish between component motion echoes and human target echoes when movable spatial components switch states, and maintains continuous adaptation between radar sensing parameters and human target tracks. Attached Figure Description

[0011] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart illustrating the first embodiment of the residential space sensing and configuration method of this application; Figure 2 A schematic diagram of two-way state switching and sensing scenario provided in the first embodiment of the residential space sensing configuration method of this application; Figure 3 A flowchart illustrating the second embodiment of the residential space sensing and configuration method of this application; Figure 4 This is a schematic diagram of the module structure of the living space sensing configuration device according to an embodiment of this application; Figure 5This is a schematic diagram of the device structure of the hardware operating environment involved in the residential space perception configuration method in this application embodiment.

[0014] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0015] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0016] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0017] It should be noted that the executing entity of this application embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or radar sensing configuration system capable of realizing the above functions. The radar sensing configuration system may include a processor, a memory, a radar communication interface, a component displacement data interface, and a component drive control interface. The radar communication interface is used to receive radar echoes or range-angle echo maps output by each radar sensing node; the component displacement data interface is used to receive displacement feedback from movable spatial components; and the component drive control interface is used to send state switching commands to the drive device of the movable spatial components. The following uses a radar sensing configuration system as an example to describe this embodiment and the following embodiments.

[0018] Based on this, the first embodiment of this application provides a method for sensing and configuring living space, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the residential space perception configuration method of this application.

[0019] In this embodiment, the living space sensing configuration method includes steps S10 to S50: Step S10: Establish a spatial perception model based on living space data, movable space component data, and radar sensing node parameters; It should be noted that residential space data refers to data used to characterize the spatial composition and sensing range of residential areas; movable space component data refers to data used to characterize the location, range of motion, and state changes of components; radar sensing node parameters refer to data used to characterize the detection capability and operational status of radar sensing nodes; and the spatial sensing model refers to the data structure that links residential space, movable space components, and radar sensing nodes.

[0020] Understandably, the process involves acquiring data on living space, movable spatial components, and radar sensing node parameters; unifying the spatial reference used for all types of data; and establishing a spatial perception model based on the positional relationship between movable spatial components and living space, as well as the observational relationship between radar sensing nodes and living space. The spatial perception model records the position of movable spatial components during state changes, and the range of living space observable by each radar sensing node. This step, by linking living space, movable spatial components, and radar sensing nodes, provides a data foundation for subsequently determining the impact of component state changes on radar observation results.

[0021] Step S20: Control the movable spatial components to switch states in opposite directions, and synchronously collect component displacement and radar echo to obtain bidirectional state switching data; It should be noted that component displacement refers to the amount of movement of a movable spatial component relative to its initial position during state switching; bidirectional state switching data refers to the corresponding data of component displacement and radar echo collected synchronously when a movable spatial component switches states in two opposite directions.

[0022] Understandably, the process involves controlling a movable spatial component to switch states in one direction, simultaneously acquiring component displacement and radar echoes. Then, the component is controlled to switch states in the opposite direction, and corresponding data is acquired simultaneously. The data from the two motion directions are aligned according to the component displacement, ensuring that the same component displacement corresponds to radar echoes in both directions, and this is combined to obtain bidirectional state-switching data. This step, by aligning the radar echoes from the two motion directions according to component displacement, reduces the impact of differences in component speed and acquisition time on echo comparison, providing a data foundation for identifying echo responses that repeatedly change with component position.

[0023] Step S30: Based on the spatial perception model and the bidirectional state switching data, determine the state-sensitive observation unit, the non-sensitive observation unit, and the affected nodes; It should be noted that a state-sensitive observation unit refers to an observation unit whose radar echo changes accordingly with the position of movable spatial components; a non-sensitive observation unit refers to an observation unit that has the conditions for observing human targets and is less affected by changes in the state of movable spatial components; and an affected node refers to a radar sensing node whose observation results are affected by changes in the state of movable spatial components.

[0024] Understandably, the observation range corresponding to each radar sensing node is determined based on the spatial perception model, and the radar echoes corresponding to the displacement of the same component in the two-way state switching data are compared. Observation units that change accordingly with component displacement in both motion directions are identified as state-sensitive observation units, while observation units with effective observation data and small changes are identified as non-sensitive observation units. The distribution of state-sensitive observation units in the observation range of each radar sensing node determines the affected nodes. This step distinguishes between observation areas affected by component state changes and relatively stable observation areas by comparing the repeated echo changes in the two motion directions, providing a basis for determining the parameter update targets and the reference source for human target tracks.

[0025] Step S40: Establish parameter transfer relationship that changes with component displacement based on the state-sensitive observation unit, determine anchoring node based on the non-sensitive observation unit, establish track constraint relationship between anchoring node and affected node, and generate state switching configuration scheme; It should be noted that the parameter transfer relationship refers to the correspondence between component displacement and radar sensing node configuration parameters; the anchoring node refers to the radar sensing node that can provide human target tracks during the state switching of movable spatial components; the track constraint relationship refers to the correspondence between the human target tracks of the anchoring node and the human target echo search range of the affected node; and the state switching configuration scheme refers to the node parameter call and human target echo association rules during component state switching.

[0026] Understandably, based on the echo changes of state-sensitive observation units under different component displacements, the configuration parameters of radar sensing nodes corresponding to each component displacement are determined, and parameter transfer relationships are established. Anchor nodes are determined based on the distribution of non-sensitive observation units in the sensing area, and the human target trajectory output by the anchor nodes is used to limit the human target echo search range of affected nodes, establishing trajectory constraint relationships. A state switching configuration scheme is generated based on the parameter transfer relationships, anchor nodes, affected nodes, and trajectory constraint relationships. This step, by establishing the correspondence between component displacements and node configuration parameters, enables affected nodes to invoke the corresponding configuration according to the current component position; by using the human target trajectory provided by the anchor nodes to constrain the echo search range of affected nodes, the interference of component motion echoes on human target association is reduced.

[0027] Step S50: When the movable spatial component switches states, the affected node is updated according to the current component displacement and the parameter transfer relationship, and the human target echo is associated according to the state switching configuration scheme and the trajectory constraint relationship to obtain the perception configuration result.

[0028] It should be noted that human target echo refers to radar echo corresponding to the position and motion state of the human target; perception configuration result refers to the radar perception node configuration and human target echo correlation result used during the state switching of movable spatial components.

[0029] Understandably, when a movable spatial component switches states, the current component displacement is obtained. Based on this displacement, the corresponding configuration parameters are retrieved from the parameter transfer relationship, and the affected nodes are updated. The human target trajectory output by the anchored node is obtained according to the state switching configuration scheme. The echo search range of the affected nodes is limited based on the trajectory constraint relationship. Human target echoes corresponding to the human target trajectory are then filtered within this range to obtain the perception configuration result. This step, by updating the affected nodes during component state switching and using the human target trajectory of the anchored node to correlate with the human target echo, enables the radar perception configuration to adjust with changes in component position, reducing the possibility of component motion echoes being mistaken for human target echoes.

[0030] This embodiment identifies the observation units and radar sensing nodes affected by the changes in component state by collecting component displacement and radar echoes when movable spatial components switch states in two opposite directions, and establishes parameter transfer relationships that change with component displacement. During component state switching, relatively stable radar sensing nodes are used to provide human target tracks, constrain the human target echo correlation of affected nodes, reduce interference caused by component motion echoes, reduce human target track interruptions, and improve the matching degree between the radar sensing configuration of the living space and the changes in spatial state.

[0031] As an example, the step of establishing a spatial perception model based on living space data, movable spatial component data, and radar sensing node parameters includes: extracting spatial structure data, sensing task area, and node mounting surface from the living space data; mapping the spatial structure data, the sensing task area, and the node mounting surface to a spatial coordinate system to obtain spatial mapping data; extracting a first state endpoint, a second state endpoint, motion trajectory, contour dimensions, material parameters, and component displacement from the movable spatial component data; determining the node installation position based on the node mounting surface and radar sensing node parameters, and associating the node installation position with the radar sensing node parameters to obtain a node parameter association relationship; establishing a pose correspondence relationship based on the first state endpoint, the second state endpoint, the motion trajectory, and the component displacement; and establishing a spatial perception model based on the spatial mapping data, the contour dimensions, the material parameters, the node parameter association relationship, and the pose correspondence relationship.

[0032] It should be noted that spatial structure data refers to the data used to represent the position, outline, and occupied area of ​​walls, door and window openings, and fixed furnishings; the perception task area refers to the spatial range that needs to be detected by radar perception nodes; the node mounting surface refers to the wall, ceiling, or fixed support surface that meets the installation conditions of the radar perception node; spatial mapping data refers to the coordinate representation of spatial structure data, perception task area, and node mounting surface in the same spatial coordinate system; the first state endpoint and the second state endpoint refer to the two boundary positions of the motion stroke of the movable spatial component; the node parameter correlation refers to the correlation data between the node installation position and the corresponding radar perception node parameters; and the pose correspondence refers to the correspondence between the component displacement and the position, orientation, and occupied area of ​​the movable spatial component.

[0033] In this example, the radar sensing node parameters may also include the operating frequency band, frequency modulation bandwidth, number of frequency-modulated signals per frame, number of sampling points for a single frequency-modulated signal, number of array channels, and radar frame period. The radar sensing node can use a frequency-modulated continuous wave radar in the 24GHz, 60GHz, or 77GHz band. The radar frame period can be determined within the range of 20ms to 200ms based on the speed of human movement, the speed of component movement, and data processing capabilities. When calibrating the same radar sensing node in two directions of motion, the same operating frequency band, frequency modulation bandwidth, number of sampling points, array channel configuration, and radar frame period are used to ensure that the range-angle echo maps corresponding to the two directions of motion have the same range and angle cell divisions.

[0034] Understandably, the process begins by extracting wall, door / window openings, and fixed furnishings from the living space data. Walls are represented as planar areas with thickness and height, door / window openings as openings within walls, and fixed furnishings as two-dimensional outlines or three-dimensional enclosed areas. The sensing task area can be determined based on the range of human activity in areas such as bedrooms, living rooms, and hallways, excluding areas inaccessible to humans, such as walls and cabinets. Node installation surfaces can be selected based on installation strength, power supply conditions, communication conditions, and spatial obstruction, avoiding surfaces that are chronically obscured by large fixed furnishings. Using a corner of a wall or a room's planar projection point as the coordinate origin, the coordinate directions and length units of walls, door / window openings, fixed furnishings, the sensing task area, and node installation surfaces are standardized to obtain spatial mapping data.

[0035] The first state endpoint, second state endpoint, motion trajectory, profile dimensions, material parameters, and component displacement are extracted from the data of movable spatial components. Movable spatial components can be sliding doors, sliding partitions, revolving doors, or electric curtains. The motion trajectory of a sliding component can be represented as a straight line or curved path between the two state endpoints, while the motion trajectory of a rotating component can be represented as an angular range formed around a rotation axis. Profile dimensions include the component's length, width, thickness, and height. Material parameters can include the relative permittivity of the component material, as well as the reflection coefficient and attenuation parameters within the radar sensing node's operating frequency band and actual incident angle range. The actual incident angle range is determined based on the radar sensing node's installation position, installation direction, and the movable spatial component's motion range. The reflection coefficient and attenuation parameters can be determined based on the component material's product data, material test results, or test data of similar materials within the same operating frequency band. Material parameters are associated with the component's profile surface and are used to characterize the component's reflection and attenuation characteristics to radar waves. The state-sensitive observation unit still determines these parameters based on the actually acquired bidirectional state switching data. Component displacement can be obtained from the encoder, stroke sensor, or position feedback interface of the drive device and converted into a unified unit of length or angle.

[0036] Based on the location and range of the node installation surface, the node installation position is determined by combining the effective detection range, beam direction, beamwidth, range resolution, and angular resolution of the radar sensing node. Specifically, the node beam should cover the sensing task area, with the beam center pointing towards the main human activity area. When a single node cannot cover the sensing task area, multiple node installation positions can be set on different node installation surfaces. The coordinates, installation height, and installation direction of each node installation position are correlated with the beam range, effective detection range, range resolution, and angular resolution of the corresponding radar sensing node to form a node parameter correlation relationship. The node installation height and installation direction can be determined based on the room height, the range of the sensing task area, and the product installation requirements of the radar sensing node.

[0037] Next, the motion direction of the movable spatial component is determined based on the first state endpoint, the second state endpoint, and the motion trajectory, and the component displacement is converted into the corresponding component pose. For translational components, the position of the component after moving along the motion trajectory is determined based on the component displacement, while the component orientation remains unchanged; for rotational components, the rotation angle is determined based on the conversion relationship between component displacement and rotation stroke, and then the component position and orientation are determined based on the rotation axis position and rotation angle; for foldable components, the rotation position of each connecting segment can be calculated separately. Combined with the component outline dimensions, the spatial occupancy range under different component displacements is obtained, and the pose correspondence is established.

[0038] Finally, using the spatial mapping data as a fixed spatial foundation, the contour dimensions, material parameters, and pose correspondences are written into the data layer of the movable spatial components, and the node installation positions and node parameter relationships are written into the data layer of the radar sensing nodes. A spatial association is then established between the movable spatial components, the sensing task area, and the radar sensing nodes, forming a spatial sensing model. This spatial sensing model can be stored hierarchically according to fixed spatial objects, movable spatial components, the sensing task area, and radar sensing nodes. The position of the components under different component displacements, the observation range of the nodes, and the spatial relationship between them can be queried using unified spatial coordinates.

[0039] This example establishes a correspondence between component displacement and component pose by mapping the living space structure, movable space components, and radar sensing nodes to the same spatial coordinate system. It also associates the node installation position with the radar sensing node parameters, enabling the spatial sensing model to represent the spatial relationship between movable space components and the radar observation range at different positions. This provides a data foundation for subsequent comparison of radar echoes during component state switching, determination of affected observation areas, and configuration of radar sensing parameters.

[0040] Please refer to Figure 2 , Figure 2 This diagram illustrates the bidirectional state switching and sensing scenario provided in the first embodiment of the residential space sensing configuration method of this application. It shows the positional relationships between radar sensing nodes, movable spatial components, sensing task areas, and human targets within the residential space. Radar sensing node R1 is located on the upper side of the residential space, radar sensing node R2 is located on the left side of the residential space, and radar sensing node R3 is located on the lower side of the residential space. The fan-shaped areas corresponding to each radar sensing node represent the radar observation range. Multiple radar observation ranges collectively cover the sensing task area located in the middle of the space, where the human target is located. A sofa and a table are also provided on the left side of the residential space. The fixed furnishings may affect the propagation of radar waves; the movable spatial component is located on the right side of the living space, and may be a sliding partition or a sliding door. The movable spatial component moves between a first state endpoint and a second state endpoint along the component displacement direction, wherein the first state endpoint corresponds to the open position of the component and the second state endpoint corresponds to the closed position of the component; the process of the movable spatial component moving from the open position to the closed position is marked as a forward switch, and the process of moving from the closed position to the open position is marked as a reverse switch, thereby representing the scenario in which each radar sensing node observes the sensing task area and human targets when the movable spatial component switches states in opposite directions.

[0041] As an example, the step of controlling the movable spatial component to switch states in opposite directions, synchronously acquiring component displacement and radar echoes to obtain bidirectional state switching data includes: during the field calibration period, controlling the radar sensing node to transmit detection radio waves and receive radar echoes; acquiring a first endpoint static echo at a first state endpoint, controlling the movable spatial component to move from the first state endpoint to a second state endpoint, synchronously acquiring component displacement and radar echoes to form a first component displacement sequence and a first radar echo sequence; acquiring a second endpoint static echo at the second state endpoint, controlling the movable spatial component to move in the opposite direction to the first state endpoint, synchronously acquiring component displacement and radar echoes to form a second component displacement sequence and a second radar echo sequence; performing time alignment and displacement resampling on the first component displacement sequence and the first radar echo sequence, and the second component displacement sequence and the second radar echo sequence, and combining the resampling results, the first endpoint static echo, and the second endpoint static echo into bidirectional state switching data.

[0042] It should be noted that the empty field calibration period refers to the time period in which there are no human targets in the living space, or no human target echoes are detected by the radar sensing nodes, and the environmental state remains stable; the unified timestamp refers to the acquisition time when the component displacement data and radar echo data are recorded using the same clock reference; displacement resampling refers to rearranging the data of the two motion directions according to the unified component displacement interval, so that the same component displacement corresponds to the radar echoes under the two motion directions; bidirectional state switching data refers to the round-trip motion data after time alignment and displacement resampling, as well as the combined data of the first endpoint static echo and the second endpoint static echo.

[0043] Understandably, before the movable spatial component begins to move, it is determined whether each radar sensing node detects human target echoes within multiple consecutive radar frames. If no human target echoes are detected for 3 to 10 consecutive seconds, and the fluctuation of the environmental echo amplitude does not exceed a preset airfield fluctuation threshold, the corresponding time period is designated as the airfield calibration period. The preset airfield fluctuation threshold can be determined based on the range of background echo fluctuations collected by the radar sensing nodes when the component is stationary, for example, by taking 2 to 4 times the standard deviation of the background echo amplitude.

[0044] The first and second state endpoints can be determined based on the limit signals of the drive device or component displacement feedback. For sliding doors, sliding partitions, or electric curtains, the two state endpoints correspond to the start and end positions of the drive stroke, respectively; for revolving doors, the two state endpoints correspond to the two boundary positions of the rotation angle range, respectively. When a component reaches a state endpoint and its displacement remains constant over multiple consecutive sampling periods, the component is determined to be stationary at the corresponding state endpoint.

[0045] The component remains stationary at the first state endpoint, and static echoes are acquired at the first endpoint according to the radar frame period. After the static echo acquisition is completed, the movable spatial component is controlled to move from the first state endpoint to the second state endpoint, and the component displacement and radar echo are synchronously recorded according to a unified timestamp, forming a first component displacement sequence and a first radar echo sequence. The movement speed of the translation component can be set in the range of 0.05m / s to 0.3m / s, and the angular velocity of the rotation component can be set in the range of 5° / s to 30° / s. The specific values ​​are determined according to the rated speed of the drive device, the radar frame period, and the component stroke to ensure that there are distinguishable component displacement changes between adjacent radar frames.

[0046] After the component reaches the second state endpoint, it remains stationary, and the static echo of the second endpoint is acquired. After the static echo acquisition is completed, the component is controlled to move in the opposite direction to the first state endpoint, and the component displacement and radar echo are recorded simultaneously to form a second component displacement sequence and a second radar echo sequence. The round-trip motion can use the same target velocity or different velocities. Subsequently, the data in the two motion directions are compared according to the component displacement rather than the acquisition time.

[0047] Component displacement data and radar echo data can be timestamped by the same controller, or they can be recorded separately and converted to a unified clock reference by the radar sensing configuration system. For data whose timestamps do not completely match, the displacements of adjacent components are found based on the acquisition time of the radar echo frame, and the corresponding component displacement is obtained by linear interpolation. When the component displacement error corresponding to the time alignment error does not exceed one displacement sampling interval, the corresponding data is retained; when it exceeds one displacement sampling interval, the corresponding radar echo frame is discarded. When the number of consecutive missing data reaches a preset number, the corresponding motion range is re-acquired.

[0048] After time alignment is completed, a uniform displacement sampling point is set between the first and second state endpoints. The displacement sampling interval can be determined based on the radar range resolution, component stroke, and component movement speed, for example, set to 5mm to 20mm; for rotating components, the angle sampling interval can be set to 0.5° to 2°. The displacement sequences of the first and second components are resampled respectively, and the data in the second motion direction are arranged in reverse order from the first state endpoint to the second state endpoint, so that each displacement sampling point corresponds to the first and second radar echoes.

[0049] When a component pauses briefly during movement, the average static echo value corresponding to the pause position can be retained, and repeatedly acquired echo frames can be merged into a single displacement sampling point. If a single displacement data point or a single radar echo frame is missing, interpolation can be performed based on adjacent sampling points. If consecutive missing data exceeds a preset number, no interpolation is performed on the missing interval, and the corresponding movement interval is re-acquired. The preset number of missing data points can be determined based on the radar frame period and displacement sampling interval, for example, set to 2 to 5 frames. Finally, the data from the two movement directions after displacement resampling, the static echo from the first endpoint, and the static echo from the second endpoint are combined into bidirectional state switching data.

[0050] This example synchronously collects the displacement of movable spatial components and radar echoes during their reciprocating motion under open field conditions, and resamples the data in both motion directions according to the uniform component displacement. This reduces the impact of differences in motion speed, acquisition time deviation, and short pauses on echo comparison, enabling the comparison of bidirectional radar echoes corresponding to the same component position. This provides a more stable data foundation for subsequent identification of observation units that repeatedly change with the component position.

[0051] As an example, the steps of determining the anchor node based on the non-sensitive observation unit, establishing the track constraint relationship between the anchor node and the affected node, and generating a state switching configuration scheme include: determining the non-sensitive coverage ratio and the common coverage area with the affected node from radar sensing nodes that do not belong to the affected node, based on the non-sensitive observation unit corresponding to each radar sensing node, and determining the anchor node based on the non-sensitive coverage ratio and the common coverage area; establishing the coordinate transformation relationship between the node coordinate system of the anchor node and the node coordinate system of the affected node; obtaining the human target track output by the anchor node, generating a target prediction area based on the human target track, extracting the observation gate parameter and component echo shielding unit from the parameter transfer relationship, and determining the constrained observation gate accordingly; establishing the track constraint relationship based on the coordinate transformation relationship, the human target track and the constrained observation gate, and generating a state switching configuration scheme based on the human target track calling order of the anchor node and the parameter update order of the affected node.

[0052] It should be noted that the non-sensitive coverage ratio refers to the ratio of the number of non-sensitive observation units corresponding to the radar sensing node within the sensing mission area to the number of effective observation units; the common coverage area refers to the overlapping area in the living space between the non-sensitive observation range of the candidate anchoring node and the effective observation range of the affected node; the target prediction area refers to the spatial range in which the human target may appear in subsequent observation periods based on the human target trajectory; the constrained observation gate refers to the human target echo search range determined by combining the target prediction area, the observation gate of the affected node, and the component echo shielding unit; and the state switching configuration scheme refers to the order of calling the human target trajectory of the anchoring node, the order of updating the parameters of the affected node, the data calling relationship, and the human target echo association rules during the state switching of movable space components.

[0053] An effective observation unit refers to a range-angle unit located within the sensing task area and whose echo signal-to-noise ratio reaches a preset signal-to-noise ratio threshold; an effective observation range refers to the spatial range formed after the effective observation unit is mapped onto the living space; a node coordinate system refers to a coordinate system established with the installation position of the radar sensing node as the origin and the node installation direction as the reference direction; and a handover area refers to the area in the common coverage area of ​​adjacent candidate anchoring nodes used to switch human target trajectory output nodes.

[0054] Understandably, the process begins by selecting candidate anchoring nodes from radar sensing nodes that are not affected nodes. For each candidate anchoring node, non-sensitive observation units are mapped to the sensing task area, and the ratio of the number of non-sensitive observation units to the number of effective observation units is calculated to obtain the non-sensitive coverage ratio. Then, the non-sensitive observation range of the candidate anchoring node and the effective observation range of each affected node are mapped to the spatial sensing model, and the overlapping portion is taken as the common coverage area. The non-sensitive coverage ratio threshold can be selected within the range of 50% to 80%, and the common coverage area should at least cover the area traversed by a human target when entering the observation range of an affected node from the observation range of a candidate anchoring node. The specific threshold is determined based on the size of the sensing task area, radar resolution, and the number of nodes.

[0055] When selecting anchor nodes based on non-sensitive coverage ratio and shared coverage area, radar sensing nodes with a non-sensitive coverage ratio reaching the threshold and sharing a coverage area with all affected nodes are given priority. If multiple radar sensing nodes meet the criteria, the minimum shared coverage area between each radar sensing node and the affected nodes is compared first, followed by the non-sensitive coverage ratio. The radar sensing node with the larger minimum shared coverage area and higher non-sensitive coverage ratio is selected as the anchor node. Anchor nodes continuously output human target tracks during the movement of movable spatial components and do not participate in echo correlation of areas affected by component movement.

[0056] Based on the node installation positions and orientations recorded by the spatial perception model, an initial coordinate transformation relationship is established between the coordinate systems of the anchored nodes and the affected nodes. When the node installation orientation has been calibrated and only coordinate translation needs correction, a fixed reference position can be set in the common coverage area. The anchored node and the affected node measure this fixed reference position respectively, and the coordinate translation is corrected based on the measurement results. When it is also necessary to correct the node installation orientation deviation, at least two non-overlapping fixed reference positions are set in the common coverage area. The anchored node and the affected node measure each fixed reference position respectively, and the coordinate rotation and coordinate translation are determined based on the corresponding measurement positions. The coordinate transformation relationship is used to transform the human target position, velocity, and direction of motion output by the anchored node to the coordinate system of the affected nodes.

[0057] The system acquires the human target trajectory output by the anchor node. Based on the current human target position and the velocity vector formed by the movement speed and direction, it performs constant-speed extrapolation according to the prediction duration to obtain the predicted human target position for the next 1 to 3 radar frames. This is then combined with the radar range resolution, angular resolution, and the possible displacement range of the human body within the corresponding time period to generate the target prediction region. The range boundary and angular boundary of the target prediction region can be extended by 1 to 3 range units and 1 to 3 angular units on both sides of the predicted position, respectively. The number of extensions is determined based on the radar frame period, the human body's movement speed, and the node measurement error.

[0058] For each component displacement, the corresponding observation gate parameters and component echo shielding units are extracted from the parameter transfer relationship. The observation gate for the affected node is determined based on the observation gate parameters. The target prediction region is transformed to the coordinate system of the affected node. The intersection region of the transformed target prediction region and the observation gate is taken, and the component echo shielding units are excluded from the intersection region to obtain the constrained observation gate for the corresponding component displacement. When the intersection region is completely covered by the component echo shielding units, the association of the human target echo with the corresponding affected node is suspended. The anchor node continues to maintain the human target track until the constrained observation gate corresponding to the subsequent component displacement reappears.

[0059] Based on this, component displacement, affected node identifiers, coordinate transformation relationships, human target tracks, and corresponding constrained observation gates are associated to form track constraint relationships. A state switching configuration scheme is generated in the following order: first reading component displacement, then calling the corresponding configuration parameters, and finally generating constrained observation gates and associating them with human target echoes. When multiple human targets exist, target prediction regions are generated based on each human target track. When multiple target prediction regions overlap, the position deviation, velocity deviation, and motion direction deviation between the echo point and each human target track are calculated, and human target tracks with any deviation exceeding the corresponding preset association range are excluded. When only one human target track is retained, the echo point is associated with that human target track. When multiple human target tracks are retained, priority is given to associating human target tracks with smaller position deviations; if the position deviations are the same, the velocity deviation and motion direction deviation are compared sequentially. The corresponding preset association range can be determined based on radar range resolution, angular resolution, frame period, radar velocity measurement error, and the maximum human movement speed.

[0060] When no single radar sensing node can cover the entire handover area, multiple candidate anchor nodes can be selected to provide human target tracks in segments. Each candidate anchor node corresponds to a continuous sensing area and retains a shared coverage area with adjacent candidate anchor nodes. After a human target enters the shared coverage area, the position and velocity of the human target output by adjacent candidate anchor nodes are compared. If the deviations of both are within a preset handover range, the subsequent track output node is switched to the next candidate anchor node. The position deviation threshold can be set to the spatial distance corresponding to 1 to 3 range resolution units, and the velocity deviation threshold is determined based on the normal walking speed range of a human and the radar velocity measurement error.

[0061] This example identifies anchor nodes from radar sensing nodes less affected by component motion and establishes coordinate and track relationships between anchor nodes and affected nodes using shared coverage areas. This allows the human target track output by the anchor nodes to limit the echo search range of the affected nodes. By combining observation gate parameters corresponding to different component displacements and component echo shielding units to generate constrained observation gates, the number of component motion echoes entering the human target echo search range is reduced. At the same time, when the coverage of a single anchor node is insufficient, multiple anchor nodes are used to maintain the human target track in segments, improving the continuity of human target echo correlation during component state switching.

[0062] As an example, the step of updating the affected nodes based on the current component displacement and the parameter transfer relationship when the movable space component switches states, and associating human target echoes with the state switching configuration scheme and the trajectory constraint relationship to obtain the perception configuration result includes: acquiring the current component displacement, current motion direction of the movable space component, and the current radar echoes of each radar perception node; determining the current configuration parameters from the parameter transfer relationship based on the current component displacement and the current motion direction, and updating the affected nodes according to the state switching configuration scheme; and associating the current configuration parameters from the parameter transfer relationship based on the current component displacement and the current motion direction. The corresponding constrained observation gate is determined in the trajectory constraint relationship, and the constrained observation gate is adjusted according to the human target trajectory output by the anchor node; the current radar echo of the affected node is processed according to the current configuration parameters, and the human target echo falling into the adjusted constrained observation gate is determined from the processed current radar echo, and the human target echo is associated with the human target trajectory; after the movable space component reaches the state endpoint, the current endpoint static echo is collected, and the parameter transfer relationship is updated according to the current endpoint static echo and the endpoint static echo in the bidirectional state switching data to obtain the perception configuration result.

[0063] It should be noted that the current configuration parameters refer to the background echo parameters, component echo shielding unit, observation gate parameters, and detection threshold corresponding to the current component displacement and current direction of motion; the endpoint echo deviation refers to the amplitude deviation and peak position deviation between the current endpoint static echo and the corresponding endpoint static echo in the bidirectional state switching data.

[0064] Understandably, after a movable spatial component begins to switch states, its displacement is read according to the radar echo acquisition cycle, and the current direction of motion is determined based on the direction of change of the component displacement in two adjacent frames. If the drive device can output the running direction, the drive direction signal can also be read directly. The reading cycle of the current component displacement can be the same as or less than the radar frame cycle, and the component displacement and the current radar echo are associated with a unified timestamp. When the component displacement reading time is inconsistent with the radar echo frame acquisition time, linear interpolation is performed on the component displacement corresponding to the radar echo frame acquisition time based on two adjacent component displacements to avoid using data from different component positions in the same radar frame.

[0065] The parameter transfer relationship is queried based on the current component displacement and the current direction of motion. When the current component displacement is consistent with the displacement index node, the corresponding configuration parameters are read directly. When the current component displacement is between two adjacent displacement index nodes, the current configuration parameters are determined according to the parameter interpolation method described later, and the current configuration parameters of the affected nodes are updated before processing the corresponding current radar echo.

[0066] Next, based on the current component displacement, the corresponding constrained observation gate is read from the trajectory constraint relationship, and the current human target trajectory output by the anchor node is obtained. Based on the current position, velocity, and direction of movement in the human target trajectory, the target position in subsequent radar frames is calculated. Then, the constrained observation gate is adjusted according to the radar range resolution, angular resolution, and the possible displacement of the human body within a radar frame. The range boundary of the constrained observation gate can be extended by 1 to 3 range units on both sides of the predicted position, and the angular boundary can be extended by 1 to 3 angular units on both sides of the predicted angle. The specific extension range is determined based on the human body's movement speed, radar range resolution, radar angular resolution, and node measurement error.

[0067] Echo points falling into the current component echo shielding unit are excluded from the current radar echoes of the affected nodes. Then, echo points that reach the current detection threshold are selected from the adjusted constrained observation gates. The deviations of the echo points from the human target tracks in terms of position, velocity, and direction of motion are compared. Echo points with deviations within a preset association range are identified as human target echoes and associated with the corresponding human target tracks. The position association range can be determined based on the radar range resolution and angular resolution, while the velocity association range can be determined based on the radar velocity measurement error and the normal human movement speed range. When multiple human targets exist, constrained observation gates are generated according to the human target track identifiers, and the echo points are associated with the corresponding human target tracks according to the comparison order of position deviation, velocity deviation, and direction of motion deviation described above.

[0068] After a movable spatial component reaches its state endpoint, endpoint determination is performed based on the limit signal from the drive device. If no limit signal is set, the component can be determined to have reached the state endpoint when its displacement remains within the endpoint tolerance range for 3 to 10 consecutive sampling cycles. The endpoint tolerance can be 0.5% to 2% of the component's total travel. After the component remains stationary at the state endpoint, radar echoes are continuously collected for 1 to 5 seconds. If a human target is present during the collection period, the constrained observation gate corresponding to each radar frame is determined based on the human target trajectory output by the anchor node. Range angle units covered by the constrained observation gates during the collection period are excluded, and the radar echoes of the remaining range angle units are averaged to obtain the current endpoint static echo. If no human target is present during the collection period, the radar echoes of each range angle unit are averaged to obtain the current endpoint static echo.

[0069] The static echo of the current endpoint is compared with the static echo of the corresponding endpoint in the bidirectional state switching data according to the distance angle unit that can be used for endpoint comparison. Based on the endpoint amplitude deviation and endpoint peak position deviation of each distance angle unit, the parameter transfer relationship is kept unchanged, the parameters of the corresponding displacement interval are locally corrected, or the bidirectional state switching data of the corresponding displacement interval is re-acquired. After the processing is completed, the updated node configuration and the human target echo association result are used as the perception configuration result.

[0070] This example ensures that the background echo processing of affected nodes, the component echo exclusion range, the human target observation range, and the detection threshold correspond to the current position of the component by calling configuration parameters according to the current component displacement and current direction of motion. It also adjusts the constrained observation gate using the human target trajectory output by the anchored node, reducing the likelihood of component motion echoes being associated with human target echoes. Furthermore, it selects to maintain parameters, perform local corrections, or recalibrate based on endpoint echo deviations, avoiding frequent parameter updates due to normal echo fluctuations and reducing the likelihood of continuing to use mismatched parameters after significant changes in the space environment.

[0071] As an example, the step of updating the parameter transfer relationship based on the current endpoint static echo and the endpoint static echo in the bidirectional state switching data includes: comparing the current endpoint static echo with the corresponding endpoint static echo in the bidirectional state switching data according to distance angle units to obtain the endpoint amplitude deviation and endpoint peak position deviation of each distance angle unit; selecting change observation units from the state-sensitive observation units where the endpoint amplitude deviation or the endpoint peak position deviation reaches the corresponding preset deviation threshold; determining the component displacement interval associated with the change observation unit based on the bidirectional echo ridge line, merging adjacent or overlapping component displacement intervals to obtain the parameter transfer relationship to be updated. New displacement interval; when the endpoint peak position deviation of the change observation unit does not reach the preset peak position update threshold, the echo amplitude reference and detection threshold corresponding to the displacement interval to be updated are corrected according to the endpoint amplitude deviation, and the corresponding echo amplitude fluctuation range, component echo shielding unit and observation gate parameters are retained; when the endpoint peak position deviation reaches the preset peak position update threshold, the bidirectional state switching data corresponding to the displacement interval to be updated is re-acquired, and the background echo parameters, directional difference, component echo shielding unit, detection threshold, observation gate parameters and directional parameter branch or bidirectional shared parameters corresponding to the displacement interval to be updated are updated according to the re-acquired data.

[0072] It should be noted that endpoint amplitude deviation refers to the normalized amplitude difference between the current endpoint static echo and the corresponding endpoint static echo in the bidirectional state switching data within the same distance-angle unit; endpoint peak position deviation refers to the normalized position difference of the corresponding local peak in the distance and angle dimensions in two sets of endpoint static echoes; change observation unit refers to the state-sensitive observation unit where the endpoint amplitude deviation or endpoint peak position deviation reaches the corresponding preset deviation threshold; displacement range to be updated refers to the component displacement range determined based on the bidirectional echo ridge line, where parameters need to be corrected or data needs to be reacquired; preset peak position update threshold refers to the peak position deviation limit used to distinguish between echo amplitude changes and echo spatial position changes.

[0073] Understandably, the current endpoint static echo is matched with the corresponding endpoint static echo in the bidirectional state switching data according to distance and angle units. The current echo amplitude, reference echo amplitude, and the position of the corresponding local peak are extracted for each state-sensitive observation unit. The endpoint amplitude deviation is calculated based on the current echo amplitude and the reference echo amplitude, and the endpoint peak position deviation is calculated based on the positional changes of the current local peak and the reference local peak in the distance and angle dimensions. The calculation formulas are as follows: In the formula, This represents the endpoint amplitude deviation of the q-th distance angle unit; This represents the current echo amplitude of the q-th distance angle unit; This represents the reference echo amplitude of the q-th distance angle unit; This represents the amplitude stability term, which can be taken as the average amplitude of the reference echo. Doubled times; This represents the peak position deviation at the endpoint of the q-th distance angle unit; and These represent the distance positions of the current local peak and the reference local peak, respectively. and These represent the angular positions of the current local peak and the reference local peak, respectively; Δr represents the distance element interval; Δθ represents the angular element interval; and q represents the sequence number of the distance angular element.

[0074] Observation units that reach a preset amplitude deviation threshold or a preset peak position deviation threshold from the state-sensitive observation units are selected to obtain change observation units. The preset amplitude deviation threshold can be determined based on the static echo amplitude fluctuations collected multiple times when the component is at the corresponding state endpoint. It is preferably the 95th percentile of the absolute value of the historical endpoint amplitude deviation, but it can also be set to 0.05 to 0.20 when historical data is unavailable. The preset peak position deviation threshold can be set to 0.5 to 1, which means that the normalized position difference between the current local peak and the reference local peak reaches 0.5 to 1 unit interval.

[0075] For each change observation unit, a bidirectional echo ridge line is found that passes through the change observation unit or its adjacent distance angle unit at the corresponding component displacement at the current state endpoint, and the component displacement change is tracked along the bidirectional echo ridge line. The component displacement range where the echo ridge line continuously passes through the change observation unit and its adjacent distance angle unit is taken as the associated component displacement interval. To cover the ridge line extraction error, the two ends of the component displacement interval can be extended by 1 to 2 displacement sampling intervals, but the extended range does not exceed the first state endpoint and the second state endpoint. When adjacent component displacement intervals overlap, or the interval does not exceed 2 displacement sampling intervals, the corresponding intervals are merged into a displacement interval to be updated.

[0076] For each displacement interval to be updated, when the peak position deviation of the endpoints of the associated change observation units does not reach the preset peak position update threshold, the change is determined to be mainly manifested as echo amplitude drift, without changing the position of the component echo in the distance-angle space. The preset peak position update threshold can be set to 1 to 2, and is not less than the preset peak position deviation threshold. Based on the echo amplitude of the bidirectional echo ridge at different component displacements, the correction weight for the propagation of the endpoint amplitude deviation to the displacement interval to be updated is determined, and the calculation formula is as follows: In the formula, This represents the correction weight of the q-th distance angle element at the component displacement x; This represents the average echo amplitude at component displacement x, corresponding to the echo ridges in the two motion directions; when only one of the two motion directions has a valid echo ridge. The echo amplitude of this effective echo ridge is used. This represents the component displacement corresponding to the current state endpoint; The stable term representing the ridge amplitude can be taken as the average ridge amplitude of the echo within the displacement interval to be updated. Doubled times; and These represent the echo amplitude references before and after the update, respectively; and These represent the detection thresholds before and after the update, respectively; 'c' represents the type of the parameter group to be updated, which can be a directional parameter branch corresponding to the current motion direction or a bidirectional shared parameter. When updating the detection threshold, the difference between the original detection threshold and the echo amplitude reference is retained, ensuring that the detection margin does not change due to overall changes in the background amplitude; the component echo shielding unit and observation gate parameters remain unchanged.

[0077] When a change in the peak position deviation at the endpoints of the displacement interval to be updated reaches the preset peak position update threshold, it is determined that the component echo position has changed. The movable spatial component is controlled to move in two opposite directions within the displacement interval to be updated, and the component displacement and radar echo are re-acquired according to the radar frame parameters and displacement sampling intervals used in the original bidirectional calibration process. The acquisition range can be increased by 1 to 2 displacement sampling intervals at both ends of the displacement interval to be updated, so that the old and new data can be seamlessly connected at the interval boundary. Based on the re-acquired data, the background echo parameters, directional difference, bidirectional echo ridge, component echo shielding unit, detection threshold, observation gate parameters, and directional parameter branch or bidirectional shared parameters corresponding to the displacement interval to be updated are redefined, while the parameters outside the interval remain unchanged.

[0078] Before updating the parameter transfer relationship, save the current parameter transfer relationship and its version number, and use the updated parameter transfer relationship as a candidate version. The candidate version needs to check whether each displacement index node in the displacement interval to be updated is associated with the echo amplitude benchmark, echo amplitude fluctuation range, component echo shielding unit, detection threshold, observation gate parameter, and parameter type identifier; check whether the difference between the parameter at the boundary of the displacement interval to be updated and the parameter of the adjacent unupdated displacement index node is lower than the preset boundary continuity threshold; and check whether the amplitude residual between the updated echo amplitude benchmark and the current endpoint static echo is lower than the preset residual threshold. If the checks pass, the candidate version is enabled; if the checks fail, the previous version is restored, and the corresponding displacement interval to be updated is marked as the interval that needs to be recalibrated bidirectionally. The preset boundary continuity threshold and the preset residual threshold can be determined based on the 95th percentile of the parameter difference between adjacent displacement index nodes and the endpoint amplitude residual in multiple calibration data.

[0079] This example distinguishes between environmental echo amplitude drift and component echo spatial position changes by separately judging the amplitude and peak position changes of the endpoint echo. For cases where the peak position does not change significantly, only the echo amplitude reference and detection threshold within the associated displacement interval are corrected, reducing repeated acquisition and overall parameter updates. For cases where the peak position change reaches the threshold, the bidirectional state switching data of the corresponding displacement interval is reacquired, and the background echo parameters, directional difference, component echo shielding unit, detection threshold, observation gate parameters, and directional parameter branch or bidirectional shared parameters of the corresponding displacement interval are redefined, avoiding the continued use of mismatched configuration parameters. At the same time, the recoverability of the parameter transfer relationship update process is preserved through version verification and rollback.

[0080] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3 , Figure 3This is a flowchart illustrating the second embodiment of the residential space sensing and configuration method of this application. Step S30 of the residential space sensing and configuration method includes steps S31 to S34: Step S31: Determine the range of the range angle unit corresponding to the radar sensing node according to the spatial sensing model, and extract the component displacement sequence, radar echo sequence and endpoint static echo of the two motion directions from the bidirectional state switching data. Step S32: Convert the radar echo sequences in the two directions of motion into range angle echo maps, and arrange the range angle echo maps according to the displacement of the same component; Step S33: Calculate the change in echo amplitude of each distance angle unit in the two motion directions based on the static echo of the endpoint, extract the echo ridges corresponding to the two motion directions, and determine the bidirectional response consistency based on the consistency between the change in echo amplitude in the two motion directions and the consistency between the echo ridges. Step S34: Determine the state-sensitive observation unit and the non-sensitive observation unit based on the bidirectional response consistency, and determine the radar sensing node containing the state-sensitive observation unit as the affected node.

[0081] It should be noted that the change in echo amplitude refers to the normalized change of the current echo amplitude of the same distance angle unit relative to the static echo amplitude at the endpoint; the echo ridge line refers to the trajectory formed by the continuous local peaks in the distance angle echo map during the movement of the movable spatial component; the bidirectional response consistency refers to a dimensionless value used to characterize the repeatability of amplitude changes, the proximity of echo ridge lines, and the magnitude of changes of the same distance angle unit in two directions of movement.

[0082] Range-angle unit refers to a radar observation unit defined by a range interval and an angle interval; range-angle echo map refers to a data map of echo amplitudes arranged according to range units and angle units; range-angle space refers to the radar observation space with range and angle as coordinate dimensions; bidirectional echo ridge refers to the combination of echo ridges matched according to the displacement of the same component in two directions of motion.

[0083] Understandably, the process begins by reading the installation location, orientation, effective detection range, and beam coverage of each radar sensing node based on the spatial perception model. This transforms the sensing task area into the corresponding node coordinate system, determining the boundaries of the sensing task area in both the range and angle dimensions. Range units are then divided according to the range resolution of the radar sensing nodes, and angle units are divided according to the angle resolution. The range and angle unit ranges are formed by combining the range units and angle units falling within the sensing task area. Subsequently, the displacement sequences of the first and second components, the first radar echo sequence, the second radar echo sequence, and the static echoes corresponding to the two state endpoints are extracted from the two-way state switching data according to the motion direction identifier and timestamp.

[0084] For each radar echo frame, a range-dimensional Fourier transform is performed on the intermediate frequency sampling data from different receiving channels. Then, an angle-dimensional Fourier transform is performed based on the phase difference between the array channels to obtain the echo amplitude of each range-angle cell, forming a range-angle echo map. When the radar sensing node can directly output the range-angle echo map, the node's output results can also be read. The range cell width uses the range resolution of the radar sensing node, and the angle cell width uses the angle resolution. Data exceeding the aforementioned range-angle cell range is not included in subsequent processing.

[0085] The data for the two motion directions are arranged according to the component displacement order from the first state endpoint to the second state endpoint. The data for the first motion direction retains its original arrangement order, while the data for the second motion direction is rearranged according to the component displacement in ascending order, so that each component displacement sampling point corresponds to the distance-angle echo map in both motion directions. When a certain displacement sampling point does not have a directly corresponding radar echo frame, linear interpolation is performed using the echo amplitude values ​​corresponding to the two adjacent component displacement sampling points. When more than 2 to 5 consecutive sampling points are missing, the corresponding missing interval is not used.

[0086] The endpoint static echo can be obtained by averaging multiple radar echo frames during the period when the component remains stationary at the endpoint. The inter-frame variance of the static echoes at the two endpoints is calculated in each range angle cell. The static echo at the endpoint with the smaller average inter-frame variance is selected as the common reference for the two directions of motion, so that the amplitude changes of the bidirectional echoes use the same reference. When the ratio of the difference between the average inter-frame variances of the two endpoints to the larger average inter-frame variance is less than 10%, the static echo corresponding to the first state endpoint can be preferentially selected as the common reference.

[0087] For each distance angle unit and each component displacement sampling point, the absolute difference between the current echo amplitude and the common reference echo amplitude is calculated, and then normalized using the common reference echo amplitude to obtain the echo amplitude change. The calculation formula is as follows: In the formula, This represents the change in echo amplitude of the q-th distance angle element under the motion direction d and component displacement x; This indicates the current echo amplitude value for the corresponding distance angle unit; Indicates the common reference echo amplitude; The amplitude stability term can be represented by the average amplitude of the common reference echo from all distance and angle units. Doubled 10 times; q represents the distance angle unit number; d represents the direction of motion.

[0088] Local peak values ​​of echo amplitude variation are extracted from the distance-angle echo maps of each component displacement sampling point. Local peak values ​​whose echo amplitude variation reaches a preset variation threshold are retained. The preset variation threshold can be the 95th percentile of the normalized echo fluctuation in the static echo of the common reference endpoint. Local peak values ​​with similar distance and angle positions in adjacent sampling points are connected according to the component displacement sequence to form echo ridges corresponding to two motion directions. The distance position difference between adjacent local peak values ​​can be limited to no more than 1 to 2 distance units, and the angle position difference can be limited to no more than 1 to 2 angle units. Peak segments that cannot continuously pass through at least 3 component displacement sampling points are not considered as echo ridges.

[0089] As an example, the step of determining the bidirectional response consistency based on the consistency between the echo amplitude changes in the two motion directions and the consistency between the echo ridges includes: for each distance angle unit, aligning the echo amplitude change sequences corresponding to the two motion directions according to the component displacement; determining the amplitude repeatability and normalized average change amplitude based on the aligned two echo amplitude change sequences, and determining the normalized ridge deviation based on the echo ridges corresponding to the two motion directions; calculating the bidirectional response consistency based on the amplitude repeatability, the normalized average change amplitude, and the normalized ridge deviation; and determining the distance angle unit whose bidirectional response consistency reaches a preset consistency threshold and whose normalized average change amplitude reaches a preset change threshold as a state-sensitive observation unit.

[0090] It should be noted that amplitude repeatability refers to the degree to which the echo amplitude change sequence of the same distance and angle unit in two motion directions shows the same trend with the component displacement; normalized average change amplitude refers to the average value of the echo amplitude change in two motion directions at the effective aligned displacement point; normalized ridge deviation refers to the average value of the position difference of the corresponding echo ridges in the distance and angle dimensions in two motion directions after normalization by radar resolution; effective aligned displacement point refers to the component displacement sampling point where there is an effective echo amplitude change in both motion directions.

[0091] Understandably, for each distance angle unit, the displacement range commonly covered by the component displacement sequences in two motion directions is first determined, and alignment displacement points are set according to the displacement sampling interval. When both motion directions have corresponding sampling values, the change in echo amplitude is directly obtained; when the sampling position in a certain motion direction is between two adjacent displacement sampling points, linear interpolation is used to obtain the corresponding value. If the missing component displacement interval exceeds two displacement sampling intervals, or if the corresponding radar echo is continuously lost for more than two frames, this interval is not considered a valid alignment displacement point.

[0092] Furthermore, the average change in echo amplitude at the effective alignment displacement point in the two motion directions is taken as the signal amplitude, and the normalized inter-frame fluctuation standard deviation of the static echo at the endpoint in the corresponding range angle cell is taken as the noise amplitude. The signal-to-noise ratio (SNR) for each motion direction is calculated, and the average of the SNRs for the two motion directions is taken as the average SNR of the range angle cell. The calculation formula is as follows: In the formula, This represents the signal-to-noise ratio of the q-th distance angle unit in the motion direction d; This represents the average value of the corresponding echo amplitude change at the effective alignment displacement point; This represents the normalized inter-frame variability standard deviation of the endpoint static echo in the q-th distance angle unit; This represents the signal-to-noise ratio stability term, which can be set to... to ; q represents the distance angle unit number; d represents the direction of motion. It has the same dimensions as the normalized echo amplitude change, and is used to avoid the signal-to-noise ratio being uncalculated when the signal amplitude or noise amplitude is 0.

[0093] If the average signal-to-noise ratio (SNR) of a range angle element in both motion directions is lower than a preset SNR threshold, or if the effective alignment displacement points are less than 60% of all alignment displacement points, the bidirectional response consistency of that range angle element will not be calculated. The preset SNR threshold can be set from 3dB to 6dB, and the specific value is determined based on the noise fluctuation of the radar sensing node in the airspace.

[0094] The amplitude repeatability is calculated based on the change in echo amplitude at the effective alignment displacement point between the two motion directions. The calculation formula is as follows: In the formula, This represents the amplitude repeatability of the q-th distance angle unit; and These represent the changes in echo amplitude at the nth effective alignment displacement point of the qth distance angle element in the two motion directions, respectively. and These represent the average changes in echo amplitude corresponding to the two directions of motion; This indicates the number of effective alignment displacement points contained in the q-th distance angle unit; n represents the index of the effective alignment displacement point; and q represents the index of the distance angle unit. The amplitude repeatability ranges from -1 to 1, with a positive value indicating that the echo amplitude changes in the two motion directions have the same trend.

[0095] When the echo amplitude change sequence in any direction of motion has no fluctuation, resulting in a denominator of 0 in the formula, the corresponding distance angle unit is considered to have no effective amplitude repeatability and is not included in the state sensitivity determination.

[0096] The normalized average change amplitude is calculated based on the change in echo amplitude at the effective aligned displacement point in both motion directions. The calculation formula is as follows: In the formula, This represents the normalized average variation amplitude of the q-th distance angle unit; the meanings of the remaining characters are the same as in the aforementioned formula. Since the change in echo amplitude has been normalized using the static echo amplitude at the endpoints, the normalized average variation amplitude is a dimensionless quantity.

[0097] The normalized ridgeline deviation is determined based on the echo ridgelines corresponding to the two motion directions. First, the ridgeline positions of the two motion directions are matched at the same component displacement. Then, the distance position difference is divided by the distance element interval, and the angular position difference is divided by the angular element interval. The calculation formula is as follows: In the formula, This represents the normalized ridgeline deviation corresponding to the q-th distance angle unit; and These represent the distance positions corresponding to the two motion directions at the m-th ridge matching position; and These represent the angular positions corresponding to the two directions of motion; Δr represents the distance unit interval. Indicates the angular unit interval; This represents the number of valid ridgeline matching positions; m represents the sequence number of the ridgeline matching position. The normalized ridgeline deviation is a dimensionless quantity.

[0098] When there are fewer than 3 valid ridgeline matching positions, or when the number of displacement points jointly covered by the echo ridgelines of the two motion directions is less than 50%, the normalized ridgeline deviation and bidirectional response consistency of the corresponding distance angle unit are recorded as invalid, and this distance angle unit does not participate in the state sensitivity determination corresponding to the current bidirectional state switching data; when the valid ridgeline matching condition is met, the bidirectional response consistency is calculated based on the amplitude repeatability, normalized average change amplitude, and normalized ridgeline deviation, and the calculation formula is as follows: in, This represents the bidirectional response consistency of the q-th distance angle unit; This represents the amplitude repeatability between the two echo amplitude change sequences corresponding to the q-th distance angle unit; This represents the normalized average change magnitude corresponding to the q-th distance angle unit; This represents the normalized ridge line deviation between the two echo ridge lines corresponding to the q-th distance angle unit; The amplitude scale term can be 0.5 to 1 times the preset change threshold, which is used to reduce the bidirectional response consistency when the normalized average change amplitude is close to the preset change threshold; q represents the sequence number of the distance angle unit.

[0099] Distance and angle units whose bidirectional response consistency reaches a preset consistency threshold and whose normalized average change amplitude reaches a preset change threshold are initially identified as state-sensitive observation units. Distance and angle units with effective echoes but not meeting the state-sensitive criteria are identified as non-sensitive observation units. The preset change threshold can be the 95th percentile of the normalized echo fluctuation in the stationary data of the open field; when historical open field data is lacking, it can be set to 0.05 to 0.2. The preset consistency threshold can be set to 0.5 to 0.8, and determined based on the recurrence ratio of state-sensitive observation units in multiple component reciprocating movements.

[0100] This example reduces the impact of different component movement speeds and acquisition times on bidirectional data comparison by aligning the echo amplitude change sequences of the two movement directions according to the component displacement. Amplitude repeatability reflects whether the change trends of the two movement directions are consistent, normalized average change amplitude eliminates background fluctuations with small change amplitudes, and normalized ridge deviation reflects the degree of positional repeatability of component echoes in distance-angle space. The three factors together determine the bidirectional response consistency, reducing the possibility that random noise, scattered echo peaks, and low signal-to-noise ratio observation units are identified as state-sensitive observation units.

[0101] For the initially identified state-sensitive observation units, connected component processing is performed according to the 8-neighborhood relationship of the range and angle dimensions. If a connected component contains fewer than 2 to 5 range or angle units, or appears only near a single component displacement sampling point, the corresponding connected component is filtered out. If two connected components are separated by no more than one range or angle unit, and the corresponding echo ridges are continuous in the component displacement direction, the two connected components are merged. After connected component processing, radar sensing nodes containing state-sensitive observation units are identified as affected nodes.

[0102] This embodiment arranges radar echoes from two directions of motion according to the displacement of the same component and uses a static echo at a unified endpoint as the amplitude variation benchmark, enabling bidirectional echoes to be compared at the same component position. By combining amplitude variation repeatability, echo ridge deviation, and variation amplitude, the consistency of bidirectional response is determined, reducing the impact of random echo fluctuations and scattered peaks on the determination of state-sensitive observation units. At the same time, by merging and filtering connected components, the observation area that continuously changes with the component movement is retained, providing a basis for determining radar sensing nodes affected by component state switching.

[0103] As an example, the step of establishing a parameter transfer relationship based on the state-sensitive observation unit that changes with the component displacement includes: determining background echo parameters based on the original echo amplitudes corresponding to the two motion directions under the same component displacement for each distance angle unit in the state-sensitive observation unit, and determining the directional difference based on the change in the bidirectional echo amplitude under the same component displacement; generating a component echo shielding unit based on the ridge center and ridge width of the bidirectional echo ridge; determining the sensing task area from the spatial sensing model, determining the detection threshold based on the background echo parameters and the endpoint static echo, and determining the observation gate parameter based on the intersection position of the component echo shielding unit and the sensing task area; generating directional parameter branches or bidirectional shared parameters based on the directional difference; and arranging the background echo parameters, the component echo shielding unit, the detection threshold, the observation gate parameter, and the directional parameter branches or bidirectional shared parameters according to the component displacement to establish a parameter transfer relationship.

[0104] It should be noted that the background echo parameters include the echo amplitude reference and the echo amplitude fluctuation range; the directional difference refers to the normalized difference between the echo amplitude changes corresponding to the two motion directions under the same component displacement for the same distance and angle unit; the ridge center refers to the distance and angle positions of the echo ridge line corresponding to a component displacement sampling point; the ridge width refers to the unit range continuously covered by the echo ridge line in the distance and angle dimensions; the observation gate parameters include the lower distance limit, upper distance limit, lower angle limit, and upper angle limit of the human target echo search range; the directional parameter branches refer to the configuration parameters saved for the motion directions of the two components respectively; the bidirectional shared parameters refer to the configuration parameters jointly invoked by the motion directions of the two components.

[0105] Understandably, for each state-sensitive observation unit, the change in echo amplitude and the original echo amplitude are read for the two motion directions under the same component displacement. For each motion direction, the original echo amplitude for that motion direction is used as the directional echo amplitude reference. The directional echo amplitude fluctuation range is determined based on the echo amplitude fluctuation within the current component displacement neighborhood and the fluctuation range of the endpoint static echo. The current component displacement neighborhood can include the current component displacement sampling point and one to two displacement sampling points before and after it. The directional echo amplitude fluctuation range can be the larger of the maximum absolute deviation of the original echo amplitude within the current component displacement neighborhood from the directional echo amplitude reference and the corresponding endpoint static echo fluctuation range.

[0106] The directional difference is calculated based on the change in echo amplitude in two directions of motion under the same component displacement. The calculation formula is as follows: In the formula, This represents the directional difference of the q-th distance angle element at the component displacement x; and These represent the changes in echo amplitude at the component displacement x and the two directions of motion for the qth distance angle element, respectively. The stable term representing the difference can be set as follows: to q represents the distance angle element number; x represents the component displacement. The directional difference is a dimensionless quantity.

[0107] For each component displacement sampling point, the peak positions of the echo ridges corresponding to the two motion directions are read, and these peak positions are taken as the ridge centers for each motion direction. The change in echo amplitude corresponding to the ridge center is used as the local peak value. Continuous units whose echo amplitude changes reach a preset proportion of the local peak value are searched along both the distance and angle dimensions to obtain the ridge width for each motion direction. The preset proportion can be selected within the range of 30% to 70%, with the specific value determined based on the airspace noise level and radar resolution.

[0108] The outer boundaries of the ridge range corresponding to each of the two motion directions are used as the component echo coverage boundaries for that motion direction. Based on the displacement alignment error and node resolution, one to two distance or angle elements are extended outwards to obtain the component echo shielding elements for each of the two motion directions. When the echo ridge line for a single motion direction is temporarily missing, the component echo shielding element corresponding to the other motion direction can be used as a temporary shielding range. When the echo ridge lines for both motion directions are missing, no new component echo shielding elements are added based on the current component displacement.

[0109] For each direction of motion, the detection threshold is determined based on the corresponding direction echo amplitude reference and the direction echo amplitude fluctuation range, and the calculation formula is as follows: In the formula, This represents the detection threshold corresponding to the direction d of motion of the q-th distance angle element at the component displacement x; This indicates the reference value for the corresponding directional echo amplitude; This indicates the range of amplitude fluctuations in the corresponding direction echo; The fluctuation margin coefficient can be selected from 2 to 4; d represents the direction of motion. The fluctuation margin coefficient can be determined based on the frequency of background echo exceeding the detection threshold during multiple air-field acquisitions, with a larger value taken when the air-field echo fluctuation is large.

[0110] The sensing task area is read from the spatial sensing model and transformed into the range-angle space of the corresponding radar sensing node. For each direction of motion, when the corresponding component echo shielding unit does not intersect with the sensing task area, the range boundary and angle boundary of the sensing task area are used as the observation gate parameters for that direction of motion. When there is an intersection, the range-angle units corresponding to the intersection are excluded from the sensing task area, and then the lower range limit, upper range limit, lower angle limit, and upper angle limit of the remaining continuous area are extracted. When the remaining area is divided into multiple discontinuous regions, the observation gate parameters corresponding to each region are saved separately.

[0111] Within the same distance-angle unit, when the directional difference reaches a preset directional difference threshold for two or more consecutive component displacement sampling points, the directional echo amplitude reference, directional echo amplitude fluctuation range, component echo shielding unit, detection threshold, and observation gate parameters corresponding to the two motion directions are combined into a first motion direction parameter group and a second motion direction parameter group, respectively, generating a directional parameter branch. The preset directional difference threshold can be selected within the range of 0.10 to 0.30, or it can be determined based on the 95th percentile of the directional difference obtained from multiple round trip calibrations.

[0112] When the directional difference does not reach the preset directional difference threshold, the bidirectional shared echo amplitude reference is determined based on the original echo amplitude values ​​of the two motion directions, and the bidirectional shared echo amplitude fluctuation range is determined based on the directional echo amplitude fluctuation range of the two motion directions and the difference between the original echo amplitude values. The calculation formula is as follows: In the formula, This indicates the bidirectional shared echo amplitude reference for the q-th distance angle element at the component displacement x; and These represent the original echo amplitude values ​​corresponding to the two directions of motion; This indicates the range of fluctuation in the bidirectional shared echo amplitude; q represents the distance angle element number; x represents the component displacement.

[0113] Based on the bidirectional shared echo amplitude reference and the bidirectional shared echo amplitude fluctuation range, the bidirectional shared detection threshold is determined according to the aforementioned calculation method. Specifically, the bidirectional shared echo amplitude reference and the bidirectional shared echo amplitude fluctuation range are used to replace the directional echo amplitude reference and the directional echo amplitude fluctuation range, respectively. The bidirectional shared component echo shielding unit is the union of the echo shielding units corresponding to the two motion directions, and the bidirectional shared observation gate is the intersection of the observation gates corresponding to the two motion directions. The bidirectional shared echo amplitude reference, the bidirectional shared echo amplitude fluctuation range, the bidirectional shared component echo shielding unit, the bidirectional shared detection threshold, and the bidirectional shared observation gate are combined into a bidirectional shared parameter.

[0114] Finally, a displacement index is established according to the order of component displacement from the first state endpoint to the second state endpoint, and each component displacement sampling point is associated with the background echo parameters, component echo shielding unit, detection threshold, observation gate parameters, and directional parameter branch or bidirectional shared parameters to form a parameter transfer relationship.

[0115] This example establishes echo amplitude benchmarks, echo amplitude fluctuation ranges, component echo shielding units, detection thresholds, and observation gate parameters for each component displacement. This allows radar sensing parameters to be adjusted according to the positional changes of movable spatial components. By setting directional parameter branches or bidirectional shared parameters based on the differences in bidirectional echoes, the directional differences generated by the reciprocating motion of components are preserved, and the amount of processing required for repeatedly storing parameters when the differences are small is reduced. This provides a configuration basis for subsequently excluding component motion echoes and associating them with human target echoes.

[0116] As an example, the step of arranging the background echo parameters, the component echo shielding unit, the detection threshold, the observation gate parameters, and the directional parameter branches or the bidirectional shared parameters according to the component displacement to establish parameter transfer relationships includes: establishing displacement index nodes according to component displacement sampling points, and associating each displacement index node with the corresponding background echo parameters, component echo shielding unit, detection threshold, and observation gate parameters; when a displacement index node is located within a displacement segment where the directional difference of two or more consecutive component displacement sampling points reaches a preset directional difference threshold, associating the displacement index node with the directional parameter branches corresponding to the two motion directions respectively; when a displacement index node is not located within the displacement segment, associating the displacement index node with the bidirectional shared parameters; when the current component displacement is located between two adjacent displacement index nodes, determining the corresponding directional parameter branch or bidirectional shared parameter according to the current motion direction, and interpolating the background echo parameters and the detection threshold according to the displacement ratio of the current component displacement between two adjacent displacement index nodes; interpolating the boundary coordinates of the component echo shielding unit and the observation gate parameters according to the displacement ratio to obtain the current configuration parameters corresponding to the current component displacement.

[0117] It should be noted that the displacement index node refers to the data node set according to the component displacement sampling point and used to associate the corresponding configuration parameters; the displacement ratio refers to the relative position of the current component displacement between two adjacent displacement index nodes; the current configuration parameters refer to the background echo parameters, component echo shielding unit, detection threshold and observation gate parameters obtained based on the current component displacement and the current direction of movement.

[0118] Understandably, displacement index nodes are established according to the displacement sequence from the first state endpoint to the second state endpoint. The setting interval of the displacement index nodes is consistent with the displacement resampling interval of the bidirectional state switching data. For translational components, the interval can be 5mm to 20mm, and for rotational components, it can be 0.5° to 2°. The specific interval is determined based on the total stroke of the component, the radar frame period, and the radar resolution. Each displacement index node records the component displacement value, parameter type identifier, and the corresponding background echo parameters, component echo shielding unit, detection threshold, and observation gate parameters.

[0119] When a displacement index node is located within a displacement segment where the directional difference between two or more consecutive component displacement sampling points reaches a preset directional difference threshold, a first motion direction parameter group and a second motion direction parameter group are stored under that displacement index node, and the corresponding directional parameter branch is pointed to by the parameter type identifier. When a displacement index node is not located within this displacement segment, a set of bidirectional shared parameters is stored under that displacement index node. The preset directional difference threshold uses the value determined above based on multiple round-trip calibration data and will not be set again here.

[0120] During the operation of a movable spatial component, the current direction of motion is determined based on the sign of the displacement changes between two adjacent measurements. When the component displacement increases, the direction parameter branch corresponding to the first direction of motion is invoked; when the component displacement decreases, the direction parameter branch corresponding to the second direction of motion is invoked. If the change in component displacement does not exceed the resolution of the position sensor, the most recently confirmed direction of motion is maintained. When displacement index nodes are associated with bidirectional shared parameters, the current direction of motion is not distinguished.

[0121] When the current component displacement matches a certain displacement index node, the parameters corresponding to that displacement index node are directly read. When the current component displacement is between two adjacent displacement index nodes, the background echo parameters and detection threshold are linearly interpolated based on the displacement ratio between the two adjacent displacement index nodes. For component echo shielding units and observation gate parameters that can establish regional correspondence based on the cross-interval ratio, the same displacement ratio is used to linearly interpolate the distance to the lower boundary, distance to the upper boundary, angle to the lower boundary, and angle to the upper boundary.

[0122] The boundary of the component echo shielding unit obtained by interpolation is taken outward to the boundary of the adjacent distance angle unit, and the boundary of the observation gate is taken inward to the boundary of the adjacent distance angle unit, so as to reduce the situation where the component echo enters the search range of the human target.

[0123] When the number of shielded regions in adjacent displacement index nodes is different, or when a corresponding relationship cannot be established based on spatial overlap, instead of directly interpolating the boundary coordinates of different regions, the region parameters corresponding to the displacement index node that is closer to the current component displacement are selected based on the displacement ratio to avoid interpolation producing non-existent shielded regions or observation regions.

[0124] When a component reverses direction during its movement, if the sign of the component displacement change changes in two consecutive reading cycles, it is confirmed that the current direction of movement has changed, and the parameter branch for the other direction is switched at the current component displacement. If the direction of movement has not yet been confirmed, the parameter branch before the switch continues to be used. After the branch switch is completed, the current configuration parameters are determined only between the adjacent displacement index nodes corresponding to the new direction of movement.

[0125] When the current component displacement exceeds the calibration range but the excess is no more than one displacement index interval, the state endpoint parameter closest to the current component displacement is called. If the excess is greater than one displacement index interval, or if it exceeds the calibration range for three consecutive reading cycles, parameter interpolation is no longer performed on the excess portion, and the corresponding component motion interval is marked as requiring recalibration. Finally, the background echo parameters obtained through interpolation or directly read, the component echo shielding unit, the detection threshold, and the observation gate parameters are combined into the current configuration parameters.

[0126] This example associates radar configuration parameters at different locations with component displacement index nodes, enabling parameter transfer relationships to be queried according to the current component displacement. It calls directional parameter branches or bidirectional shared parameters based on directional differences, preserving the echo differences generated by the component's round-trip motion. By interpolating parameter values ​​and boundary coordinates, it reduces parameter jumps between adjacent displacement index nodes. At the same time, it handles cases of reverse motion and exceeding the calibration range separately, improving the matching degree between the current configuration parameters and the component's position and motion direction during continuous component motion.

[0127] This application also provides a living space sensing configuration device, please refer to... Figure 4 The living space sensing configuration device includes: Model building module 10 is used to build a spatial perception model based on living space data, movable space component data and radar sensing node parameters. The switching acquisition module 20 is used to control the movable spatial components to switch states in opposite directions, synchronously acquiring component displacement and radar echoes to obtain bidirectional state switching data. The unit determination module 30 is used to determine the state-sensitive observation unit, the non-sensitive observation unit, and the affected node based on the spatial perception model and the bidirectional state switching data. The relationship generation module 40 is used to establish a parameter transfer relationship that changes with the displacement of the component based on the state-sensitive observation unit, determine the anchoring node based on the non-sensitive observation unit, establish a track constraint relationship between the anchoring node and the affected node, and generate a state switching configuration scheme. The configuration module 50 is used to update the affected nodes according to the current component displacement and the parameter transfer relationship when the movable space component switches states, and associate the human target echo according to the state switching configuration scheme and the track constraint relationship to obtain the perception configuration result.

[0128] The residential space sensing configuration device provided in this application, employing the residential space sensing configuration method in the above embodiments, can solve the technical problem of distinguishing between the motion echo of movable spatial components and the echo of human targets when the components switch states, and maintaining continuous adaptation between radar sensing parameters and human target trajectories. Compared with the prior art, the beneficial effects of the residential space sensing configuration device provided in this application are the same as those of the residential space sensing configuration method provided in the above embodiments, and other technical features in the residential space sensing configuration device are the same as those disclosed in the above method embodiments, and will not be repeated here.

[0129] This application provides a residential space sensing and configuration device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the residential space sensing and configuration method in the first embodiment described above.

[0130] The following is for reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing the residential space sensing configuration device in the embodiments of this application. The residential space sensing configuration device in the embodiments of this application may be a dedicated controller, edge computing device, industrial computer, or server communicating with radar sensing nodes, all installed within the residential space. Figure 5 The device structure shown is only one implementation method, and each hardware component can be configured according to the number of radar sensing nodes, the component driving method, and the amount of data processing.

[0131] like Figure 5 As shown, the living space sensing and configuration device may include a processing unit 1001, a ROM 1002, a storage unit 1003, a RAM 1004, a bus 1005, an I / O interface 1006, an input device 1007, an output device 1008, and a communication device 1009. The processing unit 1001 executes computer programs stored in the ROM 1002, storage unit 1003, or RAM 1004; the I / O interface 1006 connects to component displacement sensors, component driving devices, and a local operation terminal; the communication device 1009 receives radar echo data output from each radar sensing node and sends state switching commands to the driving devices of movable space components; the output device 1008 can output human target tracks, node configuration status, and sensing and configuration results. The processing unit 1001, ROM 1002, storage unit 1003, RAM 1004, I / O interface 1006, and communication device 1009 exchange data via the bus 1005.

[0132] The residential space sensing configuration device provided in this application, employing the residential space sensing configuration method in the above embodiments, can solve the technical problem of distinguishing between the motion echo of movable spatial components and the echo of human targets when the components switch states, and maintaining continuous adaptation between radar sensing parameters and human target trajectories. Compared with the prior art, the beneficial effects of the residential space sensing configuration device provided in this application are the same as those of the residential space sensing configuration method provided in the above embodiments, and other technical features in this residential space sensing configuration device are the same as those disclosed in the above method embodiments, and will not be repeated here.

[0133] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the living space sensing configuration method in the above embodiments.

[0134] The computer-readable storage medium provided in this application can be a tangible medium capable of storing computer programs, such as a hard disk, random access memory, read-only memory, flash memory, optical storage device, or magnetic storage device. When the computer program is executed by a processor, it implements the above-described living space sensing configuration method.

[0135] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described residential space sensing configuration method. This solves the technical problem of distinguishing between the motion echo of a movable spatial component and the echo of a human target when the component switches states, while maintaining continuous adaptation between radar sensing parameters and the trajectory of the human target. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the residential space sensing configuration method provided in the above embodiments, and will not be elaborated upon here.

[0136] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for sensing and configuring living space, characterized in that, The method includes: A spatial perception model is established based on living space data, movable space component data, and radar sensing node parameters. Control the movable spatial components to switch states in opposite directions, and simultaneously collect component displacement and radar echo to obtain bidirectional state switching data; Based on the spatial perception model and the bidirectional state switching data, determine the state-sensitive observation unit, the non-sensitive observation unit, and the affected nodes; Based on the state-sensitive observation unit, a parameter transfer relationship is established that changes with the displacement of the component. Based on the non-sensitive observation unit, the anchoring node is determined, a track constraint relationship is established between the anchoring node and the affected node, and a state switching configuration scheme is generated. When the movable spatial component switches states, the affected node is updated according to the current component displacement and the parameter transfer relationship, and the human target echo is associated according to the state switching configuration scheme and the trajectory constraint relationship to obtain the perception configuration result.

2. The method as described in claim 1, characterized in that, The step of determining the state-sensitive observation unit, the non-sensitive observation unit, and the affected node based on the spatial perception model and the bidirectional state switching data includes: The range of the range angle unit corresponding to the radar sensing node is determined according to the spatial sensing model, and the component displacement sequence, radar echo sequence and endpoint static echo in two motion directions are extracted from the bidirectional state switching data. The radar echo sequences in the two directions of motion are converted into range-angle echo maps, and the range-angle echo maps are arranged according to the displacement of the same components. The change in echo amplitude of each distance angle unit in the two motion directions is calculated based on the static echo at the endpoint. The echo ridges corresponding to the two motion directions are extracted, and the bidirectional response consistency is determined based on the consistency between the change in echo amplitude in the two motion directions and the consistency between the echo ridges. Based on the bidirectional response consistency, state-sensitive and non-sensitive observation units are determined, and radar sensing nodes containing the state-sensitive observation units are identified as affected nodes.

3. The method as described in claim 2, characterized in that, The step of establishing the parameter transfer relationship as the component displacement changes based on the state-sensitive observation unit includes: For each distance angle unit in the state-sensitive observation unit, the background echo parameters are determined based on the original echo amplitude values ​​corresponding to the two motion directions under the same component displacement, and the directional difference is determined based on the change in the bidirectional echo amplitude under the same component displacement. The component echo shielding unit is generated based on the ridge center and ridge width of the bidirectional echo ridge. The sensing task area is determined from the spatial sensing model, the detection threshold is determined based on the background echo parameters and the endpoint static echo, and the observation gate parameters are determined based on the intersection position of the component echo shielding unit and the sensing task area. Generate directional parameter branches or bidirectional shared parameters based on the directional difference; The background echo parameters, the component echo shielding unit, the detection threshold, the observation gate parameters, and the directional parameter branches or the bidirectional shared parameters are arranged according to the component displacement to establish parameter transfer relationships.

4. The method as described in claim 1, characterized in that, The steps for establishing a spatial perception model based on living space data, movable space component data, and radar sensing node parameters include: Spatial structure data, sensing task area, and node installation surface are extracted from residential space data. The spatial structure data, sensing task area, and node installation surface are then mapped to a spatial coordinate system to obtain spatial mapping data. Extract the first state endpoint, second state endpoint, motion trajectory, contour dimensions, material parameters, and component displacement from the data of movable spatial components; The node installation position is determined based on the node mounting surface and the radar sensing node parameters, and the node installation position is associated with the radar sensing node parameters to obtain the node parameter association relationship. A pose correspondence is established based on the first state endpoint, the second state endpoint, the motion trajectory, and the component displacement. A spatial perception model is then established based on the spatial mapping data, the contour dimensions, the material parameters, the node parameter associations, and the pose correspondence.

5. The method as described in claim 1, characterized in that, The steps of controlling the movable spatial component to switch states in opposite directions, synchronously acquiring component displacement and radar echoes, and obtaining bidirectional state switching data include: During the airfield calibration period, the radar sensing nodes are controlled to transmit detection radio waves and receive radar echoes. The static echo of the first endpoint is collected at the first state endpoint, and the movable spatial component is controlled to move from the first state endpoint to the second state endpoint. The component displacement and radar echo are collected simultaneously to form the first component displacement sequence and the first radar echo sequence. At the second state endpoint, static echoes of the second endpoint are collected, and the movable spatial component is controlled to move in the opposite direction to the first state endpoint. The component displacement and radar echoes are collected simultaneously to form a second component displacement sequence and a second radar echo sequence. The displacement sequence of the first component and the first radar echo sequence, and the displacement sequence of the second component and the second radar echo sequence are time-aligned and resampled. The resampling results, the static echo of the first endpoint and the static echo of the second endpoint are combined into bidirectional state switching data.

6. The method as described in claim 1, characterized in that, The steps of determining the anchoring node based on the non-sensitive observation unit, establishing the track constraint relationship between the anchoring node and the affected node, and generating a state switching configuration scheme include: From radar sensing nodes that do not belong to the affected nodes, determine the non-sensitive coverage ratio and the common coverage area with the affected nodes based on the non-sensitive observation units corresponding to each radar sensing node, and determine the anchoring node based on the non-sensitive coverage ratio and the common coverage area. Establish the coordinate transformation relationship between the node coordinate system of the anchored node and the node coordinate system of the affected node; The human target trajectory output by the anchor node is obtained, a target prediction region is generated based on the human target trajectory, the observation gate parameters and component echo shielding units are extracted from the parameter transfer relationship, and the constrained observation gate is determined accordingly. A trajectory constraint relationship is established based on the coordinate transformation relationship, the human target trajectory, and the constrained observation gate. A state switching configuration scheme is generated based on the human target trajectory calling order of the anchor node and the parameter update order of the affected node.

7. The method according to any one of claims 1 to 6, characterized in that, The steps of updating the affected nodes based on the current component displacement and the parameter transfer relationship when the movable spatial component switches states, and associating human target echoes with the state switching configuration scheme and the trajectory constraint relationship to obtain the perception configuration result include: Obtain the current displacement, current direction of motion, and current radar echo of each radar sensing node of the movable spatial component; The current configuration parameters are determined from the parameter transfer relationship based on the current component displacement and the current direction of motion, and the affected nodes are updated according to the state switching configuration scheme; The corresponding constrained observation gate is determined from the trajectory constraint relationship based on the current component displacement, and the constrained observation gate is adjusted according to the human target trajectory output by the anchor node. The current radar echo of the affected node is processed according to the current configuration parameters. The human target echo that falls into the adjusted constrained observation gate is determined from the processed current radar echo. The human target echo is associated with the human target track. After the movable spatial component reaches the state endpoint, the static echo of the current endpoint is collected. The parameter transfer relationship is updated based on the static echo of the current endpoint and the static echo of the endpoint in the bidirectional state switching data to obtain the perception configuration result.

8. A living space sensing configuration device, characterized in that, The device applies the living space sensing configuration method as described in any one of claims 1 to 7, and the device comprises: The model building module is used to build a spatial perception model based on living space data, movable space component data and radar sensing node parameters. The switching acquisition module is used to control the movable spatial components to switch states in opposite directions, synchronously acquiring component displacement and radar echoes to obtain bidirectional state switching data; The unit determination module is used to determine the state-sensitive observation units, non-sensitive observation units, and affected nodes based on the spatial perception model and the bidirectional state switching data. The relationship generation module is used to establish parameter transfer relationships that change with the displacement of the component based on the state-sensitive observation unit, determine the anchoring node based on the non-sensitive observation unit, establish the track constraint relationship between the anchoring node and the affected node, and generate a state switching configuration scheme. The operation configuration module is used to update the affected nodes according to the current component displacement and the parameter transfer relationship when the movable space component switches states, and associate the human target echo according to the state switching configuration scheme and the track constraint relationship to obtain the perception configuration result.

9. A residential space sensing configuration device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the living space sensing configuration method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the living space sensing configuration method as described in any one of claims 1 to 7.

Images