Intelligent crib monitoring system based on wireless transmission

Abstract

The invention discloses an intelligent baby crib monitoring system based on wireless transmission, and relates to the technical field of baby physiological monitoring, which comprises the following steps: dynamically distributing node communication time slots through wireless networking, adjusting the inclination angle of a laser array, activating a dual-band metamaterial antenna, synchronously collecting light intensity fluctuation signals and electrocardiosignals, and generating a mixed signal flow; carrying out physiological feature extraction on the mixed signal by adopting adaptive filtering and noise conjoint analysis, and generating a clean breathing waveform and an electrocardio waveform through wavelet decomposition and dynamic threshold suppression; respiration and heart rate abnormity is identified based on zero cross point detection and R wave energy verification, and a transmission priority is determined in combination with chaotic key generation and a Markov decision model. Through multi-band collaborative networking and dynamic time slot allocation, hardware-level synchronous acquisition of light intensity fluctuation signals and electrocardiosignals is realized, and signal synchronization errors are controlled in combination with metamaterial antenna dual-channel noise capture and differential common-mode rejection.

Description

Technical Field

[0001] The present invention relates to the technical field of infant physiological monitoring, and in particular to an intelligent baby crib monitoring system based on wireless transmission. Background Art

[0002] In recent years, infant health monitoring technology has gradually developed from single-sensor monitoring (such as piezoelectric film respiration detection, photoplethysmography) to multi-modal fusion. In the prior art, a physiological signal acquisition system based on a wireless body area network can already achieve preliminary synchronous monitoring of electrocardiogram and respiration, and realize data transmission through Bluetooth or ZigBee protocols. However, such systems generally adopt fixed-frequency band communication and static encryption strategies, and there are bottlenecks in multi-node cooperation, dynamic anti-interference, and real-time anomaly response.

[0003] The main deficiencies of the prior art are concentrated in two aspects: First, the synchronous acquisition of multi-modal signals is limited by hardware-level clock deviation and radio frequency interference, and the time-domain alignment error between the light intensity fluctuation signal and the electrocardiogram signal often leads to misjudgment of the respiration cycle; Second, traditional encryption mechanisms (such as AES static encryption) and fixed-priority transmission strategies are difficult to balance security and real-time performance in high-concurrency data stream scenarios, and key update delay and transmission conflicts may exacerbate the system response delay. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an intelligent baby crib monitoring system based on wireless transmission to solve the problems of insufficient accuracy of synchronous acquisition of multi-modal physiological signals and low efficiency of secure transmission.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: The present invention provides an intelligent baby crib monitoring system based on wireless transmission, which includes a networking and acquisition module that dynamically allocates node communication time slots through wireless networking, adjusts the inclination angle of the laser array, and activates the dual-band metamaterial antenna to synchronously acquire the light intensity fluctuation signal and the electrocardiogram signal, and generate a mixed signal stream; a signal optimization module that extracts physiological characteristics from the mixed signal by using adaptive filtering and noise joint analysis, and generates a clean respiration waveform and electrocardiogram waveform through wavelet decomposition and dynamic threshold suppression; an anomaly decision module that identifies respiration and heart rate anomalies based on zero-crossing detection and R-wave energy verification, and determines the transmission priority in combination with chaotic key generation and Markov decision model; a transmission control module that triggers a multi-band collaborative transmission strategy according to the priority, implements differential encryption and dual-link protocol encapsulation, and establishes a time-slot response and dynamic retransmission mechanism; The data fusion module uses an extended Kalman filter to fuse multi-node observation data and generates bed attitude adjustment instructions through a model predictive controller; The intervention execution module, based on a fuzzy logic controller, hierarchically triggers an acoustic-optic-mechanical composite intervention to execute the bed attitude adjustment instructions.

[0007] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the generation of the hybrid signal stream includes the following steps, Dynamically allocate node communication time slots through wireless networking, adjust the inclination angle of the quantum cascade laser array, synchronously activate the dual-band metamaterial antenna, generate an optical intensity sequence using the laser pulse reflection signal, and simultaneously collect the electrocardiogram signal and environmental noise; Eliminate the common-mode interference through differential amplification, and combine a sliding window to compensate for the transmission delay to generate a hybrid signal stream.

[0008] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the extraction of physiological characteristics from the hybrid signal includes the following steps, Input the hybrid signal stream into a low-noise amplifier to eliminate signal reflection interference and generate an amplified analog signal stream; Parallelly input the amplified analog signal stream into a respiratory signal channel processing branch and an electrocardiogram signal channel processing branch; Filter the respiratory signal of the respiratory signal channel processing branch to generate a stable respiratory analog signal; Perform anti-interference processing on the electrocardiogram signal of the electrocardiogram signal channel processing branch and output a normalized electrocardiogram analog signal in amplitude.

[0009] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the wavelet decomposition and dynamic threshold suppression refer to using Daubechies wavelet to decompose the respiratory waveform and electrocardiogram waveform, jointly analyzing the noise characteristics based on the energy distribution of the high-frequency detail layer, dynamically suppressing the interference components through an adaptive threshold algorithm, and reconstructing a clean respiratory waveform and a complete electrocardiogram waveform of the QRS complex group that retain physiological characteristics.

[0010] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the identification of respiratory and heart rate abnormalities includes the following steps, Detect the respiratory cycle based on zero-crossing points, and trigger a respiratory abnormality event when the instantaneous frequency exceeds the safe range in multiple consecutive sliding windows; Identify the position of the R wave through first-order difference and energy verification, construct a dynamic heart rate curve, and trigger a heart rate abnormality event when the smoothed heart rate exceeds the safe limit value at multiple consecutive data points.

[0011] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the trigger multi-band cooperative transmission strategy includes the following steps, Generate the Lorenz chaotic initial key using the microsecond-level time component, and iteratively generate the dynamic key stream through the fourth-order Runge-Kutta method; Combine the Markov decision model to map the abnormal state to a four-dimensional state vector, and trigger the multi-band cooperative transmission strategy according to the transmission priority weight ratio.

[0012] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the transmission strategy refers to setting the high-priority threshold as and the medium-priority threshold as When the respiration weight is greater than or equal to , activate the high-frequency orthogonal frequency division multiplexing. When the electrocardiogram weight is in the range of , enable the relay node direct sequence spread spectrum. Otherwise, adopt the narrowband sleep mode to transmit the environmental data, and dynamically adjust the modulation order and spread spectrum code rule based on the link quality.

[0013] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the implementation of differential encryption and dual-link protocol encapsulation, and the establishment of the slotted response and dynamic retransmission mechanism includes the following steps, Bind the data shards and the chaotic key by symbol level, bit level and phase association, and perform differential encryption operations; Encapsulate the encrypted data through the dual-link protocol. Add a cyclic redundancy check code to the main link, transmit the control instruction on the backup link, and trigger dynamic retransmission based on the slotted response mechanism.

[0014] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the generation of the bed body attitude adjustment instruction refers to using the extended Kalman filter to fuse the multi-node observation data, constructing the hidden state transition model of the respiration frequency and heart rate, generating the bed body tilt angle adjustment instruction through the model predictive controller, and injecting the adjustment amount into the Kalman filter to update the state estimation.

[0015] As a preferred solution of the intelligent baby crib monitoring system based on wireless transmission according to the present invention, wherein: the execution of the bed body attitude adjustment instruction refers to triggering the actions of sound and light soothing, mechanical vibration and mechanical arm patting based on the classification of the respiration abnormality level. During the execution of the bed body attitude adjustment instruction, turn off the laser and the antenna receiving channel, and restart the data acquisition after the attitude adjustment is completed and compensate for the missing state estimation during the interruption.

[0016] The beneficial effects of the present invention are as follows: Through multi-band collaborative networking and dynamic time slot allocation, hardware-level synchronous acquisition of light intensity fluctuation signals and electrocardiogram signals is achieved. Combining dual-channel noise capture and differential common-mode suppression of metamaterial antennas, signal synchronization error control is realized; Based on the adaptive threshold algorithm of joint analysis of wavelet decomposition and statistical noise characteristics, motion artifacts and respiratory characteristic signals can be dynamically distinguished, improving the suppression rate of respiratory waveform baseline drift. At the same time, the energy integrity of QRS complexes is retained, avoiding the loss of effective signals caused by fixed threshold algorithms; Through chaotic key generation and Markov decision-making models, dynamic mapping of abnormal events and transmission priorities is realized. In the case of abnormal breathing states, the orthogonal frequency division multiplexing strategy is used to compress data transmission delays, and phase rotation encryption is used to enhance anti-interception capabilities, ensuring end-to-end security of key physiological data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a flowchart of mixed signal generation in this embodiment.

[0019] Figure 2 It is a flowchart of physiological signal optimization in this embodiment.

[0020] Figure 3 It is a flowchart of abnormal decision-making and transmission control in this embodiment.

[0021] Figure 4 It is a flowchart of data fusion and intervention execution in this embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the specific embodiments of the present invention in detail with reference to the drawings of the specification.

[0023] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0024] Second, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an individual or alternative embodiment that is mutually exclusive with other embodiments.

[0025] In this embodiment, referring to Figures 1 to 4 , this embodiment provides an intelligent baby crib monitoring system based on wireless transmission, including the following steps: Start the quantum cascade laser array, establish a Mesh network topology through the wireless networking coordination module, and dynamically allocate node communication time slots; precisely adjust the inclination angle of the laser array under the temperature feedback control system to ensure that the light beam covers the baby's chest area; synchronously activate the dual-band (e.g., 2.4 GHz and 5.8 GHz) of the metamaterial RF antenna within the wireless network to complete antenna pattern calibration and phase synchronization between networking nodes; the main control FPGA broadcasts a hardware-level synchronization pulse through the wireless network, and the laser array and the networking node antennas synchronously collect data. The laser captures the light intensity fluctuation, emits multiple light pulses within each sampling period, and receives the reflected light signal through an avalanche photodiode (APD) to generate a light intensity sequence containing respiratory characteristics. At the same time, the dual channels of the metamaterial antenna synchronously collect the electrocardiogram signal and environmental interference noise (full band). After the light intensity sequence, electrocardiogram signal, and environmental interference noise are processed by a differential amplifier to eliminate common-mode interference, two analog signals are output; the wireless network time synchronization engine matches the timestamp differences of the two analog signals through the sliding window algorithm, dynamically compensates for the transmission delay, and outputs a time-synchronized hybrid signal stream.

[0026] Further explanation, the dual channels of the metamaterial antenna include a main channel and an auxiliary channel. Among them, the main channel (e.g., 2.4 GHz) activates the H-shaped resonator array and loads the pre-stored electrocardiogram signal receiving mode. In this embodiment, the bandwidth is set to 0.5 - 40 Hz, and the gain is adjusted to 20 dB. The auxiliary channel (e.g., 5.8 GHz) turns on the full-band scanning mode to dynamically capture environmental electromagnetic noise (interference sources such as Wi-Fi and Bluetooth), and the sensitivity is set to -120 dBm.

[0027] Input the hybrid signal stream into a low-noise amplifier (LNA), and set the gain and noise figure; achieve input impedance matching through an adjustable resistor network to eliminate signal reflection interference and generate an amplified analog signal stream; the amplified analog signal stream enters the processing branch of the respiratory signal channel for filtering processing, and at the same time, it is parallelly input into the processing branch of the electrocardiogram signal channel for anti-interference processing.

[0028] Further explanation, the filtering processing of the respiratory signal channel processing branch includes pre-filtering, main band-pass filtering, and post-stage conditioning.

[0029] Pre-filtering refers to using a second-order Butterworth high-pass filter to filter out low-frequency motion artifacts caused by the baby's turning; main band-pass filtering refers to suppressing high-frequency noise caused by ambient light fluctuations by cascading a fourth-order Chebyshev band-pass filter; post-stage conditioning refers to dynamically adjusting the amplitude of the amplified analog signal through an automatic gain control (AGC) circuit to ensure a stable output peak and generate a stable respiratory analog signal.

[0030] The processing branch of the electrocardiogram signal channel includes power frequency notch filtering, dynamic band-pass filtering, and common-mode rejection.

[0031] Power frequency notch filtering refers to using a twin-T active filter to eliminate power line interference; dynamic band-pass filtering refers to using a programmable switched-capacitor filter to adaptively suppress electromyographic noise; common-mode rejection refers to eliminating electrode contact noise through an instrumentation amplifier and outputting a normalized electrocardiogram analog signal.

[0032] Quantify the stable respiratory analog signal, use modulation technology to improve the effective resolution, and generate a digital respiratory waveform; combine oversampling technology to reduce the noise floor of the electrocardiogram analog signal and output a digital electrocardiogram waveform with effective resolution; use the fourth-order Daubechies wavelet basis to decompose the digital respiratory waveform into five levels, separating the high-frequency detail coefficients containing noise and the low-frequency approximation coefficients reflecting physiological characteristics; simultaneously decompose the digital electrocardiogram waveform into six levels and retain the intermediate-frequency components reflecting the QRS complex; based on the energy distribution of the decomposed high-frequency detail layer, combine the noise characteristics between wireless networking nodes for joint analysis, identify the noise area, quantify the intensity of motion artifacts in the digital respiratory waveform and the activity level of electromyographic interference in the digital electrocardiogram waveform through statistical methods, and generate a noise characteristic distribution map; use an adaptive threshold algorithm based on statistical characteristics to dynamically suppress the noise characteristic distribution map, perform soft threshold truncation on the high-frequency interference components in the digital respiratory waveform, and introduce a protection mechanism for the QRS complex characteristic area in the digital electrocardiogram waveform to avoid loss of effective signals; after completing noise suppression, select the approximation coefficients and optimized detail coefficients that retain the effective physiological components for signal reconstruction to generate a clean respiratory waveform without baseline drift and an electrocardiogram waveform with a complete QRS complex; superimpose physiological state markers (such as apnea event markers, arrhythmia warning signs) on the clean respiratory waveform without baseline drift and the electrocardiogram waveform with a complete QRS complex to generate a pure physiological signal.

[0033] It should be noted that by dynamically allocating communication time slots through the Mesh network topology, the probability of multi-node transmission conflicts is reduced; the inclination angle is adjusted based on the laser temperature feedback to ensure that the optical pulse covers the baby's chest area, eliminating signal loss caused by body position offset; the main channel focuses on collecting electrocardiogram signals, and the auxiliary channel captures the full-band environmental noise to achieve synchronous identification of interference sources; the common-mode interference (such as power line noise) is eliminated through a differential amplifier, and the effective physiological signal characteristics are retained; the second-order Butterworth high-pass filter (cut-off frequency 0.1 Hz) effectively suppresses the low-frequency interference caused by the baby's turning over; the dual-T active notch filter (center frequency 50 / 60 Hz) suppresses the power line interference and retains the QRS complex; the QRS complex region (mid-frequency component) is protected by a hard threshold to avoid loss of effective signals; the respiratory waveform is reconstructed to eliminate the baseline shift caused by the baby's movement.

[0034] The respiratory cycle of the clean respiratory waveform in the pure physiological signal is identified by the zero-crossing detection method. Specifically, when the waveform crosses the zero axis from the negative direction, it is marked as the starting point of inspiration, and when the waveform crosses the zero axis back from the positive direction, it is marked as the starting point of expiration. The time difference between adjacent starting points of inspiration is the duration of a single respiratory cycle, and a continuous respiratory cycle sequence is obtained; a sliding time window is established to calculate the instantaneous frequency.

[0035] Furthermore, it is explained that in this embodiment, a window is set to 5 cycles. When the instantaneous frequency is lower than 20 breaths per minute or higher than 60 breaths per minute for 3 consecutive windows (such as 15 cycles), a respiratory abnormality event is triggered, and the abnormal start timestamp and frequency deviation value are recorded; the electrocardiogram waveform of the complete QRS complex is subjected to first-order differentiation, and when the differential value exceeds 3 times the standard deviation of the baseline noise, the candidate R wave (the positive maximum peak generated during the rapid ventricular depolarization process) is marked; the signal energy within the 50 ms interval before and after the candidate R wave is calculated, and the artifact points with sudden energy changes are removed to obtain the R wave position sequence; a dynamic heart rate curve is constructed based on the R wave sequence, the instantaneous heart rate is calculated, and a moving average is performed on the heart rate data in the past 10 seconds to obtain a smooth heart rate curve; when the smooth heart rate is lower than 100 beats per minute or higher than 160 beats per minute for 5 consecutive data points (≥10 seconds), a heart rate abnormality event is triggered, and the abnormal start time and the maximum heart rate deviation value are recorded.

[0036] It should be noted that through the double verification of the negative crossing of the waveform (starting point of inspiration) and the positive re-crossing (end point of expiration), the detection error of the respiratory cycle is reduced; based on the preprocessed clean respiratory waveform, the baseline fluctuation interference caused by the baby's movement is eliminated; through the continuous determination of the over-limit trigger, false judgment of transient interference is avoided.

[0037] Obtain the microsecond component of the current time, generate a 256-bit entropy source by applying the SHA-256 hash function to the microsecond component of the current time, and intercept the last 24 bits as the initial condition of the Lorenz chaotic key sequence, thus completing the hard-coded loading of the chaotic key parameters; use the fourth-order Runge-Kutta method to perform iterative calculations on the initial conditions to generate a continuous sequence of chaotic variables; perform variable normalization and significant bit extraction processing on each chaotic variable to obtain chaotic variable segments; perform bitwise exclusive-OR confusion on the chaotic variable segments to generate a dynamic key block.

[0038] Further explanation, the bitwise exclusive-OR confusion includes preliminary splicing, non-linear confusion, and cyclic shift.

[0039] Preliminary splicing is to combine the chaotic variable segments into a key block; non-linear confusion is to replace each byte in the key block through S-box substitution (using the S-Box of AES); cyclic shift is to perform a 3-bit left shift operation on the key after non-linear confusion.

[0040] Establish a double-buffer queue to alternately store the newly generated key blocks. When the main control module issues an encryption request, extract the keys from the active buffer in FIFO order, and at the same time inject the last 4-bit random perturbation (replace the last 4 bits of each byte) through a hardware true random number generator, and finally generate a dynamic Lorenz encryption key stream.

[0041] Map the triggered respiratory abnormality and heart rate abnormality event states (normal, respiratory abnormality, heart rate abnormality, combined abnormality) to four-dimensional state variables of the Markov decision model; establish a three-dimensional action space containing three actions: preferentially transmit respiratory data, preferentially transmit electrocardiogram data, and balanced transmission based on the activation of the four-dimensional state vector, and define an immediate reward rule according to the matching degree between the action and the state. For example, when preferentially transmitting respiratory data in the respiratory abnormality state, a reward value of +5 is given to form a reward matrix; use the numerical integration method to approximate the continuous-time value function and divide the future time axis into discrete intervals; based on the historical state transition data (such as the average time for the respiratory abnormality to turn normal within the past 1 hour), count the transition frequencies within the discrete intervals to generate a state transition rate matrix; combine the immediate benefits provided by the reward matrix and the cumulative expected benefits of each state-action pair in the state transition rate matrix under the action of the discount factor; set the initial weight matrix, and calculate the marginal benefits of respiratory data and electrocardiogram data according to the cumulative expected benefits; quantify the weight ratio of the transmission priorities of infant respiratory data and electrocardiogram data based on the marginal benefits; obtain the infant crying situation (crying mode and quiet mode) through an acoustic sensor, and dynamically correct the quantified weight ratio to generate a priority weight vector.

[0042] It should be noted that by using the SHA-256 hash function to generate the initial key and adopting the Lorenz chaotic system and the fourth-order Runge-Kutta method to iteratively generate the dynamic key block, the complexity and unpredictability of the encryption algorithm are increased, and the security of data transmission is improved; the Markov decision process (MDP) is used to determine the optimal transmission strategy for respiratory data and electrocardiogram data, which can intelligently adjust the priority of data transmission according to the current state of the baby (normal, abnormal breathing, abnormal heart rate or combined abnormality), so as to ensure the timely transmission of critical information in case of emergency; through the design of the double-buffer queue and the perturbation injected by the hardware true random number generator, the dynamic update and real-time performance of the key stream are guaranteed, which helps to quickly respond to the changes in the baby's state.

[0043] Based on the priority weight vector, analyze the weight ratio and divide the priorities, and trigger the multi-band cooperative transmission strategy of the wireless network. In this embodiment, set the high-priority threshold Greater than or equal to 0.6, and the medium-priority threshold The interval range is [0.3, 0.6), and the specific operations are as follows: When the respiratory weight ≥ 0.6, trigger the high priority, activate the high-frequency band orthogonal frequency division multiplexing of the wireless network, and at the same time turn off the non-essential radio frequency channels, load the orthogonal frequency division multiplexing parameters, configure multiple subcarriers and set a specific interval, adopt high-order quadrature amplitude modulation, and convert the respiratory data into a time-domain orthogonal frequency division multiplexing baseband signal through the inverse fast Fourier transform; When 0.3 ≤ electrocardiogram weight < 0.6, trigger the medium priority, enable the direct sequence spread spectrum of the relay node of the wireless network, adopt the Gold code to enhance the anti-interference ability, select the lowest-delay path within 3 hops through the network topology optimization, adopt a sequence code of a specific length for direct sequence spread spectrum, and perform a logical operation on the phase modulation symbol and the spread spectrum code sequence to generate a baseband chip stream resistant to narrowband interference; When the triggering conditions of both the high priority and the medium priority are not met, enable the narrowband sleep mode of the wireless network for the default allocation of low-priority environmental data, implement a semi-static scheduling strategy in the narrowband channel, encapsulate the environmental data into fixed-length data packets, and perform differential phase modulation at periodic intervals.

[0044] Input the modulated baseband signals of each mode into the multi-channel radio frequency front end of the metamaterial antenna, where the high-frequency band channel boosts the signal to a high power level through a power amplifier, the medium-frequency band channel maintains a medium power level, and the narrowband channel limits the power spectral density.

[0045] Further explanation: During the RF signal transmission process, the link quality indicators feedback from the receiving end are collected in real time. When it is detected that the quality difference of the orthogonal frequency division multiplexing subcarriers exceeds the determination critical value of the difference, the adaptive modulation algorithm is triggered to reduce the modulation order of the low-quality subcarriers; when the direct sequence spread spectrum link encounters interference, the spread spectrum code generation rule is dynamically switched to enhance the anti-interference ability; when the narrowband transmission fails continuously, the scheduling period is shortened and the transmission power is increased.

[0046] It should be noted that different types of data (breathing, electrocardiogram, and environmental data) are prioritized based on the priority weight vector to ensure that critical information (such as data during abnormal breathing or electrocardiogram) can be transmitted with higher efficiency and quality, thereby optimizing the use of wireless network resources; orthogonal frequency division multiplexing (OFDM), direct sequence spread spectrum (DSSS), and Gold codes are used to enhance the anti-interference ability and stability of the system. Especially in a complex electromagnetic environment, signal loss or error can be effectively avoided; through measures such as adaptive modulation algorithms, dynamically switching spread spectrum code generation rules, and adjusting transmission parameters according to the link quality, the reliability and efficiency of data transmission are further improved, ensuring stable data communication even under adverse conditions; different RF channel power settings are adopted for data with different priorities, which not only helps to extend the battery life of the device, but also reduces unnecessary energy consumption, while ensuring the efficient transmission of high-priority data.

[0047] The high-frequency orthogonal frequency division multiplexing baseband signal, intermediate-frequency direct sequence spread spectrum chip stream, and narrowband environmental data generated by the RF front-end processing are segmented into data shard units, and the data shard units are integrated to form a standardized data block to be encrypted; the standardized data block to be encrypted is bound to the dynamic Lorenz encryption key stream, where the high-frequency data shards are byte-level aligned with the chaotic key, the intermediate-frequency chip stream is bit-level matched with the key, and the narrowband data symbols are phase-correlated with the key, obtaining a time-synchronized key-data pair; the dynamic Lorenz key stream is obtained through the wireless networking key distribution center to implement differential encryption.

[0048] Further explanation: The differential encryption operation refers to performing symbol-level non-linear substitution on the high-frequency signal, implementing logical operation confusion on the intermediate-frequency chip stream, and performing phase rotation encryption on the narrowband data to generate an encrypted RF baseband signal.

[0049] The encrypted RF baseband signal is processed by the dual-link protocol encapsulation module. The main link performs multi-path transmission through the wireless networking backbone nodes, encapsulates the high-frequency and intermediate-frequency encrypted data and attaches a cyclic redundancy check code. The backup link uses the Bluetooth Mesh of the networking edge nodes to encapsulate the narrowband control instructions and link status identifiers, obtaining a dual-link protocol data frame with an integrity verification label.

[0050] Furthermore, a time-slot response mechanism is established during the main link transmission. The receiving end returns an encrypted confirmation signal within a predetermined time window, and a critical condition determination value for triggering the backup link retransmission mechanism is set according to the comprehensive channel quality, historical statistics, and system requirements. When it is detected that the number of consecutive unacknowledged time slots exceeds the critical determination value, the backup link retransmission control process is triggered, and the lost data frame numbers are preferentially sent through the backup channel and the radio frequency channel resources are synchronously switched.

[0051] The receiving end performs hierarchical decryption verification on the dual-link protocol data frames with integrity verification tags. The main link data reversely recovers the original physiological signal through the chaotic key stream, and the backup link instruction drives the power regulation module after decryption to output the decrypted multi-modal physiological signal to execute the control instruction.

[0052] Furthermore, when a hash value conflict is detected between the main and backup link data, the arbitration mechanism is started to preferentially use the backup link data and mark the security event to trigger the chaotic key regeneration module to update the dynamic key sequence. A forward security protection mechanism is carried out during the encryption process, injecting the last key perturbation factor of the current encryption operation into the subsequent chaotic iteration process, and at the same time physically erasing the used key storage area.

[0053] It should be noted that by binding different types of data shards generated by the radio frequency front-end processing to the dynamic Lorenz encryption key stream and implementing differential encryption, the security of data transmission is enhanced; the dual-link protocol encapsulation module is used to process the encrypted radio frequency baseband signal, and a cyclic redundancy check code is added to the main link to ensure the integrity of data transmission. At the same time, the backup link mechanism provides additional data protection, and can quickly switch to the backup link when the main link has problems to ensure the reliable transmission of data; through measures such as establishing a time-slot response mechanism and setting the critical condition determination value for triggering the backup link retransmission mechanism, the effective management of resources and the improvement of transmission efficiency are achieved, which not only improves the system response speed but also reduces unnecessary energy consumption; when a hash value conflict is detected, the arbitration mechanism is started to preferentially use the backup link data and mark the security event, and at the same time update the dynamic key sequence to enhance the system's fault recovery ability and security; injecting the last key perturbation factor of the current encryption operation into the subsequent chaotic iteration process and physically erasing the used key storage area further strengthens the security of the encryption algorithm and prevents the risk of historical key leakage.

[0054] Taking the instantaneous frequency sequence of the respiratory waveform and the smoothed heart rate curve of the electrocardiogram as the observation vectors, the observation data of multiple networking nodes are fused based on the extended Kalman filter. The state vector is defined as the hidden state of the respiratory frequency and the hidden state of the heart rate, the state transition matrix and the control input matrix are constructed, the process noise covariance and the observation noise covariance are set, and the parameter initialization of the extended Kalman filter is completed.

[0055] Furthermore, the respiratory rate hidden state in the state transition matrix is used to describe the dynamic evolution law of the respiratory rate and reflect the physiological mechanism of the weak regulation of respiration by heart rate. For example, the current respiratory rate = 95% of the respiratory rate in the previous period + 3% of the heart rate in the previous period; the heart rate hidden state corresponds to the kinetic relationship of heart rate changes and reflects the high self-stability of heart rate. For example, the current heart rate = 2% of the respiratory influence in the previous period + 98% of the heart rate in the previous period. The control input matrix is used to map the influence of external interventions (such as bed adjustment) on the state, define the intervention intensity of bed mechanical adjustment on physiological signals, optimize the ventilation efficiency by changing the thoracic volume, and quantify the adjustment range of the heart rate for the same control action (the change in the tilt angle causes fluctuations in venous return, resulting in a slight increase in heart rate). For example, when the bed tilt angle increases by 1°, the respiratory rate increases by 0.1 times per minute.

[0056] Taking the marking of respiratory abnormal events as the trigger condition, when an abnormality is detected, the parameter iteration of the extended Kalman filter after initialization is executed as follows: According to the state estimation and control instruction execution in the previous time slot, calculate the prior state and prior covariance, combine the current observation value, update the posterior state and covariance, and obtain the hidden state estimation value of respiration / heart rate; input the hidden state estimation value of respiration / heart rate output by the extended Kalman filter into the model predictive controller. In the prediction time domain, construct the state space equation, and align the discretization step size with the physiological signal sampling period; define the cost function with the optimization objectives of physiological information tracking accuracy and control action amplitude, take the respiratory rate safety range and heart rate safety range as hard constraint conditions, and solve the optimal control instruction sequence through quadratic programming; monitor the output respiratory rate sequence in real time. When multiple consecutive prediction points are below the respiratory rate safety range and the actual observation value deviates synchronously, it is determined as persistent respiratory abnormality; dynamically calculate the adjustment amount of the bed tilt angle, drive the bed elevation angle to increase through the stepper motor, and at the same time limit the maximum amplitude to avoid excessive intervention; inject the calculated adjustment amount of the bed tilt angle into the control input matrix of the extended Kalman filter as a feedforward quantity, and at the same time feedback the actual angle through the bed pressure sensor to generate the bed attitude control instruction including the target adjustment parameter and safety limit.

[0057] Furthermore, in this embodiment, the respiratory rate safety limit is 20 ≤ respiratory rate ≤ 60 times per minute, and the heart rate safety limit is 100 ≤ heart rate ≤ 160 times per minute.

[0058] It should be noted that by using the instantaneous frequency sequence of the respiration waveform and the electrocardiogram smoothed heart rate curve as the observation vectors and performing data fusion based on the extended Kalman filter, the hidden states of the respiration rate and heart rate can be estimated more accurately, improving the accuracy of monitoring the physiological state of infants; by using the method of combining the extended Kalman filter with the model predictive controller, rapid response can be achieved and the optimal control command can be calculated when respiration or heart rate abnormalities are detected, which helps to take timely measures to prevent potential dangerous situations; by quantifying the effects of mechanical adjustments of the bed body (such as changes in tilt angle) on the respiration rate and heart rate and dynamically adjusting the bed body posture according to the real-time monitoring results, the ventilation efficiency can be effectively improved, the heart rate can be stabilized, and the risks brought by excessive intervention can be avoided.

[0059] Extract the tilt angle target value, the maximum allowable adjustment rate, and the anti-overshoot safety margin in the bed body posture control command. Input the respiration rate deviation, the heart rate variability coefficient, and the control residual into the fuzzy logic controller. Based on the respiration abnormality level classification rules, match the real-time respiration rate with the respiration rate safety limit value. When it is detected that the respiration rate is mildly abnormal, the wireless network drives the acoustic-optic composite stimulation module, encodes the acoustic-optic soothing command into a PWM dimming signal and a class D audio power amplifier drive signal, and sends them to the LED array and the speaker through the metamaterial antenna link, starts the progressive enhancement strategy, gradually increases the light intensity, and at the same time expands the white noise audio band; when the respiration rate is moderately abnormal, the wireless network issues the vibration parameters and the emergency braking conditions to the motor drive sequence, and the link sends a control frame containing the start-stop vibration period, the duty cycle, and the emergency braking conditions (such as when the bed body pressure sensor detects a reaction force > 10 N), drives the eccentric wheel mechanism to generate mechanical vibrations, and synchronously enables the current loop feedback mechanism - automatically switches to the slow start mode when the motor current fluctuation exceeds the rated value; when the apnea overrun event is recognized, decompose the robotic arm movement trajectory into a Cartesian space coordinate sequence, transmit the joint angle command through the wireless network dual-link redundancy, and verify the safe operation space through the laser rangefinder before execution. Automatically switch to the impedance control mode when an obstacle is detected, and monitor the contact force feedback in real time during the tapping action; during the execution of all control commands, turn off the power supply of the APD photodetector (disable the laser drive current) and cut off the metamaterial antenna receiving channel through the hardware interrupt signal, and only maintain the minimized communication link for the execution status feedback; after the posture adjustment is completed or the abnormality is lifted, restart the quantum cascade laser array (such as a 500 ms warm-up delay), and reactivate the full-range data acquisition after the signal-to-noise ratio of the respiration signal is restored. Based on the multi-node wireless network, compensate for the missing state estimation during the interruption of the Kalman filter, and reconstruct the physiological signals during the interruption.

[0060] It should be noted that a fuzzy logic controller is used to process complex input parameters (such as respiratory rate deviation, heart rate variability coefficient, and control residual), so as to achieve dynamic adjustment based on real-time data, improve the effectiveness and pertinence of intervention measures; the wireless networking technology is used to achieve rapid transmission of instructions, and a dual-link redundant transmission is adopted for key instructions (such as the movement trajectory of the robotic arm) to ensure reliable execution of commands in case of emergency. At the same time, unnecessary functions are turned off during the execution of control instructions to reduce potential interference sources; multiple operation modes (such as slow start mode, impedance control mode, etc.) are designed and can be automatically switched according to the actual situation, which not only protects the equipment but also avoids discomfort or harm to the baby.

[0061] In summary, through multi-band collaborative networking and dynamic time slot allocation, the present invention realizes hardware-level synchronous acquisition of light intensity fluctuation signals and electrocardiogram signals. Combining the dual-channel noise capture and differential common-mode suppression of the metamaterial antenna, it controls the signal synchronization error; based on the adaptive threshold algorithm jointly analyzed by wavelet decomposition and statistical noise characteristics, it can dynamically distinguish motion artifacts and respiratory characteristic signals, improve the suppression rate of the baseline drift of the respiratory waveform, while retaining the energy integrity of the QRS complex and avoiding the loss of effective signals caused by the fixed threshold algorithm; through chaotic key generation and Markov decision model, it realizes the dynamic mapping of abnormal events and transmission priorities. In the case of abnormal breathing, the orthogonal frequency division multiplexing strategy is adopted to compress the data transmission delay, and the anti-interception ability is enhanced by phase rotation encryption to ensure the end-to-end security of key physiological data.

[0062] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An intelligent baby crib monitoring system based on wireless transmission, characterized in that: include, The networking acquisition module dynamically allocates node communication time slots through wireless networking, adjusts the laser array inclination and activates the dual-band metamaterial antenna, synchronously collects light intensity fluctuation signals and ECG signals, and generates a mixed signal stream; The signal optimization module uses adaptive filtering and noise joint analysis to extract physiological features from mixed signals, and generates clean breathing waveforms and ECG waveforms through wavelet decomposition and dynamic threshold suppression; The abnormal decision module identifies breathing and heart rate abnormalities based on zero-crossing point detection and R-wave energy verification, and determines the transmission priority by combining chaotic key generation and Markov decision model; The transmission control module triggers the multi-band coordinated transmission strategy according to priority, implements differentiated encryption and dual-link protocol encapsulation, and establishes a time-slotted response and dynamic retransmission mechanism; The data fusion module uses the extended Kalman filter to fuse multi-node observation data and generates bed posture adjustment instructions through the model predictive controller; The intervention execution module, based on the fuzzy logic controller, triggers the acoustic-optical-mechanical composite intervention in stages to execute the bed posture adjustment instructions.

2. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The generating of the mixed signal stream comprises the following steps: Dynamically allocate node communication time slots through wireless networking, adjust the inclination angle of the quantum cascade laser array, synchronously activate the dual-band metamaterial antenna, use the laser pulse reflection signal to generate a light intensity sequence, and simultaneously collect ECG signals and environmental noise; Common-mode interference is eliminated through differential amplification, combined with a sliding window to compensate for transmission delays and generate a mixed signal stream.

3. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The physiological feature extraction of the mixed signal The following steps are included: Inputting the mixed signal stream into a low noise amplifier to eliminate signal reflection interference and generate an amplified analog signal stream; The amplified analog signal stream is input into the respiratory signal channel processing branch and the electrocardiogram signal channel processing branch in parallel; Filtering the respiratory signal of the respiratory signal channel processing branch to generate a stable respiratory simulation signal; The ECG signal of the ECG signal channel processing branch is subjected to anti-interference processing, and an ECG simulation signal with normalized amplitude is output.

4. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The wavelet decomposition and dynamic threshold suppression refer to using Daubechies wavelet to decompose the respiratory waveform and the ECG waveform, jointly analyzing the noise characteristics based on the energy distribution of the high-frequency detail layer, dynamically suppressing the interference components through an adaptive threshold algorithm, and reconstructing a clean respiratory waveform that retains physiological characteristics and an ECG waveform with a complete QRS complex.

5. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: Identifying abnormal breathing and heart rate The following steps are included: The breathing cycle is detected based on the zero-crossing point, and a breathing abnormality event is triggered when the instantaneous frequency in multiple consecutive sliding windows exceeds the safe range; The R-wave position is identified through first-order difference and energy verification, and a dynamic heart rate curve is constructed. When multiple consecutive data points of the smoothed heart rate exceed the safety limit, an abnormal heart rate event is triggered.

6. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The triggering multi-band coordinated transmission strategy includes the following steps: The Lorenz chaotic initial key is generated by using microsecond time components, and the dynamic key stream is generated by the fourth-order Runge-Kutta method iteration; The abnormal state is mapped into a four-dimensional state vector by combining the Markov decision model, and the multi-band collaborative transmission strategy is triggered according to the transmission priority weight ratio.

7. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The transmission strategy refers to setting the high priority threshold as , the medium priority threshold is , when the breathing weight is greater than or equal to When the high frequency band OFDM is activated, when the ECG weight is When the link quality is within the range of , the direct sequence spread spectrum of the relay node is enabled, otherwise the narrowband sleep mode is used to transmit the environmental data, and the modulation order and spread spectrum code rules are dynamically adjusted based on the link quality.

8. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The implementation of differentiated encryption and dual link protocol encapsulation, and establishment of slotted response and dynamic retransmission mechanism includes the following steps: Bind the data slices with the chaotic key at the symbol level, bit level and phase level to perform differentiated encryption operations; Encrypted data is encapsulated through a dual-link protocol, a cyclic redundancy check code is added to the main link, control instructions are transmitted on the backup link, and dynamic retransmission is triggered based on a slotted response mechanism.

9. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: Generating bed posture adjustment instructions refers to using an extended Kalman filter to fuse multi-node observation data, constructing a hidden state transfer model of respiratory frequency and heart rate, generating bed tilt angle adjustment instructions through a model prediction controller, and injecting the adjustment amount into the Kalman filter to update the state estimation.

10. The intelligent baby crib monitoring system based on wireless transmission as claimed in claim 1, characterized in that: The execution of the bed posture adjustment instruction refers to triggering sound and light soothing, mechanical vibration and robotic arm patting actions based on the abnormal breathing level classification, closing the laser and antenna receiving channels during the execution of the bed posture adjustment instruction, and restarting data collection after the posture adjustment is completed to compensate for the state estimation loss during the interruption.

Images