A method, device and system for detecting the quality of feed

Abstract

This invention relates to a method, apparatus, and system for detecting feed quality, addressing the technical problem that existing electronic nose functions are unsuitable for rapid detection of compound feed. The system includes: a gas acquisition path section, used to form a transmission gas path between the sample bottle and the detection gas chamber, where a controlled sampling positive pressure is formed, and a controlled connection to a clean gas source is established to clean the gas path; a sensing array section, used to construct a sensor matrix adapted to the detection gas chamber for characteristic gases formed during the spoilage process of compound feed; a main control processing section, used to provide computing power, storage resources, and wireless communication resources, and to form and control the sampling process, data processing process, and human-machine interaction process according to preset logic; and an interactive auxiliary section, used to provide a human-machine interface, a local data interface, and a power control module. By rationally planning the complex on-site acquisition process through preset control logic, a new technical framework is provided for the distributed detection of batch feed, rapid data processing, and real-time display of results.

Description

Technical Field

[0001] This invention relates to the field of feed freshness detection technology, specifically to a method, apparatus, and system for detecting feed quality. Background Technology

[0002] During storage and transportation, compound feed is susceptible to changes such as oxidative rancidity, protein degradation, and mold contamination due to factors like temperature, humidity, and light, leading to reduced nutritional value and even the formation of harmful substances. Existing methods for detecting feed freshness have objective limitations. For example, sensory evaluation is heavily influenced by the experience of the evaluators and the environment, and cannot achieve quantitative analysis. Chemical analysis methods have long testing cycles, pose safety risks and environmental pollution, and cannot perform repeated testing on the same sample. Instrumental analysis methods are expensive and complex, making them unsuitable for rapid on-site testing and widespread application. Rapid and accurate detection is crucial for ensuring feed quality and preventing economic losses. Electronic nose technology, an intelligent collection and detection technology that simulates the biological olfactory system using gas sensors, has been widely used in food, pharmaceuticals, and environmental monitoring. However, it has the following drawbacks when applied to compound feed: The gases produced during spoilage are complex in composition, and existing sensor combinations struggle to comprehensively cover these characteristic gases, resulting in insufficient detection sensitivity and accuracy. The evaluation dimension is too singular, failing to separately evaluate interrelated but distinct quality dimensions (freshness and mold metabolic contamination), thus failing to comprehensively reflect feed quality. The existing negative pressure sampling method in the gas collection path leads to unstable gas sampling volumes, affecting detection repeatability. Inappropriate selection of the gas chamber material results in severe gas adsorption, leading to long response times and slow recovery. The lack of an effective cleaning mechanism makes cross-contamination easily occur between measurements of the same sample. Summary of the Invention

[0003] To address the aforementioned problems, embodiments of the present invention provide a method, apparatus, and system for detecting feed quality, thereby resolving the technical issue that existing electronic nose functions are not suitable for rapid detection of compound feed.

[0004] The feed quality detection system of this invention includes: The gas path section is used to form a transmission gas path between the sample bottle and the detection gas chamber. A sampling positive pressure is formed in a controlled manner on the transmission gas path, and a clean gas source is connected in a controlled manner to form a gas path cleaning. The sensing array section is a sensor matrix used to construct a detection chamber adapted to the characteristic gases formed during the spoilage of compound feed. The main control processing section is used to provide computing power resources, storage resources and wireless communication resources, and to form and control the sampling process, data processing process and human-computer interaction process according to preset logic; The interactive assistance section provides a human-computer interaction interface, a local data transmission interface, and a power control module.

[0005] In one embodiment of the present invention, the gas collection path includes a sample bottle, a headspace balancing device, a sampling pipeline, and a detection gas chamber connected in series. A sampling pump and a flow meter are sequentially installed on the sampling pipeline, and a sampling solenoid valve is installed on the sampling pipeline between the sampling pump and the flow meter. The gas collection path also includes an interface for a clean gas source connected in series on the detection gas chamber. An exhaust gas pipeline, an activated carbon filter, and an exhaust port are also sequentially connected in series on the detection gas chamber. A clean solenoid valve is installed on the gas pipeline between the detection gas chamber and the activated carbon filter.

[0006] In one embodiment of the present invention, the detection chamber is a cylindrical sealed cavity with a polished inner wall. A PCB board with uniformly distributed through holes coaxial with the detection chamber is fixed thereon. Standard sockets are evenly distributed around the axis of the detection chamber on the PCB board. Each sensor is mounted on the standard socket to form a circular array. The acquisition array includes several sensors, forming four functional groups, including: The Freshness Core Group is used to detect freshness-related gases produced in the early stages of feed spoilage. The mold monitoring group is used to detect the metabolites produced during mold growth, including alcohols, ketones, esters, terpenes, and volatile organic compounds. The comprehensive indicator group is used to detect other comprehensive indicators related to feed quality; The environmental compensation group is used to monitor ambient temperature and humidity and to perform environmental compensation calibration for other sensors.

[0007] In one embodiment of the present invention, the main control processing unit includes a processor and an analog-to-digital converter (ADC). The processor and the ADC are connected via an SPI interface. The ADC is connected to some sensors via single-ended inputs of each input channel. The processor is connected to another set of sensors via corresponding GPIO interfaces. The processor establishes a data connection with a card reader via a USB interface, a data connection with an OLED display via an I2C interface, and a data connection with the sampling pump, solenoid valve, and flow meter in the gas acquisition path via corresponding GPIO interfaces.

[0008] The feed quality detection method of this invention includes: Sensors are grouped according to the composition characteristics of volatile gases from compound feed, and the grouped sensors are controlled to synchronously / asynchronously sample in positive pressure volatile gases to form feedback data; The volatile gas feedback data is used to extract the component characteristics, and the freshness and mold contamination degree are quantified in two directions based on the component characteristics to form a feed quality grade assessment. Based on the assessment data, an orderly display of feed grade-related data is generated; A data transmission link for the evaluation data is formed based on the direction of the data request.

[0009] In one embodiment of the present invention, the generation of feedback data includes: The sensors are grouped into several groups: a core group for freshness monitoring of ammonia, amines, and hydrogen sulfide; a mold monitoring group for alcohols, ketones, esters, terpenes, and volatile organic compounds; a comprehensive index group for methane, sulfides, and alcohols; and an environmental compensation group for ambient temperature and humidity. Sample pretreatment and sampling pretreatment are performed sequentially. Sensors synchronously sample data in a positive pressure sampling environment to generate feedback data.

[0010] In one embodiment of the present invention, the process of forming a feed quality grade assessment includes: The feedback data is preprocessed to establish the sensor's data response curve; Statistical features are extracted from the data response curve, and the statistical features are then subjected to dimensionality reduction processing to form low-dimensional feature vectors. Gas types and concentrations are identified based on low-dimensional feature vectors, forming freshness and mold contamination indices. Feed quality grades are assessed based on freshness index and mold contamination index.

[0011] In one embodiment of the present invention, the ordered display includes: The current FQI value and quality level, MPI value and mold level, and sensor response curve are displayed on the interactive interface as needed. The interactive interface displays FQI values ​​and quality levels, MPI values ​​and mold levels, and environmental response curves during the process as needed. The interactive interface displays comprehensive judgment and suggestion information during the process as needed.

[0012] In one embodiment of the present invention, the data transmission link for forming the evaluation data includes: Based on the feed circulation process, the detection data and evaluation data are aggregated to the cloud using wireless data links to form continuous monitoring of circulating feed; Based on the feed circulation process, the aggregated data from the cloud is displayed on the interactive interface using a wireless data link, forming relevant quantitative data for each stage of the feed circulation process.

[0013] The feed quality detection device according to an embodiment of the present invention includes: Configure the acquisition module to group sensors according to the composition characteristics of volatile gases from compound feed, and control the grouped sensors to synchronously / asynchronously sample in positive pressure volatile gases to form feedback data; The quality assessment module is used to extract the component characteristics of the volatile gas feedback data, and to quantify the freshness and mold contamination level in two directions based on the component characteristics to form a feed quality grade assessment. The data display module is used to generate an orderly display of feed grade-related data based on the evaluation data; The data synchronization module is used to establish a data transmission link for the evaluation data based on the direction of the data request.

[0014] The feed quality detection method, apparatus, and system of this invention integrate an electronic nose-like structure within a single housing. While ensuring structural stability and integration, it establishes a parallel detection structure for the complex characteristic gases in compound feed, real-time processing capabilities, and a human-machine interface. The complex on-site data collection process is rationally planned through preset control logic, detection accuracy is improved through real-time processing, and detection results are displayed in a timely manner through human-machine interaction, thus improving detection efficiency and accuracy during storage and transportation. This provides a new technical framework for the distributed detection of batch feed, rapid data processing, and real-time result display. Attached Figure Description

[0015] Figure 1 The diagram shown is a schematic representation of the architecture of a feed quality detection system according to an embodiment of the present invention.

[0016] Figure 2 The diagram shown is a schematic diagram of the gas acquisition circuit section in a feed quality detection system according to an embodiment of the present invention.

[0017] Figure 3 The diagram shown is a schematic diagram of the detection chamber in a feed quality detection system according to an embodiment of the present invention.

[0018] Figure 4 The diagram shown is a schematic diagram of the main control processing part in a feed quality detection system according to an embodiment of the present invention.

[0019] Figure 5 The diagram shown is a flowchart illustrating a method for detecting feed quality according to an embodiment of the present invention.

[0020] Figure 6 The diagram shown is a schematic representation of the structure of a feed quality detection device according to an embodiment of the present invention.

[0021] Figure 7 The diagram shown is a schematic representation of the architecture of a device according to an embodiment of the present invention. Detailed Implementation

[0022] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0023] An embodiment of the feed quality detection system of the present invention is as follows: Figure 1 As shown. In Figure 1 In this embodiment, the housing is included, and the following are also included: -The gas collection path section is used to form a transmission gas path between the sample bottle and the detection gas chamber. A sampling positive pressure is formed in a controlled manner on the transmission gas path, and a clean gas source is connected in a controlled manner to form a gas path cleaning.

[0024] A PTFE sampling tube is used to form localized gas transmission paths between the sample bottle and the detection chamber, between the detection chamber and the clean gas source, and between the positive pressure gas source and the sample bottle. Control and metering components such as a sampling pump providing positive pressure, a sampling solenoid valve, a cleaning solenoid valve, and a flow meter are installed along these corresponding localized transmission paths. A headspace balancing device adapted to the sample bottle is also provided to maintain the equilibrium time of the gas inside the sample bottle under constant temperature conditions, ensuring the full release of volatile gases from the feed.

[0025] - The sensing array section is a sensor matrix used to construct a detection chamber adapted to the characteristic gases formed during the spoilage of compound feed.

[0026] Characteristic gases include, but are not limited to, ammonia, amines, hydrogen sulfide, alcohols, ketones, aldehydes, sulfides, and fungal metabolites. Characteristic gases can be classified according to their typical source and stage of spoilage, chemical properties and reaction mechanisms, or deterioration markers. Different types of sensors can be adapted based on this classification. Furthermore, based on the sensor type and gas volatilization characteristics, sensors can be arranged in two-dimensional or three-dimensional structures within the detection chamber.

[0027] - The main control processing section is used to provide computing resources, storage resources and wireless communication resources, and to form and control the sampling process, data processing process and human-computer interaction process according to preset logic.

[0028] Those skilled in the art will understand that using a DSP (Digital Signal Processor), FPGA (Field-Programmable Gate Array), MCU (Microcontroller Unit) system board, SoC (System on a Chip) system board, or a PLC (Programmable Logic Controller) minimum system including I / O can provide computing power, storage resources, and communication resources. Processor components can provide high-frequency processors and computing memory, large-capacity cache space, flexible wired and wireless communication modules, a large number of GPIO control ports, high-precision ADC converters, and parallel channels.

[0029] - The interactive assistance section provides a human-computer interaction interface, a local data transmission interface, and a power control module.

[0030] The interactive interface includes a touch display unit for showing detection progress, results, and interactive controls. The data transmission interface includes a USB port for local data transfer. The power control module includes a BMS (Power Management System) for controlling the switching between internal and external power supplies.

[0031] The feed quality detection system of this invention integrates an electronic nose into a housing. While ensuring structural stability and integration, it forms a parallel detection structure for the complex characteristic gases of compound feed, real-time processing capabilities, and a human-machine interface. The complex on-site data collection process is rationally planned through preset control logic, detection accuracy is improved through real-time processing, and detection results are displayed in a timely manner through human-machine interaction, thus improving detection efficiency and accuracy during storage and transportation. This provides a new technical framework for the distributed detection of batch feed, rapid data processing, and real-time result display.

[0032] An embodiment of the feed quality detection system of the present invention includes a gas acquisition path such as... Figure 2 As shown. In Figure 2 The gas sampling section includes a sample vial, a headspace balancing device, a sampling pipeline, and a detection chamber connected in series. A sampling pump and a flow meter are sequentially installed on the sampling pipeline, and a sampling solenoid valve is installed on the sampling pipeline between the sampling pump and the flow meter. It also includes an interface on the detection chamber connected in series with a clean gas source. The detection chamber also includes an exhaust gas pipeline, an activated carbon filter, and an exhaust port connected in series. A clean gas solenoid valve is installed on the gas pipeline between the detection chamber and the activated carbon filter.

[0033] In practical applications, a sample of compound feed is placed in a sample bottle. A headspace balancing device is used to obtain the sample's volatile gas in a gas-liquid (gas-solid) equilibrium state under constant temperature and time (e.g., 10g, 40°C, 30 minutes). A sampling pump provides positive pressure, and a sampling solenoid valve and flow meter control the flow rate. The detection chamber is filled with volatile gas at a certain pressure for testing. After testing, a clean gas source (nitrogen or clean air) is turned on to clean the detection chamber. The clean solenoid valve is opened, and the waste gas is discharged harmlessly through a waste gas pipe, activated carbon filter, and exhaust port.

[0034] In one embodiment of the invention, the sampling pump is a miniature diaphragm pump (DC5V / 150mA) with a flow rate of 0.5-1L / min. Two 2-position 2-way solenoid valves (DC5V / 200mA) are used. A mini rotor flow meter (0.5-2L / min) is used. The gas chamber is custom-machined from 304 stainless steel. The piping is made of PTFE, with an outer diameter of 6mm, an inner diameter of 4mm, and a length determined according to the layout.

[0035] In one embodiment of the present invention, the acquisition array includes twelve sensors, which are divided into four functional groups.

[0036] The freshness core group is used to detect freshness-related gases produced in the early stages of feed spoilage, including ammonia, amines, and hydrogen sulfide. Specific selection criteria are shown in the table below: The mold monitoring kit is used to detect metabolic products produced during mold growth, including alcohols, ketones, esters, terpenes, and volatile organic compounds. Specific selection criteria are shown in the table below: The comprehensive indicator group is used to detect other comprehensive indicators related to feed quality, including methane, sulfides, and alcohols. Specific selection criteria are shown in the table below: The environmental compensation group is used to monitor ambient temperature and humidity and to perform environmental compensation calibration for other sensors. Specific selection criteria are shown in the table below: In one embodiment of the present invention, the detection chamber of the feed quality detection system is as follows: Figure 3 As shown. In Figure 3 In the process, the detection chamber adopts a cylindrical sealed cavity. Inside the cavity 001, there is a uniformly distributed through hole 002 and a PCB board 003 coaxial with the chamber. Standard sockets 004 are evenly distributed around the axis of the chamber in the circumferential direction on the PCB board. Each sensor 005 is installed on the standard socket to form a circular array layout.

[0037] In one embodiment of the present invention, the detection chamber is made of 304 stainless steel, with a diameter of 100 mm and a height of 60 mm. It possesses advantages such as good chemical stability, corrosion resistance, and high strength, providing ample reaction space (>500 mL) for the sensor. Polishing the inner wall reduces gas adsorption and improves response and recovery speeds. Both the inlet and outlet of the chamber are made of PTFE tubing, utilizing the material's inertness to prevent gas adsorption or reaction.

[0038] In one embodiment of the invention, the housing adopts an aluminum alloy frame + transparent acrylic panel design, with dimensions of 300mm × 200mm × 80mm. The interior employs a layered storage layout. Top layer: OLED display, USB port, power control switch, status LEDs, heat dissipation holes; Middle layer left side: Main control board area (processor, analog-to-digital converter, communication module, power module); Middle layer, right side: Gas system area (sampling pump, solenoid valve, flow meter); Lower layer: Detection chamber and sensor array area.

[0039] The power module is equipped with a 5V / 3.3V regulated power supply and two 18650 lithium batteries (7.4V) as backup power. The typical power consumption is about 5W, which supports outdoor mobile use.

[0040] In one embodiment of the feed quality detection system of the present invention, the main control processing part is as follows: Figure 4 As shown. In Figure 4 The main control processing section includes a processor and an analog-to-digital converter (ADC). The processor and ADC are connected via an SPI interface. The ADC is connected to some sensors via single-ended inputs on its input channels. The processor is connected to other sensors via corresponding GPIO interfaces. The processor is connected to a card reader via a USB interface and to an OLED display via an I2C interface. The processor is also connected to the sampling pump, solenoid valve, and flow meter in the gas sampling path via their respective GPIO interfaces.

[0041] In one embodiment of the present invention, the processor employs an ESP32-S3 dual-core microcontroller (integrating a wireless communication module and a storage module), and the analog-to-digital converter (ADC) is an ADS1256 ADC, which works in conjunction with the microcontroller to acquire sensor signals. The eight ADC channels of the ADS1256 ADC are allocated as follows: Sensors MQ135, MQ4, TGS2611, and DHT22 are connected to the GPIO pins of the ESP32-S3 dual-core microcontroller for data acquisition.

[0042] The ESP32-S3 dual-core microcontroller controls the electrical components in the pneumatic circuit via GPIO pins, including: In one embodiment of the present invention, the sensor acquisition circuit of the main control processing section includes: The sensor heating circuit is used to provide a stable 5V heating voltage for the MOS sensor and is equipped with an overcurrent protection resistor (10Ω / 2W). The signal sampling circuit is used to convert the sensor resistance change into a voltage signal using a voltage divider circuit, and to configure an adjustable resistor for sensitivity calibration. A low-pass filter circuit is used to form a cutoff frequency of 10Hz, and RC filtering is used to remove high-frequency noise from the sampling. The reference voltage circuit is used for analog-to-digital converters to use a built-in 2.5V reference voltage, and is equipped with decoupling capacitors to improve sampling stability.

[0043] In one embodiment of the present invention, the software environment of the processor in the main control processing section is developed using Python + PyQt5, and the functional modules include: Data communication module: Enables communication with the lower-level device via USB / Bluetooth / WiFi; Data storage module: SQLite database, supporting massive data storage and querying; Real-time visualization module: Real-time plotting of sensor response curves; Pattern recognition module: integrates algorithms such as PCA / LDA / SVM / ANN; Report generation module: Automatically generates test reports (PDF format).

[0044] An embodiment of the present invention provides a method for detecting feed quality as follows: Figure 5 As shown. In Figure 5 In this embodiment, the following are included: Step 100: Group the sensors according to the composition characteristics of the volatile gases in the compound feed, and control the grouped sensors to synchronously / asynchronously sample in the positive pressure volatile gases to form feedback data.

[0045] Those skilled in the art will understand that the volatile gases in compound feed include components such as ammonia, amines, hydrogen sulfide, aldehydes, ketones, alcohols, and fungal metabolites, and the concentration and type of these components are uncertain. Grouping sensors based on prior knowledge can cover major component categories. Sensors within the same group can distinguish differences in component characteristics within a single category. By leveraging differences in detection dimensions and sensitivity among sensors within the same group, verifiable, multi-dimensional signal acquisition via different feedback paths can be performed on one or more specific weak components within a specific category. This ensures the coverage and resolution of signals from complex mixed characteristic gases. Maintaining a positive pressure state for the volatile gases ensures pressure stability during signal acquisition, eliminating interference from temperature and humidity changes caused by unstable gas pressure. Based on the independence of the signal feedback channels, the sensor grouping leverages the flexibility of sensor acquisition within the group, allowing for flexible pre-configuration of sensors based on prior knowledge of feed volatility. A time-series-based sensor signal acquisition strategy can be developed, enabling synchronous or asynchronous signal acquisition.

[0046] Step 200: Extract the component characteristics from the feedback data of volatile gases, and quantify the freshness and mold contamination level in two directions based on the component characteristics to form a feed quality grade assessment.

[0047] This technical solution employs two opposing measurement dimensions for feed quality assessment, utilizing the Freshness Index (FQI) and the Mycotoxin Contamination Index (MPI) to evaluate feed quality from the perspectives of freshness and mycotoxin contamination, respectively. This results in more comprehensive and clear test results, facilitating users to make targeted decisions based on different indicators. The component feature extraction process includes normalizing and preprocessing the feedback data to form the feedback response curve, as well as extracting and quantifying different statistical features from the response curve.

[0048] Step 300: Based on the assessment data, form an orderly display of feed grade-related data.

[0049] Coordination of data display through human-computer interaction. The displayed content includes, but is not limited to, system status, temperature and humidity values, detection progress, FQI value and freshness level, MPI value and mold level, sensor response curve, comprehensive judgment and recommendations.

[0050] Step 400: Establish a data transmission link for the evaluation data based on the data request direction.

[0051] Leveraging the flexibility of the hardware architecture, data transmission links are established for different usage scenarios. A USB link is established for laboratory environments to ensure high speed and stability. A Bluetooth link is established for rapid on-site testing to meet low-power and portability requirements. A WiFi link is established for uploading to cloud platforms to achieve remote real-time capabilities.

[0052] The feed quality detection method of this invention addresses the complexity of characteristic gases in compound feed testing by establishing a flexible detection mechanism that progressively adjusts signal acquisition dimensions and performs bidirectional quality assessment. Sensor grouping ensures full adaptation of signal feedback to different stages of feed change during the detection process. A multi-dimensional quality assessment is established based on bidirectional indicators, guaranteeing an objective evaluation of feed changes. Simultaneously, the method enables orderly display and organized storage of detection data, achieving diversity in data presentation and application scenarios. This ensures detection flexibility and data visibility in field applications, data continuity in statistical analysis, and comparison with objective indicators in scenario judgment, forming a complete detection process.

[0053] like Figure 5 As shown, in one embodiment of the present invention, step 100 includes: Step 110: Form sensor groups, including: a freshness core group for ammonia, amines and hydrogen sulfide components; a mold monitoring group for alcohols, ketones, esters, terpenes and volatile organic compounds; a comprehensive index group for methane, sulfides and alcohols; and an environmental compensation group for ambient temperature and humidity.

[0054] Grouping is used to form a coverage range for the characteristics of certain types of gases. Sensor adaptation within the grouping is used to form an accuracy adaptation for the detection of targeted gas components. In one embodiment of the present invention, the freshness core group includes TGS826, TGS825, MQ137 and TGS2603 sensors; the mold monitoring group includes TGS822, TGS2610, TGS2602 and MQ135 sensors; the comprehensive index group includes MQ4, MQ136 and TGS2611 sensors; and the environmental compensation group includes a DHT22 temperature and humidity sensor.

[0055] Step 120: Perform sample pretreatment and sampling pretreatment sequentially.

[0056] Sample pretreatment ensures the full release of volatile gases from the sample, and sampling pretreatment ensures a stable collection environment.

[0057] In one embodiment of the present invention, sample pretreatment includes sample preparation, headspace equilibration, and gas path cleaning. In another embodiment, sample pretreatment includes: pulverizing a quantitative amount of pig compound feed sample, placing it in a sample bottle, sealing it, and equilibrating it in a constant temperature environment for a set time to allow volatile gases in the sample to be fully released into the headspace gas. Sampling pretreatment includes: gas path self-check, controlling the cleaning solenoid valve to introduce clean air, cleaning the gas path system for a set time, and establishing a sampling baseline.

[0058] Step 130: Perform synchronous sampling of the sensor in a positive pressure sampling environment to generate feedback data.

[0059] The acquisition control logic controls the gas path, flow rate, and timing during acquisition, ensuring the consistency and stability of signal acquisition.

[0060] In one embodiment of the present invention, synchronous sensor sampling includes controlling the cleaning solenoid valve to close, the sampling solenoid valve to open, starting the sampling pump to draw sample headspace gas into the gas chamber at a flow rate of 0.5-1 L / min, maintaining a sampling time of 120 seconds, and controlling the analog-to-digital converter to synchronously acquire the response signals of each sensor at a sampling rate of 10 Hz, including the temperature and humidity signals recorded simultaneously. The sampling signals are converted into feedback data including timestamps by the analog-to-digital converter.

[0061] The feed quality detection method of this invention forms a complete coverage and accurate matching of feed gas characteristic spectrum to meet the complex composition of feed and the diverse needs of the circulation process. It ensures multi-scenario adaptability in terms of collection type, collection accuracy and collection environment, and provides a stable data formation basis for the time-series comparison of collected data.

[0062] like Figure 5 As shown, in one embodiment of the present invention, step 200 includes: Step 210: Perform preprocessing on the feedback data to establish the sensor's data response curve.

[0063] In one embodiment of the present invention, the preprocessing of the time-series feedback data includes moving average filtering, outlier removal, and baseline correction. A fixed-size moving average window is used to smooth the signal and suppress high-frequency noise. Outliers are removed by calculating the mean and standard deviation of the filtered signal according to the 3σ principle. The mean of the first 30 seconds of sampling is taken as the baseline. The baseline value is subtracted from the overall signal value to eliminate DC offset and slow drift. Min-Max normalization is used to normalize the feedback data values ​​to determine an interval. Within the determined numerical interval, data response curves for each sensor are established based on the time sequence of the feedback data.

[0064] Step 220: Extract statistical features from the data response curve and perform dimensionality reduction on the statistical features to form a low-dimensional feature vector.

[0065] In one embodiment of the present invention, the extracted statistical features include: By comprehensively characterizing the concentration, composition, and kinetic response of gases through statistical features, a foundation is provided for subsequent component identification.

[0066] In one embodiment of the present invention, principal component analysis (PCA) or linear discriminant analysis (LDA) is used to reduce the dimensionality of the extracted features for each statistical feature. The variance contribution rate of each principal component is calculated, and the components are sorted from largest to smallest, then accumulated sequentially. The top k principal components with a cumulative variance contribution rate > 90% are selected. This effectively eliminates feature redundancy, reduces data dimensionality, and improves the efficiency of gas pattern recognition, providing a lightweight feature foundation for the accurate identification of ammonia, amines, hydrogen sulfide, aldehydes, ketones, alcohols, and fungal metabolites.

[0067] Step 230: Identify gas types and concentrations based on low-dimensional feature vectors to form freshness index and mold contamination index.

[0068] Support Vector Machines (SVMs) or Artificial Neural Networks (ANNs) are used for gas classification, identification, and concentration estimation, forming a regression fit for the gas concentration classification. SVMs construct an optimal classification hyperplane in a high-dimensional feature space, achieving stable classification of small samples and high-dimensional data, making them particularly suitable for gas identification using sensor arrays. Artificial Neural Networks, based on multi-dimensional features, learn the nonlinear mapping relationship between sensor response and gas concentration through multiple layers of neurons, achieving continuous concentration output.

[0069] In one embodiment of the present invention, when using support vector machine (SVM) for component identification, radial basis function (RBF) is used as kernel function, and parameters are optimized through grid search and cross-validation.

[0070] Furthermore, a two-way index calculation is formed based on the concentration characteristics of each category of gas or component.

[0071] The Freshness Index (FQI) comprehensively reflects the freshness of feed, and its calculation formula is as follows: FQI = α1×F_fresh + α2×F_amine + α3×F_H2S + α4×F_VOC in: F_fresh: Freshness feature value (based on the response of sensors TGS826 and TGS825); F_amine: Amine characteristic value (based on the response of sensors TGS2603 and MQ137); F_H2S: Characteristic value of hydrogen sulfide (based on the response of sensor TGS825); F_VOC: Characteristic value of volatile organic compounds (based on the response of sensor TGS2602); α1, α2, α3, and α4 are weighting coefficients, which were determined through experimental calibration.

[0072] The Mold Contamination Index (MPI) comprehensively reflects the degree of mold contamination in feed, and the calculation formula is as follows: MPI = β1×M_alcohol + β2×M_ketone + β3×M_ester + β4×M_terpene + β5×M_sulfide in: M_alcohol: Characteristic value of alcohols (based on the response of sensors TGS822 and TGS2610); M_ketone: Ketone characteristic value (based on the response of sensors TGS822 and MQ4); M_ester: Ester characteristic value (based on sensor MQ4 response); M_terpene: Terpene characteristic value (based on sensor MQ135 response); M_sulfide: Sulfide characteristic value (based on sensor MQ136 response); β1, β2, β3, β4, and β5 are weighting coefficients, determined through experimental calibration.

[0073] Step 240: Assess feed quality grade based on freshness index and mold contamination index.

[0074] Feed quality is quantified using a dual-index range of freshness index and mold contamination index.

[0075] In one embodiment of the present invention, the feed quality grade is quantified as follows: The feed quality detection method of this invention, through a dual-index model, can simultaneously distinguish between bacterial-dominated spoilage and fungal-dominated mold contamination, effectively solving the problem of cross-interference from multi-component gases and achieving refined and standardized grading of feed quality. It enables early, non-destructive warnings, preventing moldy feed from entering the breeding process and reducing the risk of toxins. The dual indices are calculated independently, unaffected by differences in spoilage pathways, and have good adaptability to various types of feed and different spoilage stages.

[0076] like Figure 5 As shown, in one embodiment of the present invention, step 300 includes: Step 310: Display the current FQI value and quality level, MPI value and mold level, and sensor response curve on the interactive interface as needed.

[0077] Based on the sampling time point, the current sampling data is displayed graphically and textually, showing the sensor sampling status, sampling evaluation values, and quality levels in different dimensions.

[0078] Step 320: Display the FQI value and quality level, MPI value and mold level, and environmental response curve during the process on the interactive interface as needed.

[0079] Based on the temporality of the sampling timeline nodes and the temporality of the feed circulation process, combined with the environmental conditions of the sampling timeline nodes, an assessment value and a graphic display of different quality levels throughout the feed life cycle are generated.

[0080] Step 330: Display comprehensive judgment and suggestion information during the process on the interactive interface as needed.

[0081] Based on the testing and evaluation results and the stages of the process, suggestive or informative graphic displays are created. The feed quality detection method of this invention establishes a close connection between monitoring data, working scenarios, and workflow through human-computer interaction, thereby expanding the scenarios for data collection and utilization.

[0082] like Figure 5 As shown, in one embodiment of the present invention, step 400 includes: Step 410: Based on the feed circulation process, use wireless data links to generate monitoring and evaluation data and aggregate them to the cloud to form continuous monitoring of the circulating feed.

[0083] Step 420: Based on the feed circulation process, use the wireless data link to display the aggregated data in the cloud on the interactive interface, forming relevant quantitative data for each stage of feed circulation.

[0084] The feed quality detection method of this invention utilizes wireless communication capabilities to upload and store batch feed circulation time-series detection data, providing a data foundation for further systematic analysis.

[0085] In practical applications, for feed production enterprises, it can be applied to the stages of raw material warehousing, production process control, and finished product delivery, for example: Raw material testing: Freshness testing is conducted on raw materials such as corn, soybean meal, and fishmeal, and spoiled raw materials are rejected; Process monitoring: Online detection is carried out in granulation, cooling, packaging and other stages to prevent unqualified products from entering the market; Finished product sampling inspection: Regularly sample and inspect finished products entering the warehouse, establish quality records, and achieve full traceability.

[0086] For feed distributors, this can be applied to their storage and sales processes, for example: Warehouse inspection: Conduct freshness testing on purchased feed to control the quality of incoming goods; Inventory Management: Regularly test the stock of feed and adjust sales priorities based on the test results to achieve first-in, first-out (FIFO). Customer Recommendation: Based on the test results, we recommend suitable products to farmers and provide value-added services.

[0087] For farms, it can be applied to the feed application stage, for example: Pre-use testing: Freshness testing is conducted on feed before use to ensure feed quality and safety; Storage instructions: Adjust storage conditions (temperature, humidity, ventilation) according to test results to extend shelf life; Effect evaluation: Analyze the impact of feed quality on breeding results by combining animal growth performance data.

[0088] For quality supervision departments, it can be applied to the feed market supervision process, for example: Market sampling inspection: Randomly sample feed sold in the market to crack down on the sale of spoiled feed; Rapid screening: Conduct rapid on-site screening in the event of a sudden feed safety incident; Standard setting: Accumulate testing data to provide a scientific basis for setting standards for feed freshness.

[0089] In one embodiment of the present invention, the detection result obtained by using the above-described detection system and detection method is as follows: The test results were compared and verified with the national standard method as follows: One hundred unknown samples were selected for blind testing, and the accuracy of the detection was verified as follows: the accuracy of the fresh / rotten binary classification reached 92%, and the accuracy of the five-level classification reached 87%.

[0090] The following is a comparison with existing chemical analysis methods: The following is a comparison with existing electronic nose devices: An embodiment of the present invention provides a feed quality detection device, such as... Figure 6 As shown. In Figure 6 In this embodiment, the following are included: The acquisition module 10 is configured to group sensors according to the composition characteristics of volatile gases from compound feed, and control the grouped sensors to synchronously / asynchronously sample in positive pressure volatile gases to form feedback data; The quality assessment module 20 is used to extract the component characteristics of the volatile gas feedback data, and to quantify the freshness and mold contamination level in two directions based on the component characteristics to form a feed quality grade assessment. Data display module 30 is used to generate an orderly display of feed grade-related data based on the evaluation data; The data synchronization module 40 is used to form a data transmission link for the evaluation data according to the direction of the data request.

[0091] like Figure 6 As shown, in one embodiment of the present invention, the configuration acquisition module 10 includes: The sampling strategy configuration unit 11 is used to form sensor groups, including: a freshness core group for ammonia, amines and hydrogen sulfide components; a mold monitoring group for alcohols, ketones, esters, terpenes and volatile organic compounds; a comprehensive index group for methane, sulfides and alcohols; and an environmental compensation group for ambient temperature and humidity. The preset processing control unit 12 is used to sequentially perform sample preprocessing and sampling preprocessing. The sampling process control unit 13 is used to perform synchronous sampling of sensors to generate feedback data in a positive pressure sampling environment.

[0092] like Figure 6 As shown, in one embodiment of the present invention, the quality assessment module 20 includes: The data preprocessing unit 21 is used to preprocess the feedback data and establish the data response curve of the sensor; The data feature extraction unit 22 is used to extract statistical features from the data response curve and perform dimensionality reduction processing on the statistical features to form a low-dimensional feature vector. The bidirectional index quantization unit 23 is used to identify gas types and concentrations based on low-dimensional feature vectors, and to form freshness index and mold contamination index. The quality assessment unit 24 is used to assess the feed quality grade based on the freshness index and the mold contamination index.

[0093] like Figure 6 As shown, in one embodiment of the present invention, the data display module 30 includes: The real-time detection and display unit 31 is used to display the current FQI value and quality level, MPI value and mold level and sensor response curve on the interactive interface as needed; The detection trend display unit 32 is used to display the FQI value and quality level, MPI value and mold level and environmental response curve during the circulation process on the interactive interface as needed. The detection interaction prompt unit 33 is used to display comprehensive judgment and suggestion information during the process on the interactive interface as needed.

[0094] like Figure 6 As shown, in one embodiment of the present invention, the data synchronization module 40 includes: The data uplink synchronization unit 41 is used to aggregate detection data and evaluation data to the cloud via a wireless data link according to the feed circulation process, forming a continuous detection of circulating feed. The cloud data parsing unit 42 is used to display aggregated cloud data on the interactive interface using a wireless data link based on the feed circulation process, forming relevant quantitative data for each stage of feed circulation.

[0095] This application also provides an electronic device, the structure of which is as follows: Figure 7 As shown, the electronic device 4000 includes at least one processor 4001, a memory 4002, and a bus 4003. The at least one processor 4001 is electrically connected to the memory 4002. The memory 4002 is configured to store at least one computer-executable instruction, and the processor 4001 is configured to execute the at least one computer-executable instruction to perform the steps of the feed quality detection method provided in any embodiment or optional embodiment of this application.

[0096] Furthermore, the processor 4001 can be an FPGA (Field-Programmable Gate Array) or other devices with logic processing capabilities, such as an MCU (Microcontroller Unit) or a CPU (Central Processing Unit).

[0097] This application also provides another computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the feed quality detection method provided in any embodiment or optional implementation of this application.

[0098] The computer-readable storage media provided in this application include, but are not limited to, any type of disk (including floppy disk, hard disk, optical disk, CD-ROM, and magneto-optical disk), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic cards, or optical cards. In other words, readable storage media include any medium by which a device (e.g., a computer) stores or transmits information in a readable form.

[0099] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A feed quality detection system, characterized in that, include: The gas path section is used to form a transmission gas path between the sample bottle and the detection gas chamber. A sampling positive pressure is formed in a controlled manner on the transmission gas path, and a clean gas source is connected in a controlled manner to form a gas path cleaning. The sensing array section is a sensor matrix used to construct a detection chamber adapted to the characteristic gases formed during the spoilage of compound feed. The main control processing section is used to provide computing power resources, storage resources and wireless communication resources, and to form and control the sampling process, data processing process and human-computer interaction process according to preset logic; The interactive assistance section provides a human-computer interaction interface, a local data transmission interface, and a power control module.

2. The feed quality detection system according to claim 1, characterized in that, The gas collection path includes a sample bottle, a headspace balancing device, a sampling pipeline, and a detection gas chamber connected in series. A sampling pump and a flow meter are installed in series on the sampling pipeline, and a sampling solenoid valve is installed on the sampling pipeline between the sampling pump and the flow meter. It also includes an interface for a clean gas source connected in series on the detection gas chamber. An exhaust gas pipeline, an activated carbon filter, and an exhaust port are also connected in series on the detection gas chamber. A clean solenoid valve is installed on the gas pipeline between the detection gas chamber and the activated carbon filter.

3. The feed quality detection system according to claim 1, characterized in that, The detection chamber is a cylindrical sealed cavity with a polished inner wall. A PCB board with evenly distributed through holes coaxial with the detection chamber is fixed therein. Standard sockets are evenly distributed around the axis of the detection chamber on the PCB board. The sensors are mounted on the standard sockets to form a circular array. The acquisition array includes several sensors, forming four functional groups, including: The Freshness Core Group is used to detect freshness-related gases produced in the early stages of feed spoilage. The mold monitoring group is used to detect the metabolites produced during mold growth, including alcohols, ketones, esters, terpenes, and volatile organic compounds. The comprehensive indicator group is used to detect other comprehensive indicators related to feed quality; The environmental compensation group is used to monitor ambient temperature and humidity and to perform environmental compensation calibration for other sensors.

4. The feed quality detection system according to claim 1, characterized in that, The main control processing unit includes a processor and an analog-to-digital converter (ADC). The processor and ADC are connected via an SPI interface. The ADC is connected to some sensors via single-ended inputs of each input channel. The processor is connected to other sensors via corresponding GPIO interfaces. The processor is connected to a card reader via a USB interface and to an OLED display via an I2C interface. The processor is also connected to the sampling pump, solenoid valve, and flow meter in the gas sampling path via their respective GPIO interfaces.

5. A method for detecting feed quality, characterized in that, include: Sensors are grouped according to the composition characteristics of volatile gases from compound feed, and the grouped sensors are controlled to synchronously / asynchronously sample in positive pressure volatile gases to form feedback data; The volatile gas feedback data is used to extract the component characteristics, and the freshness and mold contamination degree are quantified in two directions based on the component characteristics to form a feed quality grade assessment. Based on the assessment data, an orderly display of feed grade-related data is generated; A data transmission link for the evaluation data is formed based on the direction of the data request.

6. The method for detecting feed quality according to claim 5, characterized in that, The feedback data includes: The sensors are grouped into several groups: a core group for freshness monitoring of ammonia, amines, and hydrogen sulfide; a mold monitoring group for alcohols, ketones, esters, terpenes, and volatile organic compounds; a comprehensive index group for methane, sulfides, and alcohols; and an environmental compensation group for ambient temperature and humidity. Sample pretreatment and sampling pretreatment are performed sequentially. Sensors synchronously sample data in a positive pressure sampling environment to generate feedback data.

7. The method for detecting feed quality according to claim 5, characterized in that, The assessment of feed quality grade formation includes: The feedback data is preprocessed to establish the sensor's data response curve; Statistical features are extracted from the data response curve, and the statistical features are then subjected to dimensionality reduction processing to form low-dimensional feature vectors. Gas types and concentrations are identified based on low-dimensional feature vectors, forming freshness and mold contamination indices. Feed quality grades are assessed based on freshness index and mold contamination index.

8. The method for detecting feed quality according to claim 5, characterized in that, The ordered display includes: The current FQI value and quality level, MPI value and mold level, and sensor response curve are displayed on the interactive interface as needed. The interactive interface displays FQI values ​​and quality levels, MPI values ​​and mold levels, and environmental response curves during the process as needed. The interactive interface displays comprehensive judgment and suggestion information during the process as needed.

9. The method for detecting feed quality according to claim 5, characterized in that, The data transmission link that forms the evaluation data includes: Based on the feed circulation process, the detection data and evaluation data are aggregated to the cloud using wireless data links to form continuous monitoring of circulating feed; Based on the feed circulation process, the aggregated data from the cloud is displayed on the interactive interface using a wireless data link, forming relevant quantitative data for each stage of the feed circulation process.

10. A feed quality testing device, characterized in that, include: Configure the acquisition module to group sensors according to the composition characteristics of volatile gases from compound feed, and control the grouped sensors to synchronously / asynchronously sample in positive pressure volatile gases to form feedback data; The quality assessment module is used to extract the component characteristics of the volatile gas feedback data, and to quantify the freshness and mold contamination level in two directions based on the component characteristics to form a feed quality grade assessment. The data display module is used to generate an orderly display of feed grade-related data based on the evaluation data; The data synchronization module is used to establish a data transmission link for the evaluation data based on the direction of the data request.

Images