Multiband polarization time sequence radiation measurement method for obtaining optical fingerprints of bacterial samples

Abstract

The invention relates to the technical field of optical measurement and spectral analysis, in particular to a multiband polarization time sequence radiation measurement method for obtaining optical fingerprints of bacterial samples. Comprising the following steps: a light field construction step: selecting a multi-band light source to load an independent intensity modulation frequency, and loading a unified polarization state scanning modulation frequency to construct a composite light field; a signal detection step: enabling a light field to act on a bacterial sample, and obtaining a single-channel time sequence response signal containing a wavelength-dependent responsivity parameter; a frequency domain decoupling step: decoupling and extracting a polarization intensity time sequence change sequence of each wave band by utilizing the frequency domain orthogonality of a physical process; a correlation analysis step: carrying out self-correlation operation on the sequence, and resolving a relaxation characteristic parameter set representing the dynamic depolarization characteristic; a fingerprint generation step: generating a dynamic response fingerprint containing a mapping relation between the wavelength and the time decay rate based on the parameter set; according to the invention, the multi-dimensional light field is analyzed through single-point detection, and the physical nature of bacteria as a light field modulator is clearly clarified.

Description

Multi-band polarization time-series radiometric measurement method for obtaining optical fingerprints of bacterial samples Technical Field

[0001] This invention relates to the field of optical measurement and spectral analysis technology, specifically to a multi-band polarization time-series radiometric measurement method for obtaining optical fingerprints of bacterial samples. Background Technology

[0002] With the continuous development of optical radiation measurement technology, the demand for refined characterization of the light field properties of microscopic scatterers is increasing. Obtaining the characteristic information of samples by measuring the scattering signal after the interaction of light and matter has become an important means in the fields of material analysis and biomonitoring.

[0003] Currently, methods for obtaining the optical properties of such samples generally rely on static scattering spectroscopy measurements or low-frequency polarization analysis. Technicians typically use mechanically rotating polarization elements to modulate the incident light at low speeds, or use spectrometer arrays to record the static light intensity response of different bands, thereby calculating the scattering cross section or Stokes vector parameters of the sample.

[0004] However, traditional measurement methods mainly focus on the static or quasi-static properties of the light field, often ignoring the nanosecond to millisecond-level dynamic modulation effect of microscopic particle motion on the light field; existing mechanical modulation methods have limited scanning speed and are difficult to capture high-frequency temporal change information; at the same time, conventional detection architectures are difficult to achieve effective separation of spectral, polarization and temporal dimensions in single-point detection, resulting in problems such as multidimensional information aliasing, low signal-to-noise ratio and incomplete dynamic feature extraction when processing complex scattering signals;

[0005] Therefore, how to achieve high-speed dynamic decoupling of multi-dimensional optical field signals to obtain accurate optical fingerprints containing temporal correlation characteristics has become an urgent problem to be solved in this field.

[0006] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention discloses a multi-band polarization temporal radiometric measurement method for acquiring optical fingerprints of bacterial samples. Specifically, the technical solution of this invention is as follows:

[0008] The method aims to accurately measure the dynamic parameters of the light field modulated by the sample, including:

[0009] A modulated composite probe optical field is constructed, wherein the modulated composite probe optical field contains multiple spectral band components with independent intensity modulation frequencies, and the entire optical field is loaded with a uniform polarization state scanning modulation frequency.

[0010] The modulated composite probe light field is applied to the bacterial sample to be tested, and the single-channel time-series response signal of the composite light field returned or transmitted after being applied to the sample is measured.

[0011] The single-channel time-series response signal is decoupled in the frequency domain to extract the polarization intensity time-series change sequence corresponding to each independent spectral band component constituting the composite optical field;

[0012] A time-domain correlation analysis was performed on the polarization intensity time-series variation sequence to calculate the set of relaxation characteristic parameters characterizing the dynamic depolarization characteristics of the composite probe field;

[0013] Based on the relaxation feature parameter set and the wavelength of each spectral band component, a dynamic response fingerprint is generated to characterize the modulated composite probe light field under the action of a specific sample.

[0014] Optionally, constructing the modulated composite probe optical field includes:

[0015] According to the preset frequency domain layer isolation principle, the intensity modulation frequency and the polarization state scanning modulation frequency are set;

[0016] The frequency domain layering isolation principle ensures that each intensity modulation frequency is completely separated from the others and from the polarization state scanning modulation frequency in the frequency domain, and that all modulation frequencies are greater than the physical perturbation characteristic frequency introduced by the sample.

[0017] Optionally, the frequency domain decoupling process for the single-channel timing response signal includes:

[0018] Based on preset intensity modulation frequencies, the single-channel timing response signal is subjected to first-level digital phase-locked demodulation to separate and output the baseband signal corresponding to each spectral band.

[0019] Based on the polarization state scanning modulation frequency, a second-level sideband spectrum analysis is performed on each baseband signal to resolve the temporal variation sequence of light intensity with polarization state scanning in that spectral band.

[0020] Optionally, the temporal correlation analysis of the polarization intensity time-series variation sequence includes:

[0021] For each spectral band, calculate its autocorrelation function for the time-series variation of polarization intensity.

[0022] The autocorrelation function curve is fitted to extract the decorrelation time constant of the optical field signal in this band;

[0023] The relaxation feature parameter set includes at least a set consisting of the self-decorrelation time constants of each band.

[0024] Optionally, the method further includes:

[0025] Select the time series of polarization intensity changes corresponding to different spectral bands and calculate the cross-correlation function between each pair;

[0026] Based on the signal lag or lead relationship between the bands reflected by the peak value of the cross-correlation function, the inter-band coupling coefficient of the optical field depolarization process is calculated.

[0027] The inter-band coupling coefficient is used as an additional parameter and incorporated into the relaxation characteristic parameter set.

[0028] Optionally, generating the dynamic response fingerprint includes:

[0029] Construct a multidimensional feature space with optical field parameters as coordinates, whose dimensions include at least the radiation wavelength dimension, the polarization state change dimension, and the time dynamic dimension;

[0030] The relaxed feature parameter set is mapped to the multidimensional feature space to form a dynamic optical field fingerprint corresponding to the current measurement conditions.

[0031] Optionally, the method further includes:

[0032] Real-time monitoring of the signal-to-noise ratio of the single-channel timing response signal;

[0033] Establish a model relating the signal-to-noise ratio to the polarization state scanning modulation frequency;

[0034] Based on the model, the polarization state scanning modulation frequency is dynamically adjusted so that the signal-to-noise ratio is always maintained in the optimal measurement range.

[0035] Optionally, the bacterial sample is used as a test medium that causes dynamic changes in the light field;

[0036] In the step of generating a fingerprint to characterize the dynamic response of the modulated composite probe optical field, the generated fingerprint directly reflects the intensity relaxation correlation characteristics of each band component of the composite optical field under polarization scanning.

[0037] Optionally, constructing the modulated composite probe optical field includes:

[0038] Multiple laser sources with different center wavelengths are used to generate initial beams, and independent intensity modulation is applied to each beam.

[0039] Multiple light beams, after intensity modulation, are combined to form a composite beam;

[0040] The overall polarization state of the composite beam is periodically scanned and modulated using an electro-optic modulator or an acousto-optic modulator, and the modulated composite probe light field is finally output.

[0041] Compared with existing technologies, this invention constructs a composite probe light field containing multiple band components and loaded with uniform polarization state scanning modulation. Based on the principle of frequency domain layering and isolation, it sets a modulation frequency much higher than the physical perturbation frequency of the sample. This enables the high-dimensional light-matter interaction information in a single-channel signal to be compressed into a one-dimensional time signal, thereby effectively preventing the aliasing interference of multiple independent signals in the time domain. By performing two-stage cascaded digital phase-locked demodulation and sideband spectrum analysis on the single-channel time-series response signal, the orthogonality of the physical process in the frequency domain can be used to accurately separate each independent spectral band component and the corresponding polarization intensity time-series change sequence. This enables the simultaneous analysis of multi-dimensional light field information through a single-point detector, solving the problems of high hardware cost, complex optical path alignment, and difficulty in decoupling in traditional multi-channel stacking methods. Attached Figure Description

[0042] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0043] Figure 1 is a flowchart of the method of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0045] Example 1:

[0046] Please refer to Figure 1. A multi-band polarization time-series radiometric measurement method for obtaining optical fingerprints of bacterial samples is proposed. The method aims to accurately measure the dynamic parameters of the light field modulated by the sample. The method includes: constructing a modulated composite probe light field, wherein the modulated composite probe light field contains multiple spectral band components with independent intensity modulation frequencies, and the entire light field is loaded with a uniform polarization state scanning modulation frequency.

[0047] The modulated composite probe light field is applied to the bacterial sample to be tested, and the single-channel time-series response signal of the composite light field returned or transmitted after being applied by the sample is measured; the single-channel time-series response signal is decoupled in the frequency domain to extract the polarization intensity time-series change sequence corresponding to each independent spectral band component constituting the composite light field.

[0048] Based on this, a temporal correlation analysis was performed on the polarization intensity time-series variation sequence to calculate the relaxation characteristic parameter set characterizing the dynamic depolarization characteristics of the composite probe optical field; based on the relaxation characteristic parameter set and the wavelength of each spectral band component, a dynamic response fingerprint for characterizing the modulated composite probe optical field under the action of a specific sample was generated.

[0049] The core of this embodiment lies in constructing a time-domain frequency division multiplexing polarization coding architecture, aiming to simultaneously resolve multi-dimensional light field information through a single-point detector. The system performs the operation of constructing a modulated composite detection light field, which is simultaneously structured and encoded in the time, frequency, and polarization domains. The purpose is to compress high-dimensional light-matter interaction information into a one-dimensional time signal. Specifically, the system selects... Each with an independent center wavelength Narrowband laser sources, such as Corresponding to 280nm, Corresponding to 532nm; an independent intensity modulation frequency is applied to each beam of light. At this time, the first Initial light intensity of each band It follows the physical laws as follows:

[0050]

[0051] in, : Derived from the laser drive current setting, its physical meaning is the first Average optical power in the band, measured in mW;

[0052] Derived from preset configuration, its physical meaning is modulation depth, which is dimensionless;

[0053] : Derived from a function generator, its physical meaning is intensity-modulated carrier frequency, and the unit is Hz;

[0054] The system will this Beam combining is performed, and a uniform polarization state scanning modulation frequency is applied through a polarization state generator. This makes the Stokes vector of the entire light field... On the Poincaré ball, with frequency Periodic trajectory scanning is performed; after the light field is constructed, the light field is used to illuminate the bacterial sample. As a micrometer-sized scatterer with a complex refractive index distribution and chiral structure, the Brownian motion of bacteria dynamically modulates the incident light; the detector receives the photons scattered by the sample and outputs a single-channel time-series response signal. To ensure the mathematical rigor of the subsequent decoupling algorithm and to comply with physical conservation laws, this signal... It is modeled as a weighted superposition of components in the time domain, and a wavelength-dependent responsivity parameter is explicitly introduced:

[0055]

[0056] in, : Derived from the calibration of the detection system, its physical meaning is that the detector at the [number]th [time]... The overall photoelectric conversion efficiency of the band is expressed in V / mW; this parameter can be obtained by consulting the detector at the wavelength. Calibration spectral response curve at [location] To obtain, in order to correct the physical differences in photoelectric responsivity at different wavelengths;

[0057] : This is the normalized polarization modulation transfer function. Considering the two symmetrical projections of the intensity vector corresponding to a complete scan of the polarization state on the Poincaré sphere, this function is specifically modeled as follows: To match the requirements in subsequent embodiments Frequency demodulation requirements;

[0058] : For bacteria in the first In this model, the normalized scattering coefficient, which fluctuates with time across the band, is specifically defined as the beam interception efficiency or effective scattering modulation factor, and mathematically expressed as the physical scattering cross section. Units are With the effective illumination area of ​​the beam Units are The ratio, i.e. The physical meaning of this definition lies in quantifying the probability that an incident photon will be scattered by a sample, that is, characterizing the sample's response to the incident light power. Modulation capability;

[0059] Compared to traditional scattering efficiency factors that only describe particle properties This parameter By combining the geometric characteristics of the beam, a physical connection between microscopic scattering and macroscopic optical power attenuation / modulation was directly established, ensuring the strict consistency between the multiplication of each physical dimension (mW) and the dimensionless coefficient in the formula and the detector output voltage, thus clearly elucidating the physical nature of bacteria as a light field modulator. : Represents time-domain multiplication; This represents the additive noise component within the system;

[0060] This physical model shows that each component behaves in the frequency domain as follows: The system focuses on the sideband distribution centered on the baseband, rather than baseband aliasing; the system... Decoupling is achieved by utilizing the orthogonality of different physical processes in the frequency domain, and digital signal processing is used to separate the mixed signal into segments corresponding to different wavelengths. independent components ;

[0061] Based on this, for each Perform autocorrelation calculations to obtain the set of relaxation characteristic parameters. This parameter set characterizes the rate at which the correlation of the polarization state of the light field decays over time after being scattered by the sample, and its physical essence corresponds to the rotational diffusion coefficient of bacteria. ;based on and wavelength Generate dynamic response fingerprints The fingerprint is a feature vector that contains the mapping relationship between wavelength and time decay rate.

[0062] Constructing a modulated composite probe optical field specifically includes:

[0063] According to the preset frequency domain layering isolation principle, the intensity modulation frequency and the polarization state scanning modulation frequency are set; wherein, the frequency domain layering isolation principle ensures that each intensity modulation frequency is completely separated from each other and from the polarization state scanning modulation frequency in the frequency domain, and all modulation frequencies are greater than the physical perturbation characteristic frequency introduced by the sample.

[0064] This embodiment further defines the selection rules for the modulation frequency, namely the frequency domain layered isolation principle. This is a spectrum planning strategy to prevent signal aliasing and ensure that each physical process can be independently resolved. Specifically, the frequency setting satisfies the following inequality chain constraint:

[0065]

[0066] in, : Derived from system settings, its physical meaning is the first The intensity modulation frequency of the band is set in the MHz range, and the frequency difference between any two bands is... Must meet ,in This is the maximum polarization scan frequency allowed by the system;

[0067] : Derived from the polarization controller setting, its physical meaning is the polarization state scanning modulation frequency, set in the kHz range;

[0068] : Derived from the physical properties of the sample, the physical meaning is the characteristic frequency of bacterial Brownian motion and biological rhythm, usually in the range of Hz to hundreds of Hz;

[0069] This embodiment ensures that the signals of each spectral band do not overlap in the frequency domain through strict frequency domain layering, thereby enabling perfect signal separation through bandpass filtering. At the same time, the polarization scanning speed is set to be much faster than the movement speed of the bacteria, so that the system has completed a complete polarization state scan before the bacteria undergo significant displacement or rotation. This allows the system to capture the instantaneous polarization response, avoids measurement ambiguity caused by the movement of the organism, and ensures the accuracy of dynamic optical field parameters.

[0070] The frequency domain decoupling process for a single-channel time response signal specifically includes: performing a first-level digital phase-locked demodulation on the single-channel time response signal based on preset intensity modulation frequencies, separating and outputting the baseband signal corresponding to each spectral band; and performing a second-level sideband spectrum analysis on each baseband signal based on the polarization state scanning modulation frequency to analyze the time-series change sequence corresponding to the second harmonic of the polarization state scanning frequency under that spectral band.

[0071] This embodiment details the algorithm flow for frequency domain decoupling processing, corrects the original convolution description, and adopts a standard digital mixing and filtering architecture to ensure consistency between mathematics and physics. First, it performs a first-stage digital phase-locked demodulation to separate wavelengths, targeting single-channel signals. Using a preset intensity modulation frequency As a reference signal, perform quadrature mixing and low-pass filtering; for the first... Each band, its in-phase component Orthogonal components and the final baseband signal The calculation logic is as follows:

[0072]

[0073]

[0074]

[0075] in, : This indicates dot product operation in the time domain, also known as frequency mixing, which shifts the signal spectrum to zero frequency and... Place;

[0076] : Represents temporal convolution operation;

[0077] : This is the unit impulse response function of the first-stage low-pass filter. To balance time-domain response speed and frequency-domain isolation, this function is designed as a Hamming window-weighted Sinc function. ,in, For Hamming window functions, Set the cutoff frequency to meet the following conditions. The lower limit here is set as follows: This is to ensure the frequency is The effective polarization modulation signal can pass through without loss, avoiding accidental filtering; : Originates from demodulation output, physical meaning is the first Dynamic signal of light intensity after band separation;

[0078] Perform second-level sideband spectrum analysis for each Synchronous demodulation was performed again to extract polarization features; to extract Taking the component as an example, calculate the temporal variation sequence of light intensity with polarization state scanning. :

[0079]

[0080]

[0081]

[0082] in, The impulse response of the second-stage smoothing filter has a cutoff frequency of... Set as:

[0083]

[0084] in, For safety factor; The estimated maximum characteristic frequency of bacteria;

[0085] The sequence Accurately reflects the bacterial sample at the time The instantaneous envelope of the scattering cross section at wavelength as a function of polarization state;

[0086] This embodiment employs a two-stage cascaded quadrature mixer and filter architecture, strictly adhering to the phase-locked amplification principle, and removes the high-frequency carrier through the first mixing and demodulation. The specific polarization modulation harmonic components are extracted through a second mixing and demodulation, thereby reproducing the multi-band, multi-dimensional physical parameters in the single-point detector signal with high fidelity.

[0087] The temporal correlation analysis of the polarization intensity time series specifically includes: calculating the autocorrelation function of the polarization intensity time series for each spectral band; fitting the autocorrelation function curve to extract the decorrelation time constant of the optical field signal in that band; wherein, the relaxation feature parameter set contains at least a set consisting of the decorrelation time constants of each band.

[0088] This embodiment details the core algorithm of time-domain correlation analysis, namely the calculation and fitting of the autocorrelation function; for each spectral band, the extracted sequence The system calculates its normalized intensity autocorrelation function. :

[0089]

[0090] in, : Derived from the correlator sampling interval, its physical meaning is the delay time, and the unit is s; : Indicates time averaging;

[0091] Using the exponential decay model to Nonlinear least squares fitting of the curve:

[0092]

[0093] in, Derived from the characteristics of the optical path aperture, its physical meaning is the instrument coherence factor, which is dimensionless;

[0094] : This represents the base of the natural logarithm, also known as the Euler number, which has a value of approximately 2.71828;

[0095] : Derived from the fitting results, its physical meaning is decay rate, and the unit is Hz;

[0096] Therefore, the self-correlation time constant is extracted. ,Right now The constant It is the core element of the relaxation feature parameter set, which physically corresponds to the Brownian motion decorrelation time of bacteria at a specific wavelength;

[0097] This embodiment quantifies complex biological characteristics into precise time constants through autocorrelation analysis. Since different forms of bacteria, such as cocci, bacilli, and vibrio, have drastically different rotational diffusion coefficients, their autocorrelation time constants differ significantly. Furthermore, the differences in cell wall stiffness and internal cytoplasmic viscosity between dead and live bacteria also contribute to the slight shift in this constant. Therefore, this parameter provides a highly sensitive physical criterion for the classification and identification of bacteria.

[0098] The method also includes: selecting the time-series variation sequences of polarization intensity corresponding to different spectral bands and calculating the cross-correlation function between each pair; calculating the inter-band coupling coefficient of the optical field depolarization process based on the signal lag or lead relationship between the bands reflected by the peak value of the cross-correlation function; and incorporating the inter-band coupling coefficient as an additional parameter into the relaxation characteristic parameter set.

[0099] This embodiment introduces inter-band cross-correlation analysis based on autocorrelation; the specific steps are as follows: select bands. and band For example, sequences of ultraviolet and near-infrared bands. and Calculate the cross-correlation function :

[0100]

[0101] in, The term originates from the signal minus its mean, and its physical meaning is the fluctuation portion of the signal. In numerical calculations, to prevent division by zero errors in the denominator caused by the zero variance due to the stability of some signal bands, a minimal regularization constant needs to be introduced into the denominator term. For example ;

[0102] according to peak position and the peak height at that location Calculate the inter-band coupling coefficients during the optical field depolarization process; to strictly comply with the logical distinguishability principle and avoid highly correlated but lagging signals due to product compression. High and Large and weak correlation signals To address the issue of indistinguishable values ​​at low numerical values, this embodiment abandons a single scalar compression algorithm and instead constructs a vector coupling coefficient that includes both magnitude and phase. :

[0103]

[0104] in, The physical meaning is structural coupling strength, which characterizes the morphological similarity of scatterers at different wavelengths;

[0105] The physical meaning is dynamic phase difference, which characterizes the time causal lag of photon transmission between different microscopic structural layers, such as the cell wall and the cell nucleus.

[0106] The system no longer relies on empirical formulas for fuzzy compression, but directly uses the aforementioned vector parameters. As an additional parameter, it is incorporated into the relaxation feature parameter set; this processing method ensures that fingerprint data can distinguish between two distinct biophysical states: weak coupling in phase and strong coupling out of phase, providing complete information entropy support for subsequent high-precision classification.

[0107] Example 2:

[0108] Generating a dynamic response fingerprint specifically includes: constructing a multi-dimensional feature space with optical field parameters as coordinates, the dimensions of which include at least the radiation wavelength dimension, the polarization state change dimension, and the time dynamic dimension; mapping the relaxation feature parameter set to this multi-dimensional feature space to form a dynamic optical field fingerprint corresponding to the current measurement conditions;

[0109] This embodiment describes the generation and mapping process of dynamic response fingerprints; the system constructs a multi-dimensional feature space. Its coordinate axis definition includes: the radiation wavelength axis corresponding to the laser wavelength used. The polarization relaxation axis corresponding to the calculated self-decorrelation time constant and the coupling strength axis of the cross-correlation coefficients between corresponding bands Based on this, the system maps the measured parameter set into a state vector in space. :

[0110]

[0111] That is, the fingerprint vector explicitly includes the inter-band coupling coefficient. modulus component With phase components The vector parameters are expanded into independent feature dimensions; this vector is the dynamic light field fingerprint under the current measurement conditions.

[0112] This embodiment converts complex time-series signals into a fixed point in a high-dimensional space, achieving dimensionality reduction and feature generation of the data. Different types of bacteria will occupy different clustering regions in this space. By calculating the Euclidean or Mahalanobis distance between the vector to be tested and the vector in the standard database, the bacterial species can be quickly identified. This mapping method provides a mathematical basis for automated, high-throughput bacterial detection.

[0113] The method also includes: real-time monitoring of the signal-to-noise ratio of a single-channel timing response signal;

[0114] Establish a model relating signal-to-noise ratio to polarization state scanning modulation frequency;

[0115] Based on the model, the polarization state scanning modulation frequency is dynamically adjusted to keep the signal-to-noise ratio in the optimal measurement range.

[0116] This embodiment introduces a closed-loop signal-to-noise ratio (SNR) optimization control mechanism and clarifies the specific algorithm parameters for feedback adjustment; the system calculates the SNR of a single-channel signal in real time. Signal power Defined as the average energy of all baseband signals across all bands after the first-stage demodulation:

[0117]

[0118] in, The sliding time window length is, for example, 100ms, and is set based on the premise that it is not less than 10 times the period of the lowest modulation frequency component of the system, to ensure the stability of the integral statistics and effectively suppress the interference of short-time impulse noise; noise level Then by analyzing the original signal Perform a Fast Fourier Transform to calculate higher than The square root of the power spectral density integral in the frequency band is obtained; if the calculated That is, the minimum system noise floor, then force To prevent division by zero errors;

[0119] To address the cold start problem and data singularity risk in the fitting process of quadratic polynomial models, and to ensure the physical validity of the model assumptions, this embodiment is based on the following physical mechanism: the system's signal-to-noise ratio is limited by the low-frequency band. The competition between flicker noise and the high-frequency device bandwidth roll-off effect makes... The relationship curve exhibits a unimodal convex function shape within the effective working interval, and therefore can be accurately approximated using a second-order Taylor expansion in the neighborhood of the extreme point.

[0120] Based on this, the system constructed a capacity of For example And this capacity must meet A first-in-first-out historical data buffer to ensure the degrees of freedom of least squares fitting. and execute the following data management and model update protocol:

[0121] Initialization scan: During system startup or reset, gradient optimization is temporarily suspended, and instead, a step size is forced. Perform a full-range scan within a preset frequency range, such as 1kHz-100kHz, to construct a system containing at least... Initial historical buffer for valid data points This ensures the non-singularity of the fitted matrix;

[0122] Singularity detection and perturbation injection: Calculate the buffer before each iteration update. Standard deviation of mid-frequency values ;like Right now Set to 50% of the current frequency control step size, and force a random disturbance. To expand data diversity; among them, For symbolic functions, A random number that is uniformly distributed within the interval 0 to 1;

[0123] Model Fitting and Convexity Verification: Based on Buffer The data is used to update the quadratic polynomial model in real time using the least squares method. ;like If the gradient is upward or approximately flat, it is considered unreliable and the gradient is forcibly set to zero. And triggers a large-step random jump;

[0124] Gradient Update and Boundary Clamping: Only when When calculating the model gradient The frequency is dynamically adjusted using the momentum gradient ascent algorithm.

[0125]

[0126]

[0127] The frequency at the next moment is calculated. Then, the system executes the filter coefficient synchronization update mechanism: recalculates the cutoff frequency of the first-stage low-pass filter. and the corresponding unit impulse response To eliminate transient oscillations in the digital filter caused by abrupt changes in filter coefficients, the system applies new signal power. Before the calculation, introduce a segment of length... The settling and waiting time is determined, and the demodulated output data within this time period is discarded;

[0128] in, In this embodiment, the momentum decay factor is fixed at 0.9, a value determined by the sampling delay of the control loop. With system response time constant ratio Sure, As an adaptive step size factor, to ensure the convergence of the algorithm near the vertex of the parabola, its value is designed to be a function related to the curvature of the model: ,in The base learning rate is the critical gain obtained through offline step response testing. For calibration, usually take ;

[0129] The specific steps for offline step response testing are as follows: Place the system in open-loop mode and set the polarization scan frequency. The center frequency; towards Injection amplitude is step disturbance Record the signal-to-noise ratio The response curve over time; if the response curve exhibits damped oscillations, the feedback gain value at the critical oscillation point is determined. If the system response does not oscillate, then directly... Set as to Empirical constants between them; , This is a regularization constant, with a value slightly larger than the minimum value of the machine's floating-point precision;

[0130] This is specifically introduced The saturation function serves as a physical boundary constraint condition, where such as 1kHz and For example, 500kHz is the physical linear operating range boundary of the electro-optic modulator; this step forces the iteration frequency to be constrained within the physical range allowed by the hardware, preventing frequency command overflow such as becoming negative or overclocking due to gradient explosion or numerical accumulation errors, thereby ensuring the robustness and safety of the closed-loop control system.

[0131] Bacterial samples are used as test media to induce dynamic changes in the optical field. In the step of generating a fingerprint to characterize the dynamic response of the modulated composite probe optical field, the generated fingerprint directly reflects the intensity relaxation correlation characteristics of each band component of the composite optical field under polarization scanning.

[0132] In this method, the bacterial sample is not only the object of observation, but also regarded as a random optical modulator. The dynamic response fingerprint in this embodiment directly reflects the intensity relaxation correlation characteristics of each band component of the composite light field after being modulated by the bacteria under polarization scanning. Specifically, each data point in the fingerprint represents the statistical residual memory after the photon exchanges momentum with the bacterial cell wall, cytoplasm and flagella.

[0133] This embodiment clarifies that the physical essence of the measurement is the dynamic parameters of the light field rather than purely biological parameters, and strictly defines the method in the field of radiation measurement. This definition emphasizes the physical process of light-matter interaction, which helps to distinguish it from purely biological detection methods in application and highlights the innovation of this technical solution at the physical optics level.

[0134] The specific steps for constructing a modulated composite probe optical field include: using multiple laser sources with different center wavelengths to generate initial beams, and applying independent intensity modulation to each beam; combining the intensity-modulated beams to form a composite beam; using an electro-optic modulator or an acousto-optic modulator to periodically scan and modulate the overall polarization state of the composite beam, and finally outputting the modulated composite probe optical field.

[0135] This embodiment provides a specific hardware implementation architecture for constructing a modulated light field; regarding the light source module, it adopts... Each laser diode outputs from a pigtail, and each diode is connected to an independent driving circuit. The circuit is loaded at a frequency of [frequency value missing]. A sinusoidal current modulation signal is used to achieve independent intensity modulation; multiple light beams are coupled into a single polarization-maintaining fiber using an optical fiber combiner; based on this, an electro-optic modulator (EOM) or an acousto-optic modulator (AOM) is connected as a polarization state scanning device; the voltage signal driving the EOM is a frequency... The sawtooth or sine wave causes the polarization state of the output beam to rotate and scan at high speed on the great circle of the Poincaré sphere, ultimately outputting a modulated composite probe light field.

[0136] This embodiment adopts an all-fiber or fiber-space hybrid optical path architecture and utilizes mature communication band modulation devices to achieve high-speed polarization control. Compared with the limitation of traditional mechanical rotating polarizers that can only reach hundreds of hertz, the EOM in this solution can reach the megahertz or even gigahertz level, thus meeting the high-frequency requirements for capturing the rapid Brownian motion of bacteria. At the same time, it has a compact structure and strong anti-interference ability, making it suitable for integrated applications.

[0137] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for acquiring optical fingerprints of bacterial samples using multi-band polarization time-series radiometric measurements, characterized in that, The method aims to accurately measure the dynamic parameters of a sample-modulated optical field, comprising: constructing a modulated composite probe optical field, wherein the modulated composite probe optical field contains multiple spectral band components with independent intensity modulation frequencies, and the entire optical field is loaded with a uniform polarization state scanning modulation frequency; applying the modulated composite probe optical field to a bacterial sample to be tested, and measuring the single-channel time-series response signal of the composite optical field returned or transmitted after being processed by the sample; performing frequency domain decoupling processing on the single-channel time-series response signal to extract the polarization intensity time-series change sequence corresponding to each independent spectral band component constituting the composite optical field; performing time-domain correlation analysis on the polarization intensity time-series change sequence to calculate a set of relaxation feature parameters characterizing the dynamic depolarization characteristics of the composite probe optical field; and generating a dynamic response fingerprint characterizing the modulated composite probe optical field under the action of a specific sample based on the set of relaxation feature parameters and the wavelengths of each spectral band component.

2. The method according to claim 1, characterized in that, The construction of the modulated composite probe light field includes: setting the intensity modulation frequency and the polarization state scanning modulation frequency according to a preset frequency domain layering isolation principle; wherein, the frequency domain layering isolation principle ensures that each intensity modulation frequency is completely separated from each other and from the polarization state scanning modulation frequency in the frequency domain, and that all modulation frequencies are greater than the physical perturbation characteristic frequency introduced by the sample.

3. The method according to claim 1, characterized in that, The frequency domain decoupling processing of the single-channel timing response signal includes: performing a first-level digital phase-locked demodulation on the single-channel timing response signal based on preset intensity modulation frequencies to separate and output the baseband signal corresponding to each spectral band; and performing a second-level sideband spectrum analysis on each baseband signal based on the polarization state scanning modulation frequency to resolve the temporal variation sequence of light intensity with polarization state scanning in that spectral band.

4. The method according to claim 1, characterized in that, The temporal correlation analysis of the polarization intensity time series includes: calculating the autocorrelation function of the polarization intensity time series for each spectral band; fitting the autocorrelation function curve to extract the decorrelation time constant of the optical field signal in that band; and the relaxation feature parameter set includes at least a set composed of the decorrelation time constants of each band.

5. The method according to claim 4, characterized in that, The method further includes: selecting the time-series change sequences of polarization intensity corresponding to different spectral bands, and calculating the cross-correlation function between each pair; calculating the inter-band coupling coefficient of the optical field depolarization process based on the signal lag or lead relationship between the bands reflected by the peak value of the cross-correlation function; and incorporating the inter-band coupling coefficient as an additional parameter into the relaxation characteristic parameter set.

6. The method according to claim 1, characterized in that, Generating the dynamic response fingerprint includes: constructing a multidimensional feature space with optical field parameters as coordinates, the dimensions of which include at least the radiation wavelength dimension, the polarization state change dimension, and the time dynamic dimension; mapping the relaxation feature parameter set to the multidimensional feature space to form a dynamic optical field fingerprint corresponding to the current measurement conditions.

7. The method according to claim 1, characterized in that, The method further includes: real-time monitoring of the signal-to-noise ratio (SNR) of the single-channel timing response signal; establishing a correspondence model between the SNR and the polarization state scanning modulation frequency; and dynamically adjusting the polarization state scanning modulation frequency based on the model to ensure that the signal SNR is always maintained within the optimal measurement range.

8. The method according to claim 1, characterized in that, In the method, the bacterial sample serves as the test medium that causes dynamic changes in the light field; in the step of generating a fingerprint to characterize the dynamic response of the modulated composite probe light field, the generated fingerprint directly reflects the intensity relaxation correlation characteristics of each band component of the composite light field under polarization scanning.

9. The method according to claim 1, characterized in that, The construction of the modulated composite probe optical field includes: using multiple laser sources with different center wavelengths to generate initial beams, and applying independent intensity modulation to each beam; combining the intensity-modulated beams to form a composite beam; using an electro-optic modulator or an acousto-optic modulator to periodically scan and modulate the overall polarization state of the composite beam, and finally outputting the modulated composite probe optical field.

Images