Agricultural biomass characteristic intelligent detection method, system and equipment based on multi-frequency conductance coupling and medium

Abstract

The invention relates to an agricultural biomass characteristic intelligent detection method, system and device based on multi-frequency conductance coupling and a medium, and the method comprises the following steps: applying an electric excitation signal to an agricultural biomass sample, and synchronously collecting multi-physics coupling response data; based on the multi-physics field coupling response data, constructing a multi-scale equivalent circuit model comprising a resonance unit, a constant phase unit and a diffusion impedance unit; synchronously optimizing a resonance unit parameter, a constant phase unit parameter and an impedance unit parameter of the multi-scale equivalent circuit model through a Bayesian joint inversion algorithm to obtain an optimal parameter combination; and according to the optimal parameter combination, calculating a preliminary detection result of biomass components containing cellulose crystallinity, lignin crosslinking degree and water content, combining the collected environment temperature and humidity sensing data to implement dynamic compensation, and outputting a final detection result. Therefore, the detection precision, the environment robustness and the operation reliability are synergistically improved in three aspects of hardware design, model construction and algorithm optimization.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural biomass detection, and particularly relates to an intelligent detection method for agricultural biomass characteristics based on multi-frequency conductance coupling. BACKGROUND

[0002] In the field of agricultural biomass component detection, traditional methods mainly include chemical analysis and optical spectrum. The chemical analysis measures the contents of cellulose, lignin and the like through steps such as acidolysis, oxidation and chromatographic separation. Although the accuracy is high, it is difficult to apply to on-site rapid detection because the sample is destroyed and the time consumption is as long as several hours to several days. The optical spectrum (such as near-infrared spectrum) realizes nondestructive detection by establishing a regression model of spectral characteristics and components, but due to the complexity of biomass components and the significant overlapping of spectral absorption peaks, a large number of samples are needed for modeling and the model is sensitive to moisture content, which leads to insufficient generalization ability of the model, especially the error of high moisture content samples increases significantly.

[0003] In recent years, electrical detection technology has gradually become a research hotspot due to its rapid and nondestructive characteristics. The existing technology usually measures the impedance or dielectric response characteristics of biomass by applying a single frequency or narrow frequency band excitation signal to calculate the component content. However, due to the microscale structure of agricultural biomass (cellulose microfibril, lignin matrix, pore liquid, etc.), its electrical characteristics show complex coupling effect at different frequencies, and single frequency band detection cannot fully characterize the interaction of multiple physical fields.

[0004] In addition, in terms of hardware detection, the existing electrode is not suitable for the microstructure of biomass. Due to the anisotropy of biomass fiber and the randomness of pore distribution, the contact pressure distribution on the surface of the homogeneous electrode is uneven, which may cause local arc discharge or signal coupling distortion, especially under high voltage excitation conditions, which may cause sample carbonization and further damage the detection consistency. SUMMARY

[0005] (I) Technical problems to be solved

[0006] In view of the above shortcomings and deficiencies of the prior art, the present application provides an intelligent detection method for agricultural biomass characteristics based on multi-frequency conductance coupling, a system, a device and a medium, which solves the technical problems that the existing electrical detection method cannot effectively characterize the multi-physical field coupling effect of the multi-scale structure of agricultural biomass due to the use of single frequency or narrow frequency band excitation signal, and the homogeneous electrode design is not suitable for the microanisotropic structure of biomass, which leads to limited detection accuracy, sensitive environmental interference and unstable contact impedance.

[0007] (II) Technical solutions

[0008] In order to achieve the above purpose, the main technical solutions adopted by the present application include:

[0009] In a first aspect, the embodiments of the present application provide an intelligent detection method for agricultural biomass characteristics based on multi-frequency electric conductance coupling, comprising:

[0010] applying an electric excitation signal to the agricultural biomass sample through the interdigital electrode array with gradient pore structure, and synchronously collecting multi-physical field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution;

[0011] based on the multi-physical field coupling response data, constructing a multi-scale equivalent circuit model including a resonance unit representing the electromagnetic characteristics of cellulose microfibrils, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid;

[0012] synchronously inverting the resonance unit parameters, constant phase unit parameters and impedance unit parameters of the multi-scale equivalent circuit model through the Bayesian joint inversion algorithm to obtain the optimal parameter combination;

[0013] According to the optimal parameter combination, the preliminary detection results of the biomass components including cellulose crystallinity, lignin crosslinking degree and moisture content are calculated, and dynamic compensation is implemented combined with the collected environmental temperature and humidity sensing data to output the final detection results.

[0014] Optionally, applying an electric excitation signal to the agricultural biomass sample through the interdigital electrode array with gradient pore structure, and synchronously collecting multi-physical field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution comprises:

[0015] applying a composite electric excitation signal generated by orthogonal frequency division coupling of low-frequency polarization excitation component and high-frequency dielectric detection component to the agricultural biomass sample through the interdigital electrode array;

[0016] using the piezoelectric driving module integrated in the interdigital electrode array to real-time self-adaptive adjust the contact pressure of each interdigital electrode to compensate the sample surface deformation;

[0017] synchronously triggering the following three types of sensing units based on a unified clock source:

[0018] acquiring the frequency domain impedance spectrum in the frequency band of 0.1 Hz-50 MHz through the embedded spectrum analysis module;

[0019] capturing the time domain dielectric relaxation current through the current sensing module;

[0020] scanning the spatial electric field distribution through the microelectrode matrix scanning module;

[0021] The interdigital electrode surface is covered with a plurality of layers of gradient porous conductive layers, the porosity of the conductive layers continuously increases from the root of the electrode to the tip of the electrode along the length direction of the electrode finger, and the porosity arrangement direction is consistent with the main shaft direction of the biomass fiber, the porosity gradient direction is orthogonal to the maximum change direction of the electric field gradient in the interdigital gap area, forming a synergistic adaptive interface of electric field and porosity;

[0022] The low-frequency polarization excitation component is a sine sweep signal of 0.1 Hz-10 kHz, which is used to excite the cellulose interface polarization effect, and the high-frequency dielectric detection component is a pseudo-random phase modulation pulse sequence of 100 kHz-50 MHz, which is used to analyze the lignin and hemicellulose complex dielectric properties.

[0023] Optionally, the piezoelectric driving module is integrated inside the support substrate of the interdigital electrode array, including a plurality of independently controlled piezoelectric actuators, each piezoelectric actuator is rigidly connected with a single interdigital electrode, and the contact pressure of the electrode and the sample is adjusted by applying axial displacement;

[0024] The driving direction of the piezoelectric actuator is consistent with the normal direction of the electrode contact surface, and the displacement resolution is not less than 50 nm, and the dynamic adjustment frequency covers the biomass deformation fluctuation range of 0.1 Hz-100 Hz.

[0025] Optionally, based on the multi-physical field coupling response data, a multi-scale equivalent circuit model is constructed, which includes a resonance unit representing the electromagnetic properties of cellulose microfibrils, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid.

[0026] The amplitude-frequency characteristics of the frequency domain impedance spectrum are mapped to the RLC resonance unit representing the electromagnetic properties of cellulose microfibrils, and the resonance unit includes an inductance and capacitance coupling network, and the inductance value has a negative power law relationship with the crystallinity of cellulose.

[0027] According to the decay rate and oscillation characteristics of the time domain dielectric relaxation current, a constant phase unit is constructed to represent the interface polarization effect of lignin and hemicellulose, and the constant phase unit includes a pseudo-capacitance coefficient related to the polarization charge density and a fractional order dispersion index representing the interface disorder degree of lignin and hemicellulose.

[0028] Based on the gradient vector field of the spatial electric field distribution, a diffusion impedance unit is generated to describe the directional migration of pore liquid ions, and the diffusion impedance unit includes a chain topology structure related to the electric field gradient direction, and the weight of each branch in the chain topology structure is positively related to the local electric field gradient modulus, and the branch length corresponds to the ion migration path.

[0029] A mesoscale charge transfer interface is constructed by a distributed RC network, a resonant unit and a constant phase unit are connected by the mesoscale charge transfer interface, and a diffusion impedance unit is bridged by a diffusion migration admittance, thereby forming a multi-scale equivalent circuit model covering nanoscale fibers to millimeter-scale pores.

[0030] Optionally, the resonant unit parameters, the constant phase unit parameters and the impedance unit parameters of the multi-scale equivalent circuit model are simultaneously inverted by a Bayesian joint inversion algorithm to obtain an optimal parameter combination, including:

[0031] Based on the statistical laws of the dielectric properties and microstructure of biomass, prior probability distribution clusters of the parameters of the resonant unit, the constant phase unit and the diffusion impedance unit are respectively constructed;

[0032] The multi-physical field coupling response data are taken as observation variables, and the phase angle of the frequency domain impedance spectrum, the decay waveform of the time domain relaxation current, and the gradient vector field of the spatial electric field distribution are fused to construct a multi-physical field joint likelihood function;

[0033] The parameters of the multi-scale equivalent circuit model are simultaneously inverted by a Bayesian inference framework, and the probability density field of the parameter posterior distribution is iteratively updated under the constraint of the multi-physical field joint likelihood function and the prior probability distribution cluster;

[0034] Based on the convergence criterion, when the root mean square deviation between the measured impedance phase angle and the theoretical impedance phase angle output by the multi-scale equivalent circuit model is less than 1.5 degrees, the fitting residual energy of the time domain relaxation current decay waveform is lower than 3 times the background noise energy, and the average cosine similarity between the spatial electric field gradient direction and the branch trend of the diffusion impedance unit is greater than 0.9, the cross-scale optimal parameter combination is output.

[0035] Optionally, the preliminary detection results of the biomass components including cellulose crystallinity, lignin crosslinking degree and moisture content are calculated according to the optimal parameter combination, and dynamic compensation is implemented combined with the collected environmental temperature and humidity sensing data, and the final detection results are output, including:

[0036] Based on the inductance parameter of the resonant unit in the optimal parameter combination, the initial crystallinity value is calculated through the negative power law relationship between the inductance parameter and the cellulose crystallinity, the initial crosslinking degree value is analyzed according to the fractional order dispersion index of the constant phase unit and the lignin crosslinking degree calibration curve measured by infrared spectrum, and the initial moisture content value is inverted according to the branch weight distribution of the diffusion impedance unit and the branch weight and moisture content calibration curve established by combining the impedance test and the drying method, thereby generating the preliminary detection results of the biomass components;

[0037] The temperature sensor data integrated on the detection electrode are acquired in real time, the temperature deviation between the current temperature and the environmental reference temperature is calculated, and the initial crystallinity value is implemented temperature reverse compensation according to the temperature deviation data,

[0038] Real-time humidity change detected by the humidity sensor is obtained, a pre-stored adsorption hysteresis curve related to the biomass type is called, and the initial moisture content value is corrected according to the adsorption hysteresis based on the humidity change;

[0039] Based on the obtained lignin crosslinking degree, the pseudo-capacitance coefficient weight factor of the constant phase unit is dynamically adjusted;

[0040] The temperature-compensated crystallinity value, the humidity-corrected moisture content value, and the crosslinking degree value adjusted by the weight factor are assigned component weights using the entropy weight method, and the detection results are time-series smoothed using a sliding window filtering algorithm, and the biomass component quantitative detection results resistant to environmental interference are output.

[0041] Optionally, dynamically adjusting the pseudo-capacitance coefficient weight factor of the constant phase unit based on the obtained lignin crosslinking degree comprises:

[0042] If the lignin crosslinking degree is greater than 0.7, the pseudo-capacitance coefficient weight is increased by 15%;

[0043] If the lignin crosslinking degree is 0.4 to 0.7, the pseudo-capacitance coefficient weight is adjusted to 1+0.05x lignin crosslinking degree;

[0044] If the lignin crosslinking degree is less than or equal to 0.4, the original weight of the pseudo-capacitance coefficient is maintained.

[0045] In a second aspect, the embodiments of the present application provide an agricultural biomass characteristic intelligent detection system based on multi-frequency conductance coupling, comprising:

[0046] A data acquisition module is configured to apply an electric excitation signal to an agricultural biomass sample through an interdigital electrode array with a gradient pore structure, and synchronously acquire multi-physical field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current, and spatial electric field distribution;

[0047] A model construction module is configured to construct a multi-scale equivalent circuit model including a resonance unit representing the electromagnetic characteristics of cellulose microfibrils, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid, based on the multi-physical field coupling response data;

[0048] A parameter inversion module is configured to synchronously invert the resonance unit parameters, the constant phase unit parameters, and the impedance unit parameters of the multi-scale equivalent circuit model through a Bayesian joint inversion algorithm to obtain an optimal parameter combination;

[0049] A result output module is configured to calculate preliminary detection results of biomass components including cellulose crystallinity, lignin crosslinking degree, and moisture content based on the optimal parameter combination, and implement dynamic compensation combined with the acquired environmental temperature and humidity sensing data to output final detection results.

[0050] In a third aspect, an embodiment of the present application provides a multi-frequency electric conductance coupling-based intelligent detection device for agricultural biomass characteristics, comprising: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the multi-frequency electric conductance coupling-based intelligent detection method for agricultural biomass characteristics as described above.

[0051] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, and the computer-readable storage medium stores computer-executable instructions, wherein the executable instructions are executed by a processor to implement the multi-frequency electric conductance coupling-based intelligent detection method for agricultural biomass characteristics as described above.

[0052] (III) Beneficial Effects

[0053] The beneficial effects of the present application are: the present application introduces a gradient pore structure interdigital electrode array, which first solves the compatibility problem of traditional homogeneous electrodes and the anisotropic structure of biomass, and the gradient distribution of its porosity can adapt to the surface deformation of the sample, thereby significantly reducing the contact impedance fluctuation and improving the signal coupling stability; secondly, based on the synchronous collection of multi-physical field coupling response data, the wide frequency impedance spectrum, the polarization dynamics process of the time domain relaxation current and the ion migration path information of the spatial electric field distribution are effectively integrated, and the multi-scale electromagnetic coupling effect of biomass is comprehensively characterized; further, by constructing a multi-scale equivalent circuit model including a resonance unit, a constant phase unit and a diffusion impedance unit, the physical mechanisms of cellulose crystal region energy storage, lignin interface polarization relaxation and pore ion diffusion are accurately distinguished at the circuit topology level, and the detection deviation caused by the parameter confusion of the traditional single equivalent model is overcome; on this basis, the Bayesian joint inversion algorithm is used to optimize the cross-scale parameters synchronously, which breaks through the precision limitation caused by the hierarchical error accumulation of the traditional step-by-step inversion method, and ensures the global optimal matching of the parameters; finally, the detection results are dynamically compensated combined with the environmental sensing data, through the double mechanisms of temperature drift correction and humidity adsorption delay compensation, not only the detection drift caused by environmental fluctuations in the traditional method is suppressed, but also the adaptive compensation of moisture content detection under different humidity conditions is realized, so that the detection precision, environmental robustness and operation reliability are synergistically improved at the three levels of hardware design, model construction and algorithm optimization. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 The flowchart of the method provided by the embodiment of the present application is shown in the figure;

[0055] Figure 2 The specific flowchart of step S1 of the method provided by the embodiment of the present application is shown in the figure;

[0056] Figure 3 A specific flowchart of step S2 of the method provided by the embodiment of the present application is shown in the following;

[0057] Figure 4 A specific flowchart of step S3 of the method provided by the embodiment of the present application is shown in the following;

[0058] Figure 5 A specific flowchart of step S4 of the method provided by the embodiment of the present application is shown in the following. DETAILED DESCRIPTION

[0059] In order to better explain the present application, so as to be understood, the present application is described in detail by specific embodiments in combination with the accompanying drawings.

[0060] The agricultural biomass characteristic intelligent detection method based on multi-frequency electric conductance coupling provided by the embodiment of the present application comprises: applying an electric excitation signal to an agricultural biomass sample through an interdigital electrode array with a gradient pore structure, and synchronously collecting multi-physical field coupling response data containing frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution; based on the multi-physical field coupling response data, a multi-scale equivalent circuit model containing a resonance unit representing the electromagnetic characteristics of cellulose microfibril, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid is constructed; the resonance unit parameters, the constant phase unit parameters and the impedance unit parameters of the multi-scale equivalent circuit model are synchronously inverted by a Bayesian joint inversion algorithm to obtain an optimal parameter combination; the preliminary detection results of the biomass components containing cellulose crystallinity, lignin crosslinking degree and moisture content are calculated according to the optimal parameter combination, and dynamic compensation is implemented in combination with the collected environmental temperature and humidity sensing data to output the final detection results.

[0061] The present application solves the compatibility problem of traditional homogeneous electrodes and the anisotropic structure of biomass by introducing a gradient-pore structure interdigital electrode array. The gradient distribution of its porosity can adapt to the surface deformation of the sample, thereby significantly reducing the contact impedance fluctuation and improving the signal coupling stability. Secondly, based on the synchronous acquisition of multi-physical field coupling response data, the wide frequency dielectric properties of the frequency domain impedance spectrum, the polarization dynamics process of the time domain relaxation current and the ion migration path information of the spatial electric field distribution are effectively integrated, realizing the comprehensive characterization of the multi-scale electromagnetic coupling effect of biomass. Further, by constructing a multi-scale equivalent circuit model including resonance units, constant phase units and diffusion impedance units, the physical mechanisms of cellulose crystal region energy storage, lignin interface polarization relaxation and pore ion diffusion are accurately distinguished at the circuit topology level, overcoming the detection bias caused by the parameter confusion of traditional single equivalent model. On this basis, the Bayesian joint inversion algorithm is used to optimize the cross-scale parameters, breaking through the precision limitation caused by the cumulative error of traditional step-by-step inversion method, and ensuring the global optimal matching of parameters. Finally, combined with environmental sensing data, the detection results are dynamically compensated, through the double mechanism of temperature drift correction and humidity adsorption delay compensation, not only inhibiting the detection drift caused by environmental fluctuations in traditional methods, but also realizing the adaptive compensation of moisture content detection under different humidity conditions, thereby synergistically improving the detection accuracy, environmental robustness and operation reliability at three levels of hardware design, model construction and algorithm optimization.

[0062] In order to better understand the above technical solutions, the exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer, more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0063] Specifically, the embodiment of the present application provides an intelligent detection method for agricultural biomass characteristics based on multi-frequency conductance coupling, comprising:

[0064] S1, applying an electric excitation signal to an agricultural biomass sample through an interdigital electrode array with a gradient-pore structure, and synchronously acquiring multi-physical field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution.

[0065] In the embodiments of the present application, the agricultural biomass samples include: corn stalk samples (moisture content 22%, fiber diameter 80-150 μm, component content including 38% cellulose, 15% lignin, 28% hemicellulose), rice straw samples (moisture content: 18%; fiber diameter: 50-120 μm; component content including 34% cellulose, 14% lignin, 25% hemicellulose), sugarcane residue samples (moisture content: 25%; fiber diameter: 100-200 μm; component content including 40% cellulose, 18% lignin, 28% hemicellulose), and the like.

[0066] Further, as shown in Figure 2 Step S1 includes:

[0067] S11, applying a composite electric excitation signal generated by coupling of a low-frequency polarization excitation component and a high-frequency dielectric detection component through an interdigital electrode array to the agricultural biomass sample. The surface of the interdigital electrode is covered with a multi-layer gradient-pore conductive layer, the porosity of the conductive layer continuously increases from the root to the tip of the electrode along the length direction of the electrode finger, the pore arrangement direction is consistent with the main axis of the biomass fiber, the porosity gradient direction is orthogonal to the maximum change direction of the electric field gradient in the interdigital gap region, and a synergistically adaptive interface of the electric field and the pore is formed.

[0068] In specific embodiments, the composite electric excitation signal generation step includes: generating a low-frequency polarization excitation component and a high-frequency dielectric detection component, the low-frequency polarization excitation component is a sine sweep signal of 0.1 Hz-10 kHz for exciting the cellulose interface polarization effect, and the high-frequency dielectric detection component is a pseudo-random phase modulation pulse sequence of 100 kHz-50 MHz for analyzing the lignin and hemicellulose composite dielectric properties.

[0069] Then, the low-frequency and high-frequency signals are coupled through orthogonal frequency division multiplexing technology to isolate the signal components in the frequency domain and form a composite excitation signal; a Hamming window function is applied to the coupled composite excitation signal to eliminate waveform mutations at the frequency band junctions and generate a smooth transition time-domain excitation waveform.

[0070] The composite excitation signal is applied by the interdigital electrode array with an axial gradient-pore structure on the surface. Specifically, a platinum / carbon nanotube composite conductive layer is deposited on the surface of the interdigital electrode to form a three-layer gradient-pore structure: bottom layer: porosity 35%±2%, pore size 50-100 nm, thickness 50 μm; middle layer: porosity 50%±3%, pore size 100-200 nm, thickness 80 μm; surface layer: porosity 65%±2%, pore size 200-300 nm, thickness 120 μm, the pore arrangement direction is parallel to the main axis of the straw fiber (included angle ≤5°) by laser micromachining, and the porosity gradient from the root to the tip of the electrode is 35%→65%.

[0071] Finally, the timing of the excitation application and response acquisition is matched based on the phase-locked loop synchronization mechanism to ensure the phase consistency of the time domain, frequency domain, and spatial domain data.

[0072] S12. A piezoelectric drive module integrated into the interdigital electrode array is used to adaptively adjust the contact pressure of each interdigital electrode in real time to compensate for sample surface deformation.

[0073] The piezoelectric drive module is integrated inside the support substrate of the interdigital electrode array and includes multiple independently controlled piezoelectric actuators. Each piezoelectric actuator is rigidly connected to a single interdigital electrode and adjusts the contact pressure between the electrode and the sample by applying axial displacement. The driving direction of the piezoelectric actuator is consistent with the normal direction of the electrode contact surface, and the displacement resolution is not less than 50nm. The dynamic adjustment frequency covers the biomass deformation fluctuation range of 0.1Hz-100Hz.

[0074] Specifically, the piezoelectric drive module is embedded within the insulating support substrate (such as alumina ceramic) of the electrode array, forming an integrated structure to avoid external mechanical interference. Each interdigital electrode is equipped with an independent piezoelectric actuator to achieve precise local pressure control. It also features axial drive, allowing the piezoelectric actuator to extend and retract along the normal direction of the electrode contact surface (displacement stroke ±50μm), directly adjusting the electrode-sample distance. The dynamic adjustment frequency covers the biomass deformation fluctuation range of 0.1Hz-100Hz, thus compensating for the slow moisture absorption and expansion of biomass at the low frequency of 0.1Hz and suppressing mechanical vibration interference at the high frequency of 100Hz.

[0075] S13. Based on a unified clock source, synchronously trigger the following three types of sensing units: obtain the frequency domain impedance spectrum of the 0.1Hz-50MHz band through the embedded spectrum analysis module; capture the time domain dielectric relaxation current through the current sensing module; and scan and generate the spatial electric field distribution through the microelectrode matrix scanning module.

[0076] In one embodiment, an embedded spectrum analysis module is located on the electrode array interface board, ≤5cm away from the electrode contacts. The embedded spectrum analysis module includes a high-speed ADC (16bit, 100MSPS) + digital downconverter (DDC) + FFT processor (4096 points); the current sensing module is directly connected in series in the electrode excitation circuit, including a zero flux current probe (bandwidth DC-100MHz, sensitivity 1mV / mA) + 24bit high dynamic range ADC (1MSPS); and the microelectrode matrix scanning module covers the sample surface and is arranged parallel to the electrode array, including a 16×16 tungsten microelectrode array (500μm spacing) + multiplexer switch (switching time <100ns).

[0077] It should be clear that the frequency domain impedance spectrum is a graph of the complex impedance obtained by measurement in a specific frequency range, which characterizes the resistance and capacitance coupling characteristics of the measured material under the action of an alternating electric field, including the response characteristics of the impedance amplitude and phase angle with frequency.

[0078] The time domain dielectric relaxation current is a transient current that decays over time after the excitation electric field is applied or removed due to the release of internal polarization charges in the material, reflecting the dynamic process and time delay characteristics of the dielectric polarization effect.

[0079] The spatial electric field distribution is a quantitative description of the electric field intensity vector (size and direction) at each spatial point on the surface or inside the measured sample, revealing the modulation effect of material microstructure heterogeneity on electromagnetic field distribution, which can be visualized in the form of two-dimensional / three-dimensional electric field gradient fields.

[0080] S2, based on the multi-physical field coupling response data, a multi-scale equivalent circuit model is constructed, which includes a resonance unit representing the electromagnetic characteristics of cellulose microfibrils, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid.

[0081] Further, as shown in Figure 3 , step S2 includes:

[0082] S21, map the amplitude-frequency characteristics of the frequency domain impedance spectrum to the RLC resonance unit representing the electromagnetic characteristics of cellulose microfibrils, the resonance unit includes an inductance and capacitance coupling network, and the inductance value is negatively related to the crystallinity of cellulose.

[0083] S22, according to the decay rate and oscillation characteristics of the time domain dielectric relaxation current, a constant phase unit representing the interface polarization effect of lignin and hemicellulose is constructed, the constant phase unit includes a pseudo-capacitance coefficient related to the polarization charge density and a fractional order dispersion index representing the interface disorder degree of lignin and hemicellulose.

[0084] S23, based on the gradient vector field of the spatial electric field distribution, a diffusion impedance unit describing the directional migration of pore liquid ions is generated, the diffusion impedance unit includes a chain topology structure related to the direction of the electric field gradient, and the weight of each branch in the chain topology structure is positively related to the local electric field gradient modulus, and the branch length corresponds to the ion migration path.

[0085] S24, build mesoscale charge transfer interface through distributed RC network, connect resonant unit and constant phase unit by mesoscale charge transfer interface, and bridge diffusion impedance unit via diffusion migration admittance, to form a multi-scale equivalent circuit model covering nanoscale fibers to millimeter scale pores. The mesoscale charge transfer interface is built through a distributed RC network, wherein the resistance component represents the ion migration impedance of the cellulose-lignin interface, and the capacitance component quantifies the double-layer charge storage effect of the interface; and the diffusion impedance unit is bridged by a migration path weighted admittance, and the admittance value is positively correlated with the cosine value of the angle between the electric field gradient direction.

[0086] In a specific embodiment, the characteristic frequency (such as the inflection point frequency f r ) is extracted from the frequency domain impedance spectrum, which is mapped to the RLC resonance frequency:

[0087]

[0088] wherein L cell is the inductance parameter, C cell is the capacitance parameter, such as the impedance phase peak of corn straw at 10 kHz, corresponding to the resonant unit parameters L cell = 22 μH, C cell = 15 pF.

[0089] The inductance is calibrated to be negatively related to the crystallinity C r of cellulose:

[0090]

[0091] wherein k1 is calibrated by XRD, and the value range is 13.3 to 16.3 μH.

[0092] Then, the relaxation current decay curve (I0 represents the initial value of the relaxation current at t = 0, τ is the time required for the current to decay to 1 / e of the initial value, and 0 < β ≤ 1) is fitted, and the non-exponential factor β (reflecting the polarization heterogeneity, the smaller the β value, the wider the relaxation time distribution, and the more disordered the polarization interface structure, such as corn straw β = 0.78 and sugarcane residue β = 0.62) is extracted; further, the fractional order diffusion index n = 1-β (quantifying the interface disorder degree) is obtained, and the pseudo-capacitance coefficient Q is related to the polarization charge density: Q ∝ ∈0∈ r A / d0, (∈0 is the vacuum permittivity; ∈ r : the relative dielectric constant of the interface region, the typical value of biomass is 2-6; A: the polarization interface area, d0: the polarization charge layer spacing).

[0093] Then, the sample surface two-dimensional potential distribution data of the potential distribution V(x, y) is obtained by microelectrode matrix scanning, and the electrostatic potential of each space point is characterized. The potential value of each point (x, y) reflects the local electrochemical environment, and the electric field gradient vector is calculated Branch weight The contribution weight of each branch in the diffusion impedance chain is characterized, and is positively correlated with the local electric field gradient modulus. The greater the weight, the more significant the contribution of the path to ion migration. For the corn straw sample, 0.15≤w i ≤0.38, the branch length l i =τ2·dz (τ2: bending degree, dz: straight line distance) corresponds to the effective distance of the ion migration path, and the average pore bending degree τ2 of the corn straw is 1.5 and the straight line distance dz is 67 μm.

[0094] Therefore, in the embodiment, the cross-scale coupling from micro (cellulose nanofiber) to meso (lignin-hemicellulose interface) to macro (pore millimeter-level system) is realized, which is accurately matched with the hierarchical structure of biomass.

[0095] S3, the parameters of the resonance unit, the constant phase unit and the impedance unit of the multi-scale equivalent circuit model are simultaneously inverted by the Bayesian joint inversion algorithm to obtain the optimal parameter combination.

[0096] Further, as shown in Figure 4 , step S3 includes:

[0097] S31, based on the dielectric properties and the statistical rules of the microstructure of biomass, the prior probability distribution clusters of the parameters of the resonance unit, the constant phase unit and the diffusion impedance unit are respectively constructed.

[0098] In this step, based on the cellulose crystallinity data obtained by X-ray diffraction (XRD), the statistical distribution of the inductance parameters of the resonance unit is established, such as the inductance value decreases in negative power law with the increase of the crystallinity. The lignin crosslinking degree measured by infrared spectrum (FTIR) is used to construct the truncated Gaussian distribution of the pseudo-capacitance coefficient and the dispersion index of the constant phase unit; and the pore fractal dimension obtained by micro-CT scanning is used to generate the logarithmic normal distribution of the diffusion impedance branch weight. Further, the prior probability distribution clusters of the parameters of the resonance unit, the constant phase unit and the diffusion impedance unit are respectively constructed. Thus, the physical properties (such as fiber orientation and pore complexity) of biomass are quantified as parameter value ranges in the embodiment of the application, the search space is reduced, the rationality of the initial guess of the parameters is improved, and the number of iterations is reduced.

[0099] S32, the multi-physical field coupling response data is taken as an observation variable, the phase angle of the frequency domain impedance spectrum, the decay waveform of the time domain relaxation dielectric relaxation current and the gradient vector field of the spatial electric field distribution are fused to construct a multi-physical field joint likelihood function.

[0100] Specifically, the multi-dimensional likelihood function is constructed with the error of real and imaginary parts of frequency domain impedance, the residual of time domain relaxation current, and the difference of spatial electric field gradient as observation variables:

[0101]

[0102] where M i is the model prediction value of the ith physical field, D i is the observation variable, i = 1: real and imaginary parts of frequency domain impedance, i = 2: instantaneous value of time domain relaxation current, i = 3: modulus of spatial electric field gradient vector, σ i is the measurement uncertainty weight of each field data, specifically, it is the impedance measurement error (instrument accuracy) in the frequency domain, the current noise level in the time domain, and the electric field gradient sampling error in the space, and ∏ is the joint action of the three physical field likelihood functions.

[0103] S33, synchronously invert each unit parameter of the multi-scale equivalent circuit model through the Bayesian inference framework, and iteratively update the probability density field of the parameter posterior distribution under the constraint of the multi-physical field joint likelihood function and the prior probability distribution cluster.

[0104] S34, based on the convergence criterion, when the root mean square deviation of the measured impedance phase angle and the theoretical impedance phase angle output by the multi-scale equivalent circuit model is less than 1.5 degrees, the fitting residual energy of the time domain relaxation current decay waveform is lower than 3 times the background noise energy, and the average cosine similarity between the spatial electric field gradient direction and the branch trend of the diffusion impedance unit is greater than 0.9, the optimal parameter combination across scales is output.

[0105] S4, calculate the preliminary detection results of the biomass components including cellulose crystallinity, lignin crosslinking degree and water content according to the optimal parameter combination, and implement dynamic compensation combined with the collected environmental temperature and humidity sensing data, and output the final detection results.

[0106] Further, as shown in Figure 5 , step S4 includes:

[0107] S41, based on the inductance parameter of the resonant unit in the optimal parameter combination, calculate the initial crystallinity value through the negative power law relationship between the inductance parameter and the cellulose crystallinity, according to the fractional order diffusion index of the constant phase unit and the lignin crosslinking degree calibration curve measured by infrared spectrum, analyze the initial crosslinking degree value, and according to the branch weight distribution of the diffusion impedance unit and the branch weight and water content calibration curve established by combining the drying method and impedance test, invert the initial water content value, to generate the preliminary detection results of the biomass components.

[0108] Wherein, the lignin cross-linking degree calibration curve is a quantitative relationship model between the characteristic peak area of lignin measured by infrared spectrum (FTIR) and the fractional order diffusion index (n) of constant phase unit, which represents the correlation between interface polarization effect and chemical cross-linking degree. The linear relationship between different cross-linking degrees Lx and fractional order diffusion index n is fitted, L x = k2· n + k3 (k2 = 1.22, k3 = -0.15).

[0109] The branch weight-moisture content calibration curve is a mapping relationship model between the branch weight (w i ) of diffusion impedance unit and the real moisture content (M w ) of biomass sample established by combining drying method and impedance test. The specific implementation steps are as follows: preparing standardized biomass samples (corn stalks, cotton stalks, etc.) with different moisture contents (5% to 60%), measuring the real moisture content M w of each group of samples by oven method, applying a wide frequency excitation of 10 Hz-10 MHz, and measuring the 16-level branch weight w1, w2,..., w 16 of diffusion impedance unit. The weight average value The relationship between w and Mw is established as follows:

[0110] S42, real-time acquisition of temperature sensor data integrated on the detection electrode, calculation of temperature deviation of current temperature and environmental reference temperature, and implementation of temperature reverse compensation on the initial crystallinity value according to the temperature deviation data.

[0111] Specifically, the temperature sensor is a micro platinum resistance temperature probe embedded in the electrode substrate to realize millisecond-level response detection of the temperature at the interface between the electrode and the sample. Real-time acquisition of temperature sensor data integrated on the detection electrode, calculation of the deviation ΔT of current temperature and environmental reference temperature (25℃), and reverse compensation of the initial crystallinity value according to the negative temperature coefficient characteristic of cellulose dielectric constant: for every 1℃ increase in temperature, the crystallinity compensation increases by 0.015%; for every 1℃ decrease in temperature, the compensation decreases by 0.015%.

[0112] S43, real-time acquisition of humidity change amount detected by humidity sensor, calling of pre-stored biomass type related adsorption hysteresis curve, and implementation of adsorption hysteresis correction on the initial moisture content value according to the humidity change amount.

[0113] Specifically, the humidity sensor adopts a MEMS capacitive humidity meter integrated in the negative pressure air extraction channel of the sample table to monitor the environmental humidity fluctuation in real time. The humidity hysteresis correction of the initial moisture content value is based on the pre-stored adsorption-desorption curves of different biomass types to dynamically adjust the correction amount. Based on the environmental humidity change amount ΔH detected by the humidity sensor, the pre-stored adsorption hysteresis curve related to the biomass type is called, and the initial moisture content value is dynamically corrected. When the environmental humidity rises, the correction amount ΔM a = +0.1% x ΔH x D 2 (D is the pore fractal dimension); when the humidity decreases, ΔM a = -0.08% x ΔH x D 2 .

[0114] S44, dynamically adjusting the pseudo-capacitance coefficient weight factor of the constant phase unit based on the analyzed lignin crosslinking degree. Among them, dynamically adjusting the pseudo-capacitance coefficient weight factor of the constant phase unit based on the analyzed lignin crosslinking degree includes: if the lignin crosslinking degree > 0.7, the pseudo-capacitance coefficient weight is increased by 15%; if 0.4 < lignin crosslinking degree ≤ 0.7, the pseudo-capacitance coefficient weight is adjusted to 1 + 0.05 x lignin crosslinking degree; if the lignin crosslinking degree ≤ 0.4, the original weight of the pseudo-capacitance coefficient is maintained.

[0115] S45, the crystallinity value after temperature compensation, the moisture content value after humidity correction, and the crosslinking degree value after weight factor adjustment are distributed with component weight by using the entropy weight method, and the detection result is time-series smoothed by using the sliding window filtering algorithm, and the biomass component quantitative detection result resistant to environmental interference is output.

[0116] The compensated and corrected crystallinity, moisture content, and weight-adjusted crosslinking degree are input into a weighted fusion algorithm, and the weight distribution ratio is: cellulose crystallinity 48% ± 3%, lignin crosslinking degree 32% ± 2%, and moisture content 20% ± 1%; the detection result is time-series smoothed by using the sliding window filtering algorithm, and the stabilized final detection value is output.

[0117] Additionally, the embodiment of the present application provides an agricultural biomass characteristic intelligent detection system based on multi-frequency conductance coupling, comprising:

[0118] A data acquisition module is configured to apply an electric excitation signal to an agricultural biomass sample through an interdigital electrode array with a gradient pore structure, and synchronously acquire multi-physical field coupling response data including a frequency domain impedance spectrum, a time domain dielectric relaxation current, and a spatial electric field distribution.

[0119] A model construction module is configured to construct a multi-scale equivalent circuit model including a resonance unit representing the electromagnetic characteristics of cellulose microfibrils, a constant phase unit representing the interface polarization effect of lignin and hemicellulose, and a diffusion impedance unit representing the ion migration of pore liquid based on the multi-physical field coupling response data.

[0120] A parameter inversion module is configured to synchronize optimization of the resonance unit parameters, the constant phase unit parameters and the impedance unit parameters of the multi-scale equivalent circuit model by a Bayesian joint inversion algorithm to obtain an optimal parameter combination.

[0121] A result output module is configured to calculate a preliminary detection result of the biomass components including cellulose crystallinity, lignin crosslinking degree and water content according to the optimal parameter combination, and output a final detection result by combining the collected environmental temperature and humidity sensing data for dynamic compensation.

[0122] Meanwhile, the embodiment of the present application provides an agricultural biomass characteristic intelligent detection device based on multi-frequency conductance coupling, comprising: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the agricultural biomass characteristic intelligent detection method based on multi-frequency conductance coupling as described above.

[0123] Furthermore, the embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores computer executable instructions, and the executable instructions are executed by the processor to realize the agricultural biomass characteristic intelligent detection method based on multi-frequency conductance coupling as described above.

[0124] Since the system / device described in the above embodiment of the present application is the system / device used for implementing the method of the above embodiment of the present application, the specific structure and modification of the system / device can be understood by those skilled in the art based on the method described in the above embodiment of the present application, and thus will not be described here. Any system / device used for the method of the above embodiment of the present application belongs to the scope of the present application.

[0125] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can be in the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0126] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system) and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions.

[0127] It should be noted that the description using "comprises" or "comprising" does not exclude the presence of elements or steps other than those listed in a claim. The words "a" or "an" preceding the disclosure of a plurality of elements or steps do not exclude the presence of a plurality of such elements or steps. The application can be implemented by means of both hardware and software, and any combination thereof. In a claim reciting a means, the term "means" is intended to refer to a combination of the elements in the claim, and is not intended to refer to a specific structure or composition of the combination. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage. The word "comprising" does not exclude not excluding other elements or steps. The words "first", "second", "third", and the like in the description do not necessarily have an ordinal meaning. These words are used to distinguish between similar elements or steps. The implementation of any of the measuring methods described can be performed by means of any appropriate techniques.

[0128] Furthermore, it is noted that the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments or examples. It will also be readily understood to those skilled in the art that the above description is merely illustrative of the application and should not be

[0129] While the application has been described with respect to preferred embodiments, those skilled in the art will readily appreciate that various modifications and substitutions can be made hereto without departing from the spirit and scope of the application. Accordingly, it is intended that all such alterations and modifications be considered as falling within the scope of the application as defined by the claims and their equivalents.

[0130] Obviously, many modifications and variations of the present application are possible in light of the above teachings. It is, therefore, to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A smart detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling, characterized in that, include: An electrical excitation signal was applied to an agricultural biomass sample by an interdigitated electrode array with a gradient pore structure, and multi-physics field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution were simultaneously acquired. Based on multi-physics field coupling response data, a multi-scale equivalent circuit model is constructed, which includes a resonant unit characterizing the electromagnetic properties of cellulose microfibrils, a constant phase unit characterizing the polarization effect at the lignin-hemicellulose interface, and a diffusion impedance unit characterizing the migration of pore liquid ions. The optimal parameter combination is obtained by simultaneously inverting the resonant unit parameters, constant phase unit parameters, and impedance unit parameters of the multi-scale equivalent circuit model using the Bayesian joint inversion algorithm. Preliminary test results of biomass components, including cellulose crystallinity, lignin crosslinking degree and moisture content, are calculated based on the optimal parameter combination. Dynamic compensation is then implemented by combining the collected environmental temperature and humidity sensor data to output the final test results.

2. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in claim 1, characterized in that, An electrical excitation signal was applied to an agricultural biomass sample using an interdigitated electrode array with a gradient porosity structure. Simultaneously, multi-physics coupled response data, including frequency domain impedance spectra, time-domain dielectric relaxation current, and spatial electric field distribution, were acquired. A composite electrical excitation signal, generated by orthogonal frequency division coupling of a low-frequency polarization excitation component and a high-frequency dielectric detection component, is applied to agricultural biomass samples via an interdigitated electrode array. A piezoelectric drive module integrated into the interdigital electrode array is used to adaptively adjust the contact pressure of each interdigital electrode in real time to compensate for sample surface deformation. The following three types of sensing units are synchronously triggered based on a unified clock source: The frequency domain impedance spectrum of the 0.1Hz-50MHz band was obtained using an embedded spectrum analysis module; The time-domain dielectric relaxation current is captured by a current sensing module. The spatial electric field distribution is generated by scanning using a microelectrode matrix scanning module; Among them, the surface of the interdigitated electrode is covered with a multi-layer gradient porous conductive layer. The porosity of the conductive layer increases continuously from the root of the electrode to the tip of the electrode along the length of the electrode finger, and the pore arrangement direction is consistent with the direction of the main axis of the biomass fiber. The porosity gradient direction is orthogonal to the direction of the maximum change of the electric field gradient in the interdigitated gap region, forming a cooperative adaptation interface between the electric field and the pores. A sinusoidal sweep signal with a low-frequency polarization excitation component of 0.1 Hz to 10 kHz was used to excite the polarization effect at the cellulose interface, while a pseudo-random phase-modulated pulse sequence with a high-frequency dielectric detection component of 100 kHz to 50 MHz was used to analyze the dielectric properties of the lignin-hemicellulose composite.

3. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in claim 2, characterized in that, The piezoelectric drive module is integrated inside the support substrate of the interdigital electrode array and includes multiple independently controlled piezoelectric actuators. Each piezoelectric actuator is rigidly connected to a single interdigital electrode and the contact pressure between the electrode and the sample is adjusted by applying axial displacement. Among them, the driving direction of the piezoelectric actuator is consistent with the normal direction of the electrode contact surface, and the displacement resolution is not less than 50nm, and the dynamic adjustment frequency covers the range of biomass deformation fluctuations from 0.1Hz to 100Hz.

4. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in claim 1, characterized in that, Based on multi-physics coupling response data, a multi-scale equivalent circuit model was constructed, comprising a resonant unit characterizing the electromagnetic properties of cellulose microfibrils, a constant-phase unit characterizing the polarization effect at the lignin-hemicellulose interface, and a diffusion impedance unit characterizing the migration of ions in the porous liquid. The amplitude-frequency characteristics of the frequency domain impedance spectrum are mapped to an RLC resonant unit that characterizes the electromagnetic properties of cellulose microfibrils. The resonant unit contains an inductive and capacitive coupling network, and the inductance value has a negative power-law relationship with the crystallinity of cellulose. Based on the decay rate and oscillation characteristics of the time-domain dielectric relaxation current, a constant-phase unit is constructed to characterize the polarization effect at the lignin-hemicellulose interface. The constant-phase unit includes a pseudo-capacitance coefficient that correlates the polarization charge density and a fractional-order dispersion index that characterizes the disorder at the lignin-hemicellulose interface. Based on the gradient vector field of the spatial electric field distribution, a diffusion impedance unit describing the directional migration of ions in porous liquid is generated. The diffusion impedance unit contains a chain topology structure associated with the direction of the electric field gradient. The weight of each branch in the chain topology structure is positively correlated with the local electric field gradient modulus, and the branch length corresponds to the ion migration path. A mesoscale charge transfer interface is constructed by a distributed RC network. The mesoscale charge transfer interface is used to connect the resonant unit and the constant phase unit. The diffusion impedance unit is then bridged by the diffusion migration admittance to form a multi-scale equivalent circuit model covering nanofibers to millimeter-scale pores.

5. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in claim 1, characterized in that, The optimal parameter combinations are obtained by simultaneously inverting the resonant element parameters, constant phase element parameters, and impedance element parameters of the multi-scale equivalent circuit model using a Bayesian joint inversion algorithm. Based on the dielectric properties and statistical laws of biomass microstructure, prior probability distribution clusters of parameters covering resonant units, constant phase units, and diffusion impedance units are constructed respectively. Multi-physics field coupled response data are used as observation variables, and the phase angle of the frequency domain impedance spectrum, the decay waveform of the time domain relaxation dielectric relaxation current, and the gradient vector field of the spatial electric field distribution are integrated to construct a multi-physics field joint likelihood function. The parameters of each unit of the multi-scale equivalent circuit model are synchronously inverted through the Bayesian inference framework. The probability density field of the posterior distribution of the parameters is iteratively updated with the joint likelihood function of the multi-physics field and the prior probability distribution cluster as constraints. Based on the convergence criterion, the optimal cross-scale parameter combination is output when the root mean square deviation between the measured impedance phase angle and the theoretical impedance phase angle output by the multi-scale equivalent circuit model is less than 1.5 degrees, the residual energy of the time-domain relaxation current decay waveform fitting is less than 3 times the background noise energy, and the mean cosine similarity between the spatial electric field gradient direction and the branch direction of the diffuse impedance unit is greater than 0.

9.

6. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in any one of claims 1-5, characterized in that, Preliminary detection results of biomass components, including cellulose crystallinity, lignin crosslinking degree, and moisture content, are calculated based on the optimal parameter combination. Dynamic compensation is then implemented using collected environmental temperature and humidity sensor data to output the final detection results, including: Based on the inductance parameters of the resonant unit in the optimal parameter combination, the initial crystallinity value is calculated through the negative power law relationship between the inductance parameters and the crystallinity of cellulose. The initial crosslinking degree value is analyzed based on the fractional diffusion index of the constant phase unit and the lignin crosslinking degree calibration curve measured by infrared spectroscopy. The initial moisture content value is inverted based on the branch weight distribution of the diffusion impedance unit and the branch weight and moisture content calibration curve established by the drying method combined with impedance testing. Preliminary detection results of biomass components are generated. The system acquires real-time temperature sensor data integrated on the detection electrode, calculates the temperature deviation between the current temperature and the ambient reference temperature, and performs reverse temperature compensation on the initial crystallinity value based on the temperature deviation data. The humidity change detected by the humidity sensor is acquired in real time, and the pre-stored adsorption hysteresis curves related to the biomass type are called up to correct the initial moisture content value based on the humidity change. The pseudo-capacitance coefficient weighting factor of the constant phase unit is dynamically adjusted based on the lignin crosslinking degree obtained from the analysis. The crystallinity value after temperature compensation, the moisture content value after humidity correction, and the crosslinking degree value after weight factor adjustment are used to allocate component weights using the entropy weight method. The detection results are then smoothed over time using a sliding window filtering algorithm to output quantitative detection results of biomass components that are resistant to environmental interference.

7. The intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in claim 6, characterized in that, The weighting factors for the pseudo-capacitance coefficient of the constant-phase unit, dynamically adjusted based on the lignin crosslinking degree obtained from analysis, include: If the degree of lignin crosslinking is >0.7, the weight of the pseudo capacitance coefficient is increased by 15%; If 0.4 < lignin crosslinking degree ≤ 0.7, the weight of the pseudo-capacitance coefficient is adjusted to 1 + 0.05 × lignin crosslinking degree; If the degree of lignin crosslinking is ≤0.4, the original weight of the pseudo-capacitance coefficient is maintained.

8. An intelligent detection system for agricultural biomass characteristics based on multi-frequency conductivity coupling, characterized in that, include: The data acquisition module is used to apply an electrical excitation signal to agricultural biomass samples through an interdigitated electrode array with a gradient pore structure, and simultaneously acquire multi-physics field coupling response data including frequency domain impedance spectrum, time domain dielectric relaxation current and spatial electric field distribution. The model building module is used to construct a multi-scale equivalent circuit model based on multi-physics field coupling response data, which includes a resonant unit characterizing the electromagnetic properties of cellulose microfibrils, a constant phase unit characterizing the polarization effect at the lignin-hemicellulose interface, and a diffusion impedance unit characterizing the migration of pore liquid ions. The parameter inversion module is used to simultaneously invert the resonant unit parameters, constant phase unit parameters, and impedance unit parameters of the multi-scale equivalent circuit model through a Bayesian joint inversion algorithm, and obtain the optimal parameter combination. The results output module is used to calculate the preliminary detection results of biomass components, including cellulose crystallinity, lignin crosslinking degree and moisture content, based on the optimal parameter combination, and to perform dynamic compensation by combining the collected environmental temperature and humidity sensor data to output the final detection results.

9. An intelligent detection device for agricultural biomass characteristics based on multi-frequency conductivity coupling, characterized in that, include: At least one processor; and memory that is communicatively connected to at least one processor; The memory stores instructions that can be executed by at least one processor, which are executed by at least one processor to enable the at least one processor to perform the intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in any one of claims 1-7.

10. A computer-readable storage medium storing computer-executable instructions thereon, characterized in that, When the executable instructions are executed by the processor, they implement the intelligent detection method for agricultural biomass characteristics based on multi-frequency conductivity coupling as described in any one of claims 1-7.

Images