A method for detecting physicochemical indexes in food can production

Abstract

The application discloses a physicochemical index detection method for food can production and relates to the technical field of food detection. The method comprises the following steps: deploying a near-infrared spectrum, eddy current electromagnetic and biological impedance three-mode sensing unit in a ring arrangement, and non-contact collecting can body reflection spectrum, eddy current loss and complex impedance response signals; extracting sulfur dioxide, heavy metal and microbial characteristic parameters through wavelet packet decomposition, Hilbert transform and principal component analysis; constructing a three-element coupling correction equation group to cross correct original data and output accurate detection results; and integrating a self-calibration mechanism and an abnormal product automatic sorting function. The application realizes nondestructive, efficient and multi-index parallel detection by constructing four core technical systems, namely, multi-modal synchronous acquisition, cross-modal feature coupling, dynamic threshold decision and online self-calibration, has stable precision and reduces costs by 60%, and meets the real-time quality control and traceability requirements of a full-automatic production line.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of food detection, in particular to a physicochemical index detection method for food can production. BACKGROUND

[0002] With the continuous improvement of the automation and intelligence level of the food industry, as an important processed food category, the quality control requirements in the production process of canned food are becoming increasingly strict, especially for the detection accuracy, efficiency and coverage dimension of physicochemical indexes. The current mainstream detection system is mostly built around a single index, relying on independent technical paths such as chemical analysis, electrochemical sensing or microbial culture. Although it has certain detection capability in specific parameters such as sulfur dioxide residue, heavy metal leaching or sealing performance, the overall architecture presents a highly fragmented characteristic, lacking a systematic design for multi-index collaborative detection. This fragmented detection mode leads to multiple pre-treatments, sub-packaging and instrument switching for samples, not only prolonging the detection period and increasing the risk of human error, but also making it difficult to support real-time quality feedback and closed-loop regulation requirements of the production line.

[0003] For the core detection direction of canned food physicochemical safety, the focus has gradually shifted to the simultaneous monitoring of sulfur dioxide preservative residues, lead and cadmium heavy metal migration and pathogenic microbial contamination levels. By building an integrated detection process, seamless connection from sample pretreatment to multi-parameter output is achieved, thereby ensuring the food safety bottom line while improving the production line detection throughput and decision response speed. However, existing technical solutions are generally limited by the physical isolation of detection principles, such as the insensitivity of optical detection methods to heavy metals, the difficulty of electrochemical methods to accommodate microbial activity determination, and the inability of gas chromatography to directly count live bacteria, leading to methodological conflicts and equipment compatibility bottlenecks in multi-index integration.

[0004] Existing technologies generally have structural defects such as single detection dimension, fragmented process, strong device dependency and high cost. Some technologies achieve rapid color development and quantification of sulfur dioxide through special devices, but cannot simultaneously obtain heavy metal data. Other technologies use modified electrodes to improve lead and cadmium detection sensitivity, but due to short electrode life and the need for frequent calibration, they are difficult to adapt to continuous production rhythm. Microbial detection still relies on culture methods, which take 24 to 72 hours, completely unable to meet the online quality control time requirement. When the production line produces thousands of products per hour, the traditional separate detection mode not only causes batch determination delay, but also causes missed detection risk due to sample representativeness decay. SUMMARY

[0005] The present application provides a physicochemical index detection method for food can production to solve the problems of destructive detection method, low detection efficiency, complex operation and high equipment maintenance cost in the prior art.

[0006] To solve the above technical problems, the present application provides the following technical solutions: a physicochemical index detection method for food can production, comprising the following steps:

[0007] Step S1, deploy a multi-modal sensing array at an online detection station of a food can production line, the multi-modal sensing array is composed of a near-infrared spectroscopy sensing unit, an eddy current electromagnetic sensing unit and a bioimpedance sensing unit, the three are arranged in a ring shape and coaxially aligned with the center axis of the to-be-detected can body;

[0008] Step S2, after starting the detection process, the near-infrared spectroscopy sensing unit emits a continuous modulated light beam with a wavelength range of 800nm to 2500nm to the surface of the can body, synchronously collects the reflected spectrum signal and extracts the absorption intensity parameter corresponding to the sulfur dioxide molecular vibration characteristic peak;

[0009] Step S3, the eddy current electromagnetic sensing unit applies an alternating magnetic field with a frequency of 50kHz to 500kHz, the inductor receives the eddy current loss signal caused by the metal ions in the can body, and the comprehensive concentration index of cadmium, lead and mercury heavy metal ions is calculated according to the phase shift and amplitude attenuation rate;

[0010] Step S4, the bioimpedance sensing unit applies a sinusoidal excitation current with an amplitude of 0.5V and a frequency of 1kHz to 10kHz, measures the complex impedance response curve of the can body content in the time domain, and determines the microbial pollution grade according to the curve inflection point slope and relaxation time constant;

[0011] Step S5, input the absorption intensity parameter, comprehensive concentration index and pollution grade determination result into the intelligent analysis engine, execute the time-space alignment preprocessing, start the feature extraction submodule, and generate the sulfur dioxide feature vector, heavy metal feature matrix and microbial feature tensor respectively;

[0012] Step S6, construct a three-element coupling equation group containing sulfur dioxide-heavy metal inhibition factor, heavy metal-microbial synergistic factor and microbial-sulfur dioxide feedback factor, and solve the correction coefficient of each index by iterative least squares method;

[0013] Step S7, output the coupled and corrected sulfur dioxide content value, heavy metal comprehensive concentration index and microbial pollution grade determination result.

[0014] The near-infrared spectroscopy sensing unit includes a light source module, a light splitting module, a detector module and a temperature control compensation module; the light source module uses a halogen tungsten lamp as a wide-spectrum radiation source, the output light beam is focused by a gold-coated concave mirror and then passes through a rotating grating light splitting module to realize wavelength scanning; the detector module selects an indium gallium arsenide photodiode array.

[0015] The temperature control compensation module stabilizes the working temperature of the detector at 25℃±0.5℃ by a thermoelectric cooling sheet.

[0016] The eddy current electromagnetic sensing unit comprises an excitation coil group, a differential receiving coil group, a lock-in amplifier circuit and a digital phase detector; the differential receiving coil group adopts a common-mode rejection structure to eliminate environmental electromagnetic interference.

[0017] The excitation coil group is composed of 3 layers of tightly wound copper wires with an outer diameter of 50mm and 300 turns; the lock-in amplifier circuit outputs a direct current component after coherent demodulation of the input signal and the reference signal; the digital phase detector calculates the phase difference value according to the direct current component.

[0018] The bioimpedance sensing unit comprises a constant current source circuit, a four-electrode probe array, a high-speed analog-to-digital converter and an impedance spectrum analysis module; the distance between the two excitation electrodes in the four-electrode probe array is 30mm, and the two detection electrodes are located inside the excitation electrodes with a distance of 10mm; the sampling rate of the high-speed analog-to-digital converter is 100kHz; the impedance spectrum analysis module uses the Cole-Cole model to fit the measured data and extracts the characteristic relaxation time.

[0019] When performing spatiotemporal alignment preprocessing, the intelligent analysis engine timestamps match the three types of sensing data according to the sampling time, and compensates for the spatial displacement deviation according to the tank conveying speed; for near-infrared spectrum data, the wavelet packet decomposition is used to extract the 5th order detail coefficient as the sulfur dioxide feature vector; for eddy current electromagnetic data, the Hilbert transform is used to extract the instantaneous amplitude envelope and instantaneous phase trajectory as the heavy metal feature matrix; for bioimpedance data, the principal component analysis is used to reduce the dimension and retain the first three principal components as the microorganism feature tensor.

[0020] The specific form of the three-coupling equation set is: the sulfur dioxide correction value is equal to the original sulfur dioxide measurement value multiplied by 1 minus the square root of the ratio of the heavy metal concentration index to the preset threshold; the heavy metal correction value is equal to the original heavy metal measurement value multiplied by 1 plus the absolute value of the difference between the microorganism pollution level and the reference value divided by 5; the microorganism correction value is equal to the original microorganism determination level multiplied by 1 minus the cube of the ratio of the sulfur dioxide correction value to the upper limit of safety.

[0021] Self-calibration mechanism; before the start of each batch of detection, 3 types of calibration tanks in the standard sample library are automatically called, the full-process detection is sequentially performed, and the reference output values of each sensing unit are recorded; in subsequent actual detection, the relative deviation of the current output value and the reference output value is calculated in real time, if the deviation exceeds 3%, the sensing unit parameter re-adjustment program is triggered, and the system is returned to the calibration state by adjusting the light source intensity, excitation current amplitude or lock-in loop gain.

[0022] Abnormal sample isolation mechanism; when any two of the three indicators output by the intelligent analysis engine exceed the preset safety threshold, the high-speed sorting execution mechanism is automatically activated, the corresponding tank is moved out of the main conveying belt by the pneumatic push rod and is introduced into the isolation warehouse; at the same time, an electronic report containing the over-standard index type, the over-standard amplitude and the detection timestamp is generated and pushed to the quality traceability system database.

[0023] Compared with the prior art, the present application has the following advantages:

[0024] The detection method of the present application can be completed in 5 minutes per detection in actual production line application, and the production enterprise does not need to significantly slow down the production rhythm for detection. The detection precision is that the sulfur dioxide concentration error is controlled within 3%, the heavy metal concentration error is less than 5%, and the accuracy of microorganism pollution judgment is as high as 98% or more. Compared with the traditional cumbersome method of detecting each index respectively, the detection efficiency is improved by more than 4 times, and the manual intervention requirement is reduced by 90%, which truly realizes the high-speed, accurate and fully automatic detection goal expected by modern food can production line.

[0025] The present application has good adaptability, and can meet the detection requirements by replacing the corresponding sensing probe for common fruit cans, meat cans, relatively special seafood cans and vegetable cans, thereby providing a flexible solution for enterprises.

[0026] The present application fundamentally solves many pain points faced by traditional detection methods by constructing four core technical systems of multi-modal synchronous acquisition, cross-modal feature coupling, dynamic threshold decision and online self-calibration. Firstly, the multi-index synchronous detection is truly realized, the detection efficiency is greatly improved, and the enterprise says goodbye to the cumbersome process of detecting each item; secondly, the sample pretreatment link is saved, which not only reduces the operation complexity, but also reduces the error source from the source; thirdly, through the unique feature coupling and dynamic threshold mechanism, the stability and environmental adaptability of detection are significantly improved, and the detection result is more reliable; fourthly, the built-in self-calibration function always guarantees the long-term stable operation of the system. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. The following drawings only show some embodiments of the present application, and should not be considered as limiting the scope. For those skilled in the art, other related drawings can also be obtained without creative labor.

[0028] Figure 1 It is a physicochemical index detection method flow chart for food can production provided by the embodiments of the present application. DETAILED DESCRIPTION

[0029] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only for selected embodiments of the present application.

[0030] Please refer to Figure 1 , Figure 1 is a food can production physicochemical index detection method flow chart provided by an embodiment of the present application, comprising the following steps:

[0031] Step S1, deploying a multi-modal sensing array at an online detection station of a food can production line, the multi-modal sensing array being composed of a near-infrared spectroscopy sensing unit, an eddy current electromagnetic sensing unit and a bioimpedance sensing unit, the three being arranged in a ring shape and coaxially aligned with a center axis of a to-be-detected can body;

[0032] The three sensing units of the multi-modal sensing array respectively adopt different detection mechanisms: the near-infrared spectroscopy sensing unit detects the vibration absorption characteristics of sulfur dioxide molecules by emitting a 800nm to 2500nm wavelength light beam; the eddy current electromagnetic sensing unit detects the eddy current loss caused by heavy metal ions by applying a 50kHz to 500kHz alternating magnetic field; the bioimpedance sensing unit detects the impedance change caused by microbial contamination by applying a 1kHz to 10kHz sinusoidal excitation current. The sampling frequency of the three sensing units is uniformly set to 10 times per second, and the sampling time length is 300s, ensuring that the signals of each mode are strictly aligned in the time dimension.

[0033] Step S2, after starting the detection process, emitting a continuously modulated light beam with a wavelength range of 800nm to 2500nm to the surface of the can body through the near-infrared spectroscopy sensing unit, synchronously collecting the reflected spectrum signal and extracting the absorption intensity parameter corresponding to the vibration characteristic peak of the sulfur dioxide molecule;

[0034] After starting the detection process, the multi-modal sensing array synchronously collects the original physicochemical signals of the to-be-detected food can, the original physicochemical signals including near-infrared spectroscopy absorption signals, eddy current electromagnetic response signals and bioimpedance change signals. The near-infrared spectroscopy sensing unit detects the characteristic absorption peak of sulfur dioxide molecules in particular, extracts the absorption intensity characteristics in a specific waveband, which directly reflects the concentration level of sulfur dioxide.

[0035] Step S3, an alternating magnetic field with a frequency of 50 kHz to 500 kHz is applied by the eddy current electromagnetic sensing unit, a vortex loss signal caused by metal ions inside the receiving coil is received, and the comprehensive concentration index of cadmium, lead and mercury heavy metal ions is calculated according to the phase shift and amplitude attenuation rate;

[0036] The eddy current electromagnetic sensing unit generates eddy current effect through the interaction of alternating magnetic field and metal ions. Different heavy metal ions produce characteristic phase shift and amplitude attenuation due to the difference in conductivity and magnetic permeability. The system extracts the eddy current response characteristics in the frequency range of 50 kHz to 500 kHz, and calculates the comprehensive concentration index of cadmium, lead and mercury heavy metal ions by analyzing the combination mode of phase shift and amplitude attenuation rate.

[0037] Step S4, a sine excitation current with an amplitude of 0.5 V and a frequency of 1 kHz to 10 kHz is applied by the bio-impedance sensing unit, the complex impedance response curve of the contents in the tank in time domain is measured, and the microbial pollution level is determined according to the curve inflection point slope and relaxation time constant;

[0038] The bio-impedance sensing unit utilizes the difference in dielectric properties of microbial cell membrane and cytoplasm. When there is microbial pollution, the impedance response of excitation current at different frequencies will change characteristically. The system extracts the complex impedance curve in the frequency range of 1 kHz to 10 kHz, analyzes the curve inflection point slope and relaxation time constant, which directly reflect the number and activity level of microorganisms, and determines the microbial pollution level accordingly.

[0039] Step S5, the absorption intensity parameters, comprehensive concentration index and pollution level determination results are input into the intelligent analysis engine, the feature extraction sub-module is started after time-space alignment preprocessing, and the sulfur dioxide feature vector, heavy metal feature matrix and microbial feature tensor are generated respectively;

[0040] After completing the original signal collection, the original physical and chemical signals are preprocessed in time domain and the frequency domain features are extracted. The time domain preprocessing includes zero point calibration, baseline drift correction and abnormal pulse elimination. The zero point calibration establishes the reference output value of each sensing unit by collecting blank can samples; the baseline drift correction uses the sliding window least square fitting method to eliminate the temperature drift and long-term stability fluctuation of the sensor itself; the abnormal pulse elimination identifies and filters out the transient noise caused by electromagnetic interference or mechanical vibration based on the 3 times standard deviation criterion. The frequency domain feature extraction adopts the combination of short-time Fourier transform and wavelet packet decomposition, and all feature vectors are normalized to zero mean unit variance space to eliminate the interference of dimension difference on the subsequent fusion process.

[0041] Step S6, a ternary coupling equation set containing sulfur dioxide-heavy metal inhibitors, heavy metal-microorganism synergistic factors and microorganism-sulfur dioxide feedback factors is constructed, and the correction coefficient of each index is solved by iterative least squares method;

[0042] The near-infrared spectrum feature vector, the eddy current electromagnetic feature vector and the bioimpedance feature vector are input into the cross-modal feature coupling network to generate a multi-index coupling feature encoding vector. The cross-modal feature coupling network adopts a double-path attention mechanism architecture, including an intra-modal self-attention module and an inter-modal cross-attention module. The intra-modal self-attention module respectively models the internal correlation of the three feature vectors, and strengthens the weight of the key feature nodes in each modal. The inter-modal cross-attention module constructs a bidirectional interaction weight matrix between the three modes, and the calculation formula is as follows:

[0043] I. Inter-modal attention weight calculation formula:

[0044]

[0045] Wherein, Xi represents the feature vector of the i-th mode, Xj represents the feature vector of the j-th mode, and W is a learnable weight matrix, Xi represents the attention weight of the i-th mode to the j-th mode.

[0046] II. Cross-modal feature fusion formula:

[0047]

[0048] Wherein, Xi represents the enhanced feature vector of the i-th mode after cross-modal interaction.

[0049] III. Final coupling feature encoding formula:

[0050]

[0051] Wherein, represents a vector concatenation operation, represents element-wise addition, is a three-layer fully connected network, is a cross-modal average pooling operation.

[0052] The coupling mechanism enables dynamic adjustment of the sensitivity threshold of the heavy metal detection channel based on the sulfur dioxide detection result. When high concentration of sulfur dioxide is detected, the determination threshold of heavy metal detection is automatically lowered to compensate for possible matrix interference. At the same time, the change of microbial contamination signal can reverse correct the baseline compensation coefficient of near-infrared spectrum, avoiding false judgment of sulfur dioxide quantification caused by microbial metabolites. The final output of the multi-index coupling feature encoding vector z has a dimension of 512, which fully retains the discriminative features of each modality and their interaction.

[0053] Step S7, output the coupled corrected sulfur dioxide content value, heavy metal comprehensive concentration index and microbial contamination level determination result.

[0054] After obtaining the multi-index coupling feature encoding vector, the multi-index coupling feature encoding vector is input into a dynamic threshold adaptation decision maker to output the final detection results and risk level determination of each physical and chemical index. The dynamic threshold adaptation decision maker is composed of two parallel sub-networks: an index quantitative regression network and a risk level classification network. The index quantitative regression network adopts a residual regression structure, and outputs four continuous values of sulfur dioxide concentration, cadmium ion concentration, lead ion concentration and microbial contamination index. The risk level classification network adopts a hierarchical Softmax structure, and outputs three levels of risk determination of "qualified", "slightly over standard" and "severely over standard". The dynamic threshold adaptation mechanism is embodied in the environmental parameter compensation module embedded in the decision maker. The module receives the production line environmental temperature, humidity and can information in real time, retrieves the statistical distribution parameters of the corresponding historical batch through table lookup method, and dynamically adjusts the output bias term of the regression network and the decision boundary of the classification network. For example, in a high temperature and high humidity environment, the determination threshold of microbial contamination index is automatically floated by 15% to avoid false positive alarm caused by environmental factors; for newly produced batches, conservative threshold strategy is adopted, and after accumulating enough samples, the determination standard is gradually relaxed. The decision maker output results are pushed to the production line control terminal and quality traceability database simultaneously to realize real-time closed-loop feedback of detection results.

[0055] To ensure the long-term stability and accuracy of the detection system, the application introduces an online self-calibration mechanism, which automatically starts the standard sample re-detection process after each detection task is completed. The standard sample is a pre-packaged known concentration sulfur dioxide solution, a composite of cadmium lead mixed standard solution and sterilized culture medium, and its physical and chemical parameters are calibrated by the national metrology institutions. The system compares the detection results of the standard sample with the calibration value, calculates the deviation coefficient and sensitivity decay factor of each sensing unit. If the deviation exceeds the preset tolerance range of 5%, the sensor parameter recalibration process is triggered: the near-infrared spectral unit performs light source intensity calibration and wavelength accuracy verification, the eddy current electromagnetic unit performs excitation frequency and phase reference correction, and the bioimpedance unit performs electrode cleaning and impedance reference reset. After updating the calibration parameters, the system automatically resets the normalization parameters of the feature extraction module and the compensation coefficient of the decision maker to ensure the accuracy of the next round of detection is not affected by sensor aging.

[0056] The detection method of the application can be completed in 5 minutes in actual production line application, and the production enterprise does not need to significantly slow down the production rhythm for detection. The detection precision: the sulfur dioxide concentration error is controlled within 3%, the heavy metal concentration error is less than 5%, and the accuracy of microorganism pollution judgment is as high as 98% or more. Compared with the traditional cumbersome method of detecting each index respectively, the detection efficiency is improved by more than 4 times, and the manual intervention requirement is reduced by 90%, which truly realizes the high-speed, accurate and fully automatic detection goal expected by modern food can production line. In addition, the application has good adaptability, whether it is common fruit cans, meat cans, or relatively special seafood cans, vegetable cans, can meet the detection needs by replacing the corresponding sensing probe, providing a flexible solution for enterprises.

[0057] From the perspective of technical breakthrough, the application fundamentally solves many pain points faced by traditional detection methods by constructing four core technical systems: multi-modal synchronous acquisition, cross-modal feature coupling, dynamic threshold decision and online self-calibration. First, it truly realizes multi-index synchronous detection, greatly improves detection efficiency, and lets enterprises say goodbye to the cumbersome process of detecting each item; second, it eliminates the sample pretreatment link, not only reducing the operation complexity, but also reducing the error source from the source; third, through the unique feature coupling and dynamic threshold mechanism, the stability and environmental adaptability of the detection are significantly improved, making the detection result more reliable; fourth, the built-in self-calibration function always ensures the long-term stable operation of the system.

[0058] The above only describes the preferred embodiments of the application and is not intended to limit the application. For those skilled in the art, the application has various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the application shall be included in the protection scope of the application.

Claims

1. A method for detecting physicochemical indexes in food can production, characterized in that, The method comprises the following steps: Step S1, deploying a multi-modal sensor array on an online detection station of a food can production line, the multi-modal sensor array being composed of a near-infrared spectroscopy sensing unit, an eddy current electromagnetic sensing unit and a bio-impedance sensing unit, the three units being arranged in a ring shape and coaxially aligned with the central axis of the can to be detected; Step S2, after starting the detection process, the near-infrared spectroscopy sensing unit emits a continuous modulated light beam with a wavelength range of 800nm to 2500nm to the surface of the can, synchronously collects the reflected spectrum signal and extracts the absorption intensity parameter corresponding to the sulfur dioxide molecular vibration characteristic peak; Step S3, the eddy current electromagnetic sensing unit applies an alternating magnetic field with a frequency of 50kHz to 500kHz, receives the eddy current loss signal caused by the metal ions in the can through an induction coil, and calculates the comprehensive concentration index of cadmium, lead and mercury heavy metal ions according to the phase shift and amplitude attenuation rate; Step S4, the bio-impedance sensing unit applies a sinusoidal excitation current with an amplitude of 0.5V and a frequency of 1kHz to 10kHz, measures the complex impedance response curve of the can content in the time domain, and determines the microbial pollution level according to the curve inflection point slope and relaxation time constant; Step S5, the absorption intensity parameter, comprehensive concentration index and pollution level determination result are input into an intelligent analysis engine, a feature extraction submodule is started after time-space alignment preprocessing, and a sulfur dioxide feature vector, a heavy metal feature matrix and a microorganism feature tensor are generated respectively; Step S6, a three-element coupled equation set containing sulfur dioxide-heavy metal inhibition factor, heavy metal-microorganism synergistic factor and microorganism-sulfur dioxide feedback factor is constructed, and the correction coefficient of each index is solved by iterative least squares method; Step S7, output the coupled and corrected sulfur dioxide content value, heavy metal comprehensive concentration index and microbial pollution level determination result.

2. The physicochemical index detection method for food can production according to claim 1, characterized in that, The near-infrared spectroscopy sensing unit comprises a light source module, a light splitting module, a detector module and a temperature control compensation module; the light source module uses a halogen tungsten lamp as a wide-spectrum radiation source, the output light beam is focused by a gold-coated concave mirror and then scanned in wavelength by a rotating grating light splitting module; the detector module selects an indium gallium arsenide photodiode array.

3. The physicochemical index detection method for food can production according to claim 2, characterized in that, The temperature control compensation module stabilizes the working temperature of the detector within the range of 25℃±0.5℃ by a thermoelectric cooling piece.

4. The physicochemical index detection method for food can production according to claim 1, characterized in that, The eddy current electromagnetic sensing unit comprises an excitation coil group, a differential receiving coil group, a lock-in amplifier circuit and a digital phase detector; the differential receiving coil group adopts a common-mode rejection structure to eliminate environmental electromagnetic interference.

5. The physicochemical index detection method for food can production according to claim 4, characterized in that, The excitation coil group is composed of 3 layers of tightly wound copper wire, with an outer diameter of 50mm and 300 turns; the lock-in amplifier circuit outputs a direct current component after coherent demodulation of the input signal and the reference signal; the digital phase detector calculates the phase difference value according to the direct current component.

6. The physicochemical index detection method for food can production according to claim 1, characterized in that, The biological impedance sensing unit comprises a constant current source circuit, a four-electrode probe array, a high-speed analog-to-digital converter and an impedance spectrum analysis module; the distance between the two excitation electrodes in the four-electrode probe array is 30 mm, and the two detection electrodes are located inside the excitation electrodes and have a distance of 10 mm; the sampling rate of the high-speed analog-to-digital converter is 100 kHz; the impedance spectrum analysis module uses the Cole-Cole model to fit the measured data and extracts the characteristic relaxation time.

7. The physicochemical index detection method for food can production according to claim 1, characterized in that, When the intelligent analysis engine performs the space-time alignment preprocessing, the three types of sensing data are timestamp matched according to the sampling time, and the spatial displacement deviation is compensated according to the tank conveying speed; the feature extraction submodule extracts the fifth-order detail coefficient as the sulfur dioxide feature vector by wavelet packet decomposition for near-infrared spectrum data; extracts the instantaneous amplitude envelope and instantaneous phase trajectory as the heavy metal feature matrix by Hilbert transform for eddy current electromagnetic data; and extracts the first three principal components as the microbial feature tensor after dimensionality reduction by principal component analysis for biological impedance data.

8. The physicochemical index detection method for food can production according to claim 1, characterized in that, The specific form of the three-coupling equation set is: the sulfur dioxide correction value is equal to the original sulfur dioxide measurement value multiplied by 1 minus the square root of the ratio of the heavy metal concentration index to the preset threshold; the heavy metal correction value is equal to the original heavy metal measurement value multiplied by 1 plus the absolute value of the difference between the microbial pollution level and the reference value divided by 5; and the microbial correction value is equal to the original microbial determination level multiplied by 1 minus the cube of the ratio of the sulfur dioxide correction value to the upper limit of safety.

9. The physicochemical index detection method for food can production according to claim 1, characterized in that, It also includes a self-calibration mechanism; before the start of each batch detection, the three types of calibration tanks in the standard sample library are automatically called, the full-process detection is sequentially performed, and the reference output values of each sensing unit are recorded; In subsequent actual detection, the relative deviation of the current output value and the reference output value is calculated in real time, and if the deviation exceeds 3%, the sensing unit parameter re-adjustment program is triggered, and the system is returned to the calibration state by adjusting the light source intensity, excitation current amplitude or phase-locked loop gain.

10. The method for detecting physicochemical indexes in food can production according to claim 1, characterized in that, It also includes an abnormal sample isolation mechanism; when any two of the three indicators output by the intelligent analysis engine exceed the preset safety threshold, the high-speed sorting actuator is automatically activated, the corresponding tank is moved out of the main conveying belt and introduced into the isolation bin through the pneumatic push rod; at the same time, an electronic report containing the over-standard index type, over-standard amplitude and detection timestamp is generated and pushed to the quality traceability system database.

Images