Object searching and lighting system and method based on Bluetooth AOA positioning

Abstract

The invention discloses an object searching and lighting system and method based on Bluetooth AOA positioning, and relates to the technical field of Bluetooth positioning, the system comprises a Bluetooth tag module, the Bluetooth tag module is installed on a target object, a low-power-consumption Bluetooth chip is arranged in the Bluetooth tag module, and the Bluetooth tag module is used for periodically broadcasting unique ID information; and a positioning base station array module, wherein the positioning base station array module is deployed in an indoor environment. According to the system, Bluetooth AOA positioning and multi-mode feedback are integrated, accurate object searching guiding is achieved, the coordinates of an object are calculated through AOA positioning, six-color light and a buzzer are combined, and the feedback mode is dynamically adjusted according to the distance; a sleep wake-up mechanism reduces power consumption and improves equipment endurance; a multi-base station synchronization and Kalman filtering algorithm supports multi-label concurrent positioning, and the anti-interference capability is enhanced; direction deviation prompting and shielding response are realized, object searching experience in a complex environment is optimized, and the low-power-consumption and high-efficiency intelligent object searching system is constructed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of Bluetooth positioning technology, in particular to a Bluetooth AOA positioning-based lost item light system and method. BACKGROUND

[0002] The fusion development of indoor positioning technology and intelligent lost item system has become a research hotspot in the field of Internet of Things. The Bluetooth AOA (Angle of Arrival) positioning technology is based on the principle of signal arrival angle measurement, and calculates the target direction through the phase difference of multi-antenna array received signals. It can achieve sub-meter positioning accuracy under the support of BLE5.1 and above protocols, and provides a technical basis for indoor item tracking. At the same time, the lost item light system as an intuitive visual guidance means, assists users in locating target items through LED light flickering and other ways, and has wide application needs in smart home, warehouse management and other scenarios. Limited application scenarios: AOA technology is mainly used for accurate positioning, but it does not combine other functions to improve user experience. Most AOA technology application scenarios are limited to positioning, and lack intelligent feedback mechanisms, such as being unable to guide users to find items through simple light or prompts. High complexity: AOA technology usually requires complex hardware support, such as multiple receivers and high computing power, resulting in high device cost and the need for complex configuration and debugging. Poor interactivity: With only AOA technology, users can only determine the position of the object by looking at the values on the device or screen, which lacks intuitive and immediate feedback, which is not ideal for some scenarios that require quick positioning. Therefore, we propose a Bluetooth AOA positioning-based lost item light system and method to solve the technical problems existing in the above. SUMMARY

[0003] The purpose of the present application is to provide a Bluetooth AOA positioning-based lost item light system and method to solve the problems raised in the background technology.

[0004] To achieve the above purpose, the technical solution adopted by the present application is as follows: the system comprises: A Bluetooth tag module, the Bluetooth tag module is installed on a target item, the Bluetooth tag module is built-in with a low-power Bluetooth chip, and the Bluetooth tag module is used for periodic broadcast of unique ID information; A positioning base station array module, the positioning base station array module is deployed in an indoor environment and comprises a plurality of base stations with multiple antennas, the positioning base station array module is used for receiving signals of the Bluetooth tag and extracting incident angles through phase differences of different antennas; A positioning calculation and signal processing module, the positioning calculation and signal processing module is used for fusing multi-point incident angle data and calculating two-dimensional coordinates of the target tag. The prompt light control and multi-modal feedback module is used for controlling the six-color LED light on the Bluetooth tag to flash according to target coordinates and providing feedback in combination with a bee sound and vibration.

[0005] Compared with the prior art, the present application has obvious advantages and beneficial effects, and specifically, from the above technical solution, it mainly has the following advantages: The system fuses Bluetooth AOA positioning and multi-modal feedback to realize precise object searching guidance: the object coordinates are calculated through AOA positioning, the feedback mode is dynamically adjusted according to the distance in combination with six-color light and a bee sound, the sleep-wake mechanism reduces power consumption and improves equipment endurance, multi-base station synchronization and Kalman filtering algorithm support multi-tag concurrent positioning and enhance anti-interference capability, the direction deviation prompt and the shielding response optimize the object searching experience in complex environments, and a low-power-consumption and high-efficiency intelligent object searching system is constructed.

[0006] To make the structural features and effects of the present application clearer, the present application will be described in detail below in combination with the drawings and specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 The figure is a system block diagram of the object searching light system of the embodiment of the present application. DETAILED DESCRIPTION

[0008] To make the purpose, technical solution and advantages of the present application clearer, the present application will be described in further detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0009] It should be noted that when an element is referred to as being "fixed" to another element, it can be directly on the other element or there can be a middle element. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or there can be a middle element. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0010] As shown in Figure 1 The embodiment of the present application provides an object searching light system and method based on Bluetooth AOA positioning, and the system comprises: A Bluetooth tag module is installed on a target object, a low-power-consumption Bluetooth chip is built in the Bluetooth tag module, and the Bluetooth tag module is used for periodically broadcasting unique ID information. A positioning base station array module is deployed in an indoor environment and includes a plurality of base stations with multiple antennas, which is used to receive signals of the Bluetooth tag and extract the incident angle through the phase difference of different antennas; A positioning calculation and signal processing module is used to fuse multi-point incident angle data and calculate the two-dimensional coordinates of the target tag. A prompt light control and multi-modal feedback module is used to control the six-color LED light on the Bluetooth tag to flash according to the target coordinates and provide feedback in combination with a buzzer and vibration. The core architecture of the Bluetooth AOA positioning-based light-on system is built on the basis of wireless communication theory and signal processing technology: the Bluetooth tag module uses the BLE5.1 protocol stack and periodically sends data packets containing unique IDs using the broadcast mechanism of the low-power Bluetooth chip (complying with the ISO / IEC18000-63 standard). This design realizes multi-tag concurrent identification based on the time division multiplexing (TDM) principle. The built-in six-color LED light group (red / green / blue / yellow / purple / white) and buzzer are driven by the MCU, and the brightness and frequency are controlled through the PWM pulse width modulation technology. The sleep-wake mechanism is optimized for power balance theory (sleep power consumption ≤10 μA) to improve the battery life. The positioning base station array module is deployed in accordance with the signal angle of arrival (AOA) angle measurement principle. The multi-antenna array (using a uniform circular array UCA structure with an antenna spacing of λ / 2, λ being the 2.4 GHz carrier wavelength) of each base station calculates the signal incident angle through the phase interference method. Specifically, based on the Fresnel diffraction theory, the phase difference (accuracy ≤0.5°) of the multi-antenna received signal is used to construct the azimuth vector, and the IEEE1588 clock synchronization protocol is used to achieve nanosecond-level time alignment between base stations, ensuring the spatiotemporal consistency of multi-point angle measurement data. The positioning calculation and signal processing module integrates the triangulation positioning algorithm and the Kalman filter model. The former is based on the geometric positioning principle and solves the two-dimensional coordinates by the intersection of the incident angles of at least three base stations. The latter recursively optimally estimates the dynamic noise according to the state space theory (positioning error ≤ 30 cm), and introduces the indoor map raster model (RasterModel) to exclude false positioning caused by obstacles. The concurrent processing capability (support ≥ 10 tags for synchronous positioning) of this module relies on the distributed computing theory to achieve load balancing. The design of the prompt light control and multi-modal feedback module conforms to the human-machine engineering interaction principle. The six-color light dynamically switches according to the distance-color mapping rule (D > 5m displays blue, 2m < D ≤ 5m displays yellow, and D ≤ 2m displays red). The color coding mechanism refers to the CIE1931 color space theory to improve visual recognition. The buzzer feedback (intermittent sound at a distance of 1Hz, continuous sound at a close distance) optimizes auditory guidance based on the Doppler effect principle. The purple flashing mode (5Hz frequency) when blocked is triggered based on the signal strength RSSI and the obstacle attenuation model (such as the Log-NormalShadowing model). The whole feedback mechanism realizes instruction transmission (real-time response within 100m) through the GATT service layer of Bluetooth 5.0 protocol, forming a complete theoretical support system from the hardware bottom to the application logic.

[0011] Further, the positioning base station array module supports BLE5.1 and above protocols, and realizes multi-point joint angle measurement through inter-base station synchronization; The technical implementation of the positioning base station array module deeply integrates wireless communication protocol evolution and multi-sensor cooperative positioning theory: the BLE5.1 protocol supported by it introduces channel state information (CSI) collection capability at the physical layer, and through the spatial multiplexing principle of multiple-input multiple-output (MIMO) technology, it realizes accurate measurement of the direction of arrival (DoA) of the Bluetooth signal. Each base station is configured with 4-8 receiving antennas (using uniform circular array UCA or linear array ULA structure) to build a phase interferometer model according to the Huygens-Fresnel principle, and the antenna spacing strictly follows the best measurement criterion of λ / 2 (λ is the wavelength of the 2.4GHz carrier) to eliminate 180° phase ambiguity. The IEEE1588v2 precision clock synchronization protocol is used for inter-base station synchronization, and sub-microsecond time alignment is realized through PTP (Precision Time Protocol) message interaction. This process is based on the clock offset estimation model in the network time synchronization theory, combined with the Kalman filtering algorithm to compensate for the crystal oscillator drift (typical value ±20ppm) in real time, ensuring that the distributed base station array forms a unified space-time reference system. Multi-point joint angle measurement follows the principle of spatial triangulation. After the original phase data (ADC sampling rate ≥20MSPS) is preprocessed by the edge computing node, the distributed Kalman filtering (DKF) algorithm is used for data fusion in the central server. This algorithm realizes the optimal estimation of measurement noise based on the covariance intersection theory, and introduces particle filtering (PF) to handle non-Gaussian noise scenarios, significantly improving the angle measurement accuracy in complex environments (up to 0.5° angle resolution). The system uses spatial smoothing algorithm (SSM) combined with MUSIC (multiple signal classification) algorithm to suppress multipath effect. The former destroys the coherence of the signal through subarray division technology, and the latter orthogonalizes the signal space and noise space based on the eigenvalue decomposition theory, realizing the effective separation of multipath components. In terms of concurrent processing capability, the system adopts TDOA (time difference of arrival) and AOA joint positioning architecture, and realizes conflict avoidance of ≥10 tags through time slot allocation algorithm (such as improved ALOHA protocol). This algorithm optimizes channel utilization based on queuing theory model, ensuring that the positioning refresh rate reaches more than 2Hz, meeting the needs of dynamic target tracking. The whole technical scheme forms a complete theoretical closed loop at the physical layer, data link layer and application layer, providing a solid support for high-precision indoor positioning.

[0012] Further, the prompt light control and multi-modal feedback module switches the light color according to the relative distance between the target and the user: red / green color is displayed when the distance is short, and blue / yellow color is displayed when the distance is long. The color switching mechanism of the prompt light control and multi-modal feedback module is based on the cross-application of visual perception theory and signal processing technology: the color coding scheme follows the CIE1931 color space theory. Red color (wavelength 620-750nm) and green color (500-565nm) have higher visual contrast in short distance (≤2m) due to the long-wave characteristics, which meets the resolution characteristics of human vision to near-field color in Weber's law; blue color (450-495nm) and yellow color (565-590nm) use the long-distance penetration of short-wave light (5m away), and optimize the long-distance recognition through the brightness channel (L*) in the CIELAB color model. The distance judgment is based on the Euclidean distance (D=√[(x2-x1)²+(y2-y1)²]) output by the positioning calculation module, and the RSSI signal strength is corrected by combining the logarithmic distance path loss model (PL(dB)=PL(d0)+10nlog10(d / d0)), where the path loss index n is 2.5-3.5 (typical value in indoor environment), to ensure the dynamic switching accuracy of the distance threshold (2m / 5m). The light control uses PWM pulse width modulation technology (frequency 1-1000Hz), which adjusts the LED brightness (0-100% linear response) by changing the duty cycle. When D≤2m, the red LED duty cycle is increased to more than 80% to form a high-brightness prompt, which meets the visual enhancement requirement when the target is close according to Fitts' law. The multi-modal feedback integrates the auditory guidance theory, and the frequency (1Hz→continuous sound) change of the buzzer follows the Mel frequency scale in psychoacoustics, making the frequency change linearly related to human ear perception. The color-distance mapping table is established according to the ISO9241-309 human-computer interaction standard, and the threshold interval is optimized through 200 user test data to ensure that more than 90% of users can identify the distance level within 3 seconds. The real-time control logic of this module is based on the finite state machine (FSM) model, which triggers state transition (such as blue→yellow) when the distance crosses the threshold, and eliminates frequent switching caused by positioning jitter through exponential smoothing filtering (α=0.3). The hardware implementation uses ARMCortex-M3 core MCU to control TLC5940 constant current driving chip through SPI interface, realizing independent color adjustment and synchronous control of six-color LED. From color theory, signal processing to human engineering, a multi-level theoretical support system is formed.

[0013] Further, the buzzer feedback dynamically adjusts the frequency according to the distance: low-frequency intermittent sound at long distance, and high-frequency continuous sound at short distance; The buzzer feedback mechanism is based on the Mel Scale in psychoacoustics and the theory of auditory perception design: low-frequency intermittent mode (1Hz, pulse width 200ms) at a long distance (>5m) uses the sensitive characteristics of human ears to medium-frequency signals (isoperimetric curve) in 2000-5000Hz, and generates low-frequency pulses with high audibility through AM modulation; at a short distance (≤2m), it switches to a high-frequency continuous tone (2000Hz) and follows the feedback reinforcement principle in Fitts's law when the target is close, and the frequency increase amplitude conforms to Stevens's power law (loudness is proportional to the 0.67 power of physical intensity). The distance-frequency mapping function uses a piecewise linear model f(D)=1+1999・exp(-D / 2), combined with Kalman filter to predict the distance change trend, and through the lead phase compensation (φ=arctan(2πfτ), τ=50ms) to eliminate the auditory delay. The rhythm design of intermittent sound refers to the attention blink effect in cognitive psychology, and the 1-second interval ensures the continuous perception of users in multitasking scenarios, while the frequency stability of continuous tone is controlled within ±0.5% (based on 32.768kHz crystal oscillator), which meets the specification requirements of pure tone signals in ANSI S3.6-2004 acoustic standard. The scheme outputs 12-bit PWM waveform through the DAC of STM32, and the waveform is smoothed through a second-order low-pass filter (cutoff frequency 3kHz), and finally drives an 8Ω / 0.5W loudspeaker to realize dynamic adjustment of sound pressure level 85-95dB, and the whole system forms a complete theoretical closed loop at the level of signal generation, transmission and perception.

[0014] Further, the Bluetooth tag module adopts a sleep-wake mechanism and only broadcasts signals when receiving a positioning activation instruction; The sleep-wake mechanism of the Bluetooth tag module is rooted in energy efficiency optimization theory and low-power design principles for wireless communication. It employs a dual-state model of deep sleep (power consumption ≤ 10μA) and active broadcasting, adhering to the law of energy conservation to minimize unnecessary power consumption. In sleep mode, the RF front-end, CPU core, and peripheral modules are shut down, with only the Low Power Timer (LPTIM) and the Wake-up Interrupt Controller (EXTI) remaining active. This design, based on duty cycle optimization theory, reduces average power consumption to the microampere level. The wake-up mechanism leverages the energy-saving characteristics of the BLE protocol stack, receiving activation commands via Radio Frequency Wake-up technology. The built-in wake-up circuit uses a superheterodyne receiver architecture, achieving a wake-up sensitivity of -95dBm in the 2.4GHz band (following Shannon's theorem C=B・log2(1+S / N)), ensuring reliable activation within a 100m range. Instruction detection employs a Differential Coherent Demodulation algorithm, combined with preamble synchronization technology to compress the wake-up time to ≤10ms, meeting real-time requirements. The sleep cycle is controlled by an Adaptive Timer, dynamically adjusting the broadcast interval (100ms-500ms) based on historical object-finding frequencies. This mechanism optimizes energy consumption-response balance using the Q-Learning algorithm in reinforcement learning. When a user app connection request is detected, it quickly exits sleep mode via interrupt-driven mode, avoiding power waste caused by polling detection. Hardware implementation utilizes a low-power MCU with a Cortex-M0+ core. Its power management unit (PMU) supports five low-power modes and, in conjunction with a DC-DC buck converter, controls the sleep current to the nanoamp level. During wake-up, a prefetch buffer accelerates instruction execution. The entire mechanism forms a theoretical closed loop of energy consumption optimization at the physical, protocol, and algorithm layers, complying with the ISO14543-3 standard for low-power smart home devices.

[0015] Furthermore, the object-finding method includes: Users can initiate a lost item request by selecting the target tag ID through the mobile app; The central controller activates the positioning base station array, receives tag signals, and measures the angle of incidence. The positioning engine integrates multi-point angle data to calculate the real-time coordinates of the tag; Send a light control command to the target tag to activate six-color flashing lights accompanied by a beep / vibration; The lost item search method is built on the theory of wireless sensor network (WSN) cooperative positioning and the architecture of distributed control system: the request for searching the lost item initiated by the user through the mobile APP follows the target-oriented design principle in human-computer interaction, and the request data is sent to the central controller through the MQTT protocol (QoS2 level ensures reliable transmission), which conforms to the application layer and transport layer specifications in the OSI seven-layer model. The central controller activates the base station array using the time division multiple access (TDMA) scheduling algorithm (time slot allocation follows the Max-Min fairness principle), ensuring the optimization of channel resources when multiple base stations measure concurrently. The incident angle measurement is based on the principle of angle of arrival (AOA), and the multi-antenna array of each base station calculates the angle (accuracy ≤0.5°) through the phase interference method, which follows the Huygens-Fresnel principle and realizes super-resolution direction finding combined with the MUSIC algorithm. When the positioning engine fuses multi-point angle data, it uses the weighted least squares method (WLS) to construct the target function min∑w_i・||A_i・x-b_i||², where the weight w_i is obtained by converting the signal strength RSSI through the logarithmic distance path loss model, and the Kalman filtering algorithm is used to estimate the state of the dynamic target (state transition matrix A=I, measurement matrix H=Jocobian matrix), ensuring that the coordinate update rate reaches more than 2Hz to meet the real-time tracking requirements. The light control command is sent through the GATT protocol of Bluetooth 5.0, and AES-128 encryption is used to ensure data security. The command contains color coding (based on CIE1931 color space), brightness value (PWM duty cycle 0-100%) and buzzer mode (frequency 1-2000Hz), and the dynamic adjustment of these parameters conforms to the perception feedback theory in human engineering, which improves the efficiency of searching the lost item through multi-modal stimulation. The whole process forms a closed-loop control system from request initiation, signal collection, data processing to feedback execution, and its stability is verified by Lyapunov's second method, ensuring that the system can still converge to the desired state under disturbance; When the user loses the key in a 10m×10m office environment, the key is installed with a Bluetooth tag (ID: KEY-001): the user opens the APP and clicks "Find Key", the APP sends a request to the central controller deployed in the office through Wi-Fi. The controller activates the 4 positioning base stations distributed in the four corners (forming a spatial tetrahedron structure), each base station receives the signal broadcast by the tag with a 10ms time slot interval (frequency 2.4GHz, transmission power 0dBm). Base station A's 8 antenna array detects the phase difference of the signal, calculates the incident angle as 30° (east by north); base station B measures the incident angle as 120° (south by east); base station C does not receive valid signals due to obstruction; base station D measures the incident angle as 210° (west by south). The positioning engine substitutes the three sets of angle data into the triangular positioning algorithm, excludes the obstacle area combined with the office electronic map (grid accuracy 0.5m), and predicts the tag position as (6.2m, 4.8m) through Kalman filtering, with the 95% confidence interval of the error ellipse as ±0.3m. The controller sends instructions to the tag: about 4.2m away from the user's current position (2.1m, 3.5m), triggering a yellow light (duty cycle 70%) and a 2Hz buzzer; When the user moves to (5.8m, 4.5m) according to the light and sound guidance, the system detects that the distance is shortened to 0.8m, and automatically switches to a red highlight (duty cycle 90%) and a continuous buzzer (1500Hz), and finally the user finds the key in the office desk drawer. The whole process takes 3.7 seconds, the positioning trajectory conforms to the principle of least action, and the system energy consumption optimization model shows that the tag can last for 180 days in sleep state and maintain 48 hours of search function in wake-up state.

[0016] Further, in the step of measuring the incident angle, the base station calculates the signal arrival angle through multi-antenna phase difference, and realizes multi-point joint angle measurement through clock synchronization. The incident angle measurement step is constructed based on array signal processing theory and space-time synchronization technology: the multi-antenna array (using uniform circular array UCA structure, antenna spacing λ / 2) of the base station measures the signal arrival angle through the principle of phase interferometer, and its mathematical model is based on the Huygens-Fresnel principle, the received signal can be represented as s(t)=A・exp(j(ωt+φ+2πd・sinθ / λ)), where d is the antenna spacing, θ is the incident angle; The phase difference calculation adopts a cross-correlation algorithm R(τ)=E[s1(t)s2*(t+τ)], is solved in the frequency domain through FFT transformation (resolution ≤0.5°), and the signal subspace is orthogonalized with the noise subspace by combining the MUSIC algorithm, so that super-resolution direction finding is realized. Clock synchronization adopts the IEEE1588v2 precision clock protocol, the time stamp exchange (t1-t4 four-tuple) of the synchronization message (Sync) and the follow-up message (Follow_Up) is used, the clock offset δ=(t2-t1+t3-t4) / 2 and the link delay τ=(t2-t1-t3+t4) / 2 are estimated based on the least square method, and the synchronization accuracy reaches the sub-microsecond level (σ≤100ns); When multi-point joint angle measurement is performed, the local measurement angle data (with UTC time stamp) of each base station is preprocessed by an edge computing node, a Kalman information filter (KIF) is used for data fusion in a central server, a state space model thereof is x_k=Ax_{k-1}+w_k, z_k=Hx_k+v_k, wherein the state vector x includes the target position and velocity, and the measurement matrix H is determined by the geometric positions of the base stations. The system solves the optimal angle combination by maximum likelihood estimation (MLE), introduces a consensus factor (ConsensusFactor) to process asynchronous measurement data, and ensures that the angle fusion error is ≤0.3° when the clock deviation of the base station is ≤500ns; The complete technical solution forms a complete theoretical closed loop in the aspects of signal processing, time synchronization and data fusion, and meets the requirements of the ISO / IEC24730-61 real-time positioning system standard.

[0017] Further, in the step of calculating the coordinates, a Kalman filtering algorithm is used to fuse data, and multi-tag concurrent positioning is realized. The coordinate calculation step is constructed based on the optimal estimation theory of Kalman filtering and multi-target tracking technology: for the multi-tag concurrent positioning scene, the system adopts an extended Kalman filter (EKF) group parallel processing mechanism, and each tag corresponds to an independent state space model; The state vector x=[x,y,v_x,v_y]^T includes two-dimensional position and velocity components, in the state transition equation x_k=Ax_{k-1}+w_k, the state transition matrix A=diag([1,1,Δt,Δt]) (Δt=0.5s is the sampling period), and the covariance matrix Q=σ²・∫0^ΔtΦ(t)Φ^T(t)dt of the process noise w_k~N(0,Q) (Φ is the state transition function); The measurement vector z=[θ1,θ2,...,θ_n]^T includes the incident angles measured by n base stations, in the measurement equation z_k=h(x_k)+v_k, h(x)=[arctan2(y-y_i,x-x_i)] (i=1~n) is a nonlinear function, and linearization processing is performed through the Jacobian matrix H=∂h / ∂x; The filter is iterated by a prediction-update loop: prediction step x̂_k⁻=Ax̂_{k-1}⁺, P_k⁻=AP_{k-1}⁺A^T+Q; update step K_k=P_k⁻H^T(HP_k⁻H^T+R)⁻¹, x̂_k⁺=x̂_k⁻+K_k(z_k-h(x̂_k⁻)), P_k⁺=(I-K_kH)P_k⁻, where R is the measurement noise covariance (dynamic adjustment based on AOA measurement accuracy); The multi-tag processing adopts a joint probability data association (JPDA) algorithm to solve the ambiguity of data association by calculating the association probability β_j(k) (j is the measurement index), and the weight is determined by the Mahalanobis distance d²=(z_k-h(x̂_k⁻))^T(HP_k⁻H^T+R)⁻¹(z_k-h(x̂_k⁻)). The system realizes concurrent tag differentiation through a tag ID matching mechanism (UUID field in the BLE broadcast packet) and solves the many-to-many association problem by using the Hungarian algorithm; To improve real-time performance, a distributed Kalman filter architecture is adopted, the edge node pre-computes local state estimation, and the central server fuses the global state through an information filter. The communication overhead optimization follows Shannon's information theory to ensure that the bandwidth utilization rate is ≥80%. The complete algorithm forms a complete theoretical system in terms of nonlinear estimation, data association, and distributed computing, and meets the requirements of the IEEE802.15.4z real-time positioning system standard.

[0018] Further, in the light control step, the brightness is dynamically adjusted according to the relative position of the tag and the user: the closer the distance, the higher the brightness, and the flashing frequency is switched when blocked; Further, the multi-modal feedback includes: when the user deviates from the target direction, the buzzer emits intermittent sound; and when approaching the target, the sound is switched to continuous sound; The multi-modal feedback mechanism is constructed based on the theory of auditory spatial perception and the attention guiding principle in cognitive psychology: when the angle Δθ between the line connecting the user's position and the target tag (θ_target) and the user's moving direction (θ_move) exceeds the 30° threshold, the buzzer starts the intermittent sound mode, and the frequency modulation follows the Mel Scale in psychoacoustics. Low-frequency pulses (500Hz) are emitted at an interval of 1Hz, and the pulse width (200ms) is designed to refer to the time integration characteristics of the human auditory system (Patterson integration time is about 200ms). The rhythm pattern of the intermittent sound adopts an asymmetric design (rising edge 50ms, falling edge 150ms), which conforms to the Onset Enhancement in auditory perception, and improves the perceptual significance of direction deviation; When Δθ≤15°, switch to continuous sound, frequency is raised to 2000Hz (located in the most sensitive 2-5kHz frequency band of human ear), the frequency variation amplitude (1500Hz) meets the requirement of Weber law (ΔI / I≈0.1) for perceptible difference; The direction judgment is based on the user motion trajectory estimated by the extended Kalman filter, the velocity vector v t =[Δx / Δt,Δy / Δt] of the three consecutive position points is calculated, the direction cosine matrix is constructed by combining the current coordinates of the target label, and the Mahalanobis distance is used to measure the direction similarity d_M²=(v t -μ)^TΣ⁻¹(v t -μ) (μ is the target direction vector, and Σ is the covariance matrix); In order to avoid the interference of motion noise, the system introduces a hidden Markov model (HMM) for state prediction, converts the direction judgment into maximum a posteriori probability estimation argmaxP(S|O) (S is the state space, and O is the observation sequence), and solves the optimal path through the Viterbi algorithm. The feedback intensity dynamic adjustment follows Stevens power law (ψ=k・φ^0.67), when the user approaches the target (distance threshold D=2m), the sound pressure level is linearly raised from 75dB to 85dB (in line with OSHA occupational safety standards), and at the same time, the LED brightness is increased from 50% to 90% through PWM duty cycle, forming a synergistic reinforcement feedback of vision-audition; The whole mechanism forms a complete theoretical closed loop in the aspects of signal processing, perceptual psychology and human-computer interaction, and user test verification shows that the time for searching the object can be shortened by more than 40%.

[0019] The above only describes the preferred embodiments of the present application, and does not limit the present application, any modification, equivalent replacement and improvement within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A system for lighting up based on Bluetooth AOA positioning to find a lost item, characterized in that, The system comprises: A Bluetooth tag module, which is installed on a target object, has a low-power Bluetooth chip built-in, and is used to periodically broadcast unique ID information; A positioning base station array module, which is deployed in an indoor environment and contains multiple base stations with multiple antennas, is used to receive signals of the Bluetooth tag and extract the angle of incidence through the phase difference of different antennas; A positioning calculation and signal processing module, which is used to fuse multi-point angle of incidence data and calculate the two-dimensional coordinates of the target tag; A prompt light control and multi-modal feedback module, which is used to control the six-color LED light on the Bluetooth tag to flash according to the target coordinates and provide feedback in combination with a buzzer and vibration. 2.The Bluetooth AOA positioning-based light-on system for finding an object according to claim 1, characterized in that: The positioning base station array module supports BLE5.1 and above protocols and realizes multi-point joint angle measurement through inter-base station synchronization. 3.The Bluetooth AOA positioning-based light-on system for finding an object according to claim 2, characterized in that: The prompt light control and multi-modal feedback module switches the light color according to the relative distance between the target and the user: red / green when the distance is short, and blue / yellow when the distance is far.

4. The Bluetooth AOA positioning-based light-on system for finding an object according to claim 3, characterized in that: The buzzer feedback dynamically adjusts the frequency according to the distance: low-frequency intermittent sound at a long distance, and high-frequency continuous sound at a short distance.

5. The Bluetooth AOA positioning based light-on system for finding an object according to claim 1, wherein: The Bluetooth tag module adopts a sleep-wake mechanism and only broadcasts signals when receiving a positioning activation instruction.

6. The Bluetooth AOA positioning-based light-on system for finding an object according to any one of claims 1-5, applied to a Bluetooth AOA positioning-based method for finding an object, characterized in that: The method for finding an object comprises: A user selects a target tag ID through a mobile terminal APP and initiates a request for finding an object; A central controller activates a positioning base station array, receives tag signals, and measures the angle of incidence; A positioning engine fuses multi-point angle data and calculates the real-time coordinates of the tag; A light control instruction is sent to the target tag to make it start six-color light flashing and be accompanied by a buzzer / vibration.

7. The method of claim 1, wherein the method further comprises: receiving a request for a light to be turned on; and determining a location of the light based on the request. In the step of measuring the angle of incidence, the base station calculates the angle of signal arrival through the phase difference of multiple antennas and realizes multi-point joint angle measurement through clock synchronization. 8.The method of claim 1, wherein the method further comprises: receiving a request for a light-on operation from the mobile device; and transmitting a light-on signal to the light device in response to the request for the light-on operation. In the step of calculating the coordinates, Kalman filtering algorithm is used to fuse data and realize multi-tag concurrent positioning. 9.The method of claim 1, wherein: In the light control step, the brightness is dynamically adjusted according to the relative position of the tag and the user: the closer the distance, the higher the brightness, and the flashing frequency is switched when blocked.

10. The method of claim 1, wherein the method further comprises: receiving a request for a location of the object from the user; and transmitting a signal to the object to cause the object to emit a light in response to the request. The multi-modal feedback includes: when the user deviates from the target direction, the buzzer emits intermittent sound; and when approaching the target, it switches to continuous sound.

Images