Sauce and marinade flavor detection and evaluation method, system, equipment and storage medium

Abstract

The present application provides a method, system, device and storage medium for detecting and evaluating the flavor of sauce and marinade, and relates to the field of flavor evaluation technology. The method comprises: obtaining sauce and marinade flavor sample data of the sauce and marinade product to be detected, and extracting flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; selecting a sensor based on the flavor reference data to obtain a target sensor; determining flavor detection reference data based on the sauce and marinade process data; performing flavor detection on the sauce and marinade product to be detected based on the target sensor array, and obtaining flavor detection data and detection environment data in each marinating process of the sauce and marinade product to be detected; calibrating the flavor detection data based on the detection environment data to obtain target flavor detection data, and evaluating the flavor of the sauce and marinade product to be detected based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data. The above scheme improves the accuracy of flavor detection and evaluation of sauce and marinade products.

Description

Technical Field

[0001] The present application relates to the technical field of flavor evaluation, and in particular to a method, system, device and storage medium for detecting and evaluating the flavor of a sauce or marinade. Background Art

[0002] Sauce and marinade products are one of the representatives of traditional meat products in my country, and are favored by consumers for their unique flavor. The flavor quality of sauce and marinade products directly affects consumers' sensory experience and willingness to buy, so flavor testing and evaluation of sauce and marinade products is crucial to controlling product quality and improving consumer satisfaction.

[0003] In the prior art, the flavor detection of soy sauce and marinated products mainly relies on manual sensory evaluation, in which well-trained tasters are used to subjectively score the sensory attributes of the soy sauce and marinated products, such as appearance, smell, and taste. However, the above-mentioned prior art has certain limitations. Manual sensory evaluation is easily affected by factors such as the assessor's personal physiological and psychological state, and the objectivity and consistency of the evaluation results are difficult to guarantee. In addition, the efficiency of manual evaluation is low, and it is difficult to meet the detection needs of large-scale products. It is not conducive to real-time flavor quality monitoring on the production line. Therefore, there is an urgent need for an accurate and objective method for detecting and evaluating the flavor of soy sauce and marinated products. Summary of the invention

[0004] The present application provides a method, system, device and storage medium for detecting and evaluating the flavor of sauce and marinade, which can improve the accuracy of detecting and evaluating the flavor of sauce and marinade products.

[0005] In a first aspect, the present application provides a method for detecting and evaluating the flavor of a sauce or marinade, the method comprising: Obtaining sauce and marinade flavor sample data of the sauce and marinade product to be tested, and extracting flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; Selecting a sensor based on the flavor reference data to obtain a target sensor; constructing a target sensor array based on the target sensor type and flavor reference data; Determine flavor detection reference data based on sauce and marinade process data; Based on the target sensor array, flavor detection is performed on the sauce and marinade product to be detected, and flavor detection data and detection environment data of each marinating process of the sauce and marinade product to be detected are obtained; The flavor detection data is calibrated based on the detection environment data to obtain target flavor detection data, and the flavor of the sauce-marinated product to be detected is evaluated based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data.

[0006] By adopting the above technical solution, the sauce flavor detection and evaluation method of this application obtains the sauce flavor sample data of the sauce product to be detected, and extracts the flavor reference data and sauce process data from it, laying a data foundation for the subsequent sensor selection and detection parameter determination. The flavor reference data reflects the key flavor characteristics of the sauce product, and the sauce process data contains the parameter settings of each process link. The combination of the two can provide a comprehensive and accurate reference standard for flavor detection.

[0007] In terms of sensor selection, this application locks in key flavor components based on flavor reference data, and selects sensors that are sensitive to these components as target sensors. This targeted selection can significantly improve the specificity and sensitivity of detection. At the same time, by constructing a sensor array based on the type and number of target sensors, multi-parameter and multi-dimensional perception of the flavor of the sauce and marinade can be achieved, fully capturing the subtle changes in flavor.

[0008] The determination of flavor detection reference data makes full use of the sauce and marinade process data, so that the detection and evaluation standards are closely integrated with the actual production situation. By analyzing the inherent relationship between each process parameter and the formation of flavor, the optimal flavor reference value and judgment threshold are determined, providing a scientific basis for the definition of flavor qualified products.

[0009] This application uses a target sensor array to perform online detection of marinated products, and simultaneously collects flavor detection data and detection environment data. Environmental factors such as temperature, humidity, and airflow speed may interfere with sensor response and introduce detection errors. By analyzing the correlation between environmental data and detection data and building an adaptive calibration model, environmental interference can be effectively eliminated and the reliability and consistency of detection data can be improved.

[0010] In terms of flavor evaluation, this application is based on the target flavor detection data and flavor detection reference data, and realizes the quantitative characterization and grading of flavor quality through a method combining deviation value calculation and threshold judgment. The flavor evaluation results intuitively reflect the flavor qualification and quality of the sauce and marinade products, providing a clear direction for production control and quality improvement.

[0011] In a second aspect of the present application, a method system for detecting and evaluating the flavor of a sauce or marinade is provided, comprising: A data acquisition module, used to acquire multimodal data of the liquid at the drilling fluid inlet and the flowback fluid outlet by using multimodal sensors arranged at the drilling fluid inlet and the flowback fluid outlet; A data processing module is used to preprocess the multimodal data to obtain preprocessed multimodal data, and to extract features from the preprocessed multimodal data to obtain multimodal feature data; The data analysis module is used to extract foundation soil slurry quality parameters based on multimodal feature data to obtain foundation soil slurry quality parameters, and analyze the foundation soil slurry quality based on the foundation soil slurry quality parameters to obtain sauce and marinade flavor detection and evaluation results.

[0012] In a third aspect of the present application, a computer storage medium is provided, wherein the computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above method steps.

[0013] In the fourth aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the above method.

[0014] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. The sauce flavor detection and evaluation method of the present application obtains the sauce flavor sample data of the sauce product to be tested, and extracts the flavor reference data and sauce process data from it, laying a data foundation for the subsequent sensor selection and detection parameter determination. The flavor reference data reflects the key flavor characteristics of the sauce product, and the sauce process data contains the parameter settings of each process link. The combination of the two can provide a comprehensive and accurate reference standard for flavor detection.

[0015] In terms of sensor selection, this application locks in key flavor components based on flavor reference data, and selects sensors that are sensitive to these components as target sensors. This targeted selection can significantly improve the specificity and sensitivity of detection. At the same time, by constructing a sensor array based on the type and number of target sensors, multi-parameter and multi-dimensional perception of the flavor of the sauce and marinade can be achieved, fully capturing the subtle changes in flavor.

[0016] 2. The determination of flavor detection reference data in this application makes full use of the sauce and marinade process data, so that the detection and evaluation standards are closely integrated with the actual production situation. By analyzing the intrinsic relationship between each process parameter and flavor formation, the optimal flavor reference value and judgment threshold are determined, providing a scientific basis for the definition of flavor qualified products.

[0017] 3. This application uses a target sensor array to perform online detection of marinated products, and simultaneously collects flavor detection data and detection environment data. Environmental factors such as temperature, humidity, and airflow speed may interfere with sensor response and introduce detection errors. By analyzing the correlation between environmental data and detection data and building an adaptive calibration model, environmental interference can be effectively eliminated and the reliability and consistency of detection data can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of a method for detecting and evaluating the flavor of a sauce or marinade provided in an embodiment of the present application; Figure 2 An architectural diagram of a sauce and marinade flavor detection and evaluation system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0019] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0020] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0021] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0022] In order to facilitate understanding of the method and system provided by the embodiments of the present application, before introducing the embodiments of the present application, the background of the embodiments of the present application is first introduced.

[0023] Sauce and marinade products are the best among traditional Chinese meat products, and their unique flavor characteristics are highly praised by consumers. The flavor quality of sauce and marinade products is a key factor affecting consumers' sensory experience and purchasing desire. Therefore, flavor testing and evaluation of sauce and marinade products is of vital importance to controlling product quality and improving consumer satisfaction.

[0024] At present, the flavor detection of soy sauce and marinated products mainly adopts the method of manual sensory evaluation. This method requires professionally trained tasters to subjectively score the sensory characteristics of soy sauce and marinated products, such as appearance, smell, and taste. However, manual sensory evaluation has certain limitations. The evaluation results are easily affected by factors such as the evaluator's own physiological state and psychological state, and it is difficult to ensure the objectivity and consistency of the evaluation results. In addition, the efficiency of manual evaluation is relatively low, and it is difficult to meet the needs of large-scale batch testing of products. At the same time, this method is not conducive to real-time monitoring of flavor quality on the production line.

[0025] Therefore, there is an urgent need for a method that can accurately and objectively detect and evaluate the flavor of sauce and marinated products. This method should be able to overcome the subjectivity and inefficiency of manual sensory evaluation, provide a fast, quantitative, and standardized means of flavor quality evaluation, and provide strong technical support for the production control and quality improvement of sauce and marinated products, ultimately meeting the increasing quality demands of consumers.

[0026] After the background introduction of the above content, those skilled in the art can understand the problems existing in the prior art. The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0027] Based on the above background technology, please refer to Figure 1 , Figure 1 A flowchart of a method for detecting and evaluating the flavor of a sauce and marinade provided in an embodiment of the present application is provided. The system can be implemented by a computer program or can be run as an independent tool application. Specifically, in an embodiment of the present application, the method can be applied on a server, but can also be applied to electronic devices such as a server. A method for detecting and evaluating the flavor of a sauce and marinade comprises the following steps: S101, obtaining sauce and marinade flavor sample data of the sauce and marinade product to be tested, and extracting flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; Specifically, in the sauce flavor detection and evaluation method of the present application, the sauce flavor sample data of the sauce product to be detected is first obtained. Specifically, a certain number of sauce samples are collected from the production line of the sauce product, covering each batch and production period of the product. The collected samples should be representative and able to reflect the overall characteristics and changing trends of the flavor of the sauce product.

[0028] The purpose of obtaining the flavor sample data of soy sauce and marinade is to provide a data basis for subsequent flavor detection and evaluation. By analyzing and mining the sample data, we can gain insight into the inherent laws of the flavor formation of soy sauce and marinade products, and provide guidance for subsequent steps such as sensor selection and detection parameter setting.

[0029] After obtaining the sauce and marinade flavor sample data, it is necessary to extract flavor reference data and sauce and marinade process data from it. Flavor reference data refers to data that can reflect the key flavor characteristics of sauce and marinade products, such as the content of flavoring substances and the composition of volatile components. These data can be obtained through instrument analysis or expert evaluation, and serve as the target reference for flavor detection. Sauce and marinade process data include the process parameters of each link in the production of sauce and marinade products, such as raw material ratio, marinating temperature, marinating time, etc. These parameters are closely related to the formation of sauce and marinade flavor.

[0030] The process of extracting flavor reference data and sauce and marinade process data requires the comprehensive use of multiple data processing technologies, such as feature selection, data fusion, etc. By analyzing the correlation between flavor reference data and sauce and marinade process data and establishing a quantitative correlation model between flavor quality and production technology, accurate description and prediction of sauce and marinade flavor can be achieved.

[0031] Based on the above embodiment, as an optional embodiment, extracting flavor reference data and sauce and brine process data based on sauce and brine flavor sample data includes: S301, preprocessing the sauce and marinade flavor sample data to obtain preprocessed sauce and marinade flavor sample data, and performing feature extraction based on the preprocessed sauce and marinade flavor sample data to obtain flavor sample data and sauce and marinade process sample data; Specifically, after obtaining the sauce and marinade flavor sample data, the present application method first preprocesses it to obtain the preprocessed sauce and marinade flavor sample data. This is because the original sauce and marinade flavor sample data usually has problems such as noise interference, missing data, inconsistent dimensions, etc., and direct use will affect the effect of subsequent feature extraction and model training. In order to improve data quality and availability, the sauce and marinade flavor sample data must be preprocessed and cleaned.

[0032] The preprocessing process mainly includes steps such as data screening, noise filtering, and data normalization. First, the outliers of the sauce and marinade flavor sample data are detected and eliminated to remove abnormal data points that obviously deviate from the normal range. Then, digital filtering, wavelet transform and other methods are used to reduce noise and smooth the data to eliminate high-frequency noise and random interference. Next, data from different sources and different dimensions are normalized and mapped to a unified scale space to eliminate the impact of dimensional differences. In addition, missing data need to be interpolated and repaired to ensure the integrity and continuity of the data.

[0033] After obtaining the preprocessed sauce and marinade flavor sample data, feature extraction is further performed based on these data to obtain flavor sample data and sauce and marinade process sample data. Feature extraction refers to mining key characteristic indicators that can reflect the sauce and marinade flavor quality and production process characteristics from the original data, providing data support for subsequent flavor evaluation and production optimization. This process requires full use of domain knowledge and data mining technology to extract the most essential and effective feature representation from massive, high-dimensional sample data.

[0034] S302, performing statistical screening on the flavor sample data to obtain flavor reference data; Specifically, after obtaining the flavor sample data, the method of the present application further performs statistical screening on it to obtain flavor reference data. This is because the extracted flavor sample data is usually huge in quantity and multi-dimensional, and inevitably contains some redundancy, noise and atypical samples. Directly using them for subsequent analysis and modeling not only has high computational complexity, but may also introduce misleading and interference. In order to optimize the quality and representativeness of the flavor sample data, it is necessary to use statistical analysis methods to screen out the flavor sample subsets with the most reference value and typical significance to form the flavor reference data.

[0035] The statistical screening process is mainly based on the statistical characteristics and distribution laws of flavor sample data. Descriptive statistics, inferential statistics and other methods are used to select the most representative and discriminative samples from the flavor sample data as reference standards. First, descriptive statistical analysis is performed on each characteristic indicator of the flavor sample data, such as calculating the mean, median, standard deviation, quantile, etc., to understand the central tendency, degree of dispersion and distribution form of each indicator. Then, inferential statistical methods such as hypothesis testing and variance analysis are used to examine the significant differences between different flavor samples, and preliminarily screen out key characteristic indicators that have a significant impact on flavor quality.

[0036] On this basis, we further use dimensionality reduction methods such as principal component analysis (PCA) and factor analysis (FA) to extract a few principal components or common factors that best represent the overall variation trend of the data from the high-dimensional feature space of the flavor sample data, remove redundant and noise features, and achieve dimensionality reduction compression of the flavor sample data. By analyzing the distribution of principal component scores or factor scores, we select samples with scores distributed in the central area and closest to the overall mean level as flavor reference samples. These reference samples are the most representative and concentrated in the flavor feature space, and can fully reflect the general characteristics of the flavor quality of sauce-marinated products.

[0037] S303, statistically screening the sample data of the sauce and marinade process to obtain the sauce and marinade process data.

[0038] Specifically, after obtaining the sample data of the sauce and marinade process, the present application method adopts a statistical screening method similar to the flavor sample data to select the most representative and reference-worthy sample subset to form the sauce and marinade process data. The sauce and marinade process sample data records the various process parameters and process variables of the sauce and marinade production process. The data has high dimensions and large data volumes, and there may be large differences and fluctuations between data from different batches and different time periods. In order to extract the most essential, accurate, and comprehensive process feature information from the massive amount of sauce and marinade process sample data, statistical analysis methods must be used to screen and optimize the data.

[0039] The statistical screening process of the sample data of the sauce and brine process is similar to that of the flavor sample data, and also mainly includes descriptive statistics, inferential statistics, dimensionality reduction analysis, cluster analysis and other steps. First, descriptive statistical analysis is performed on the various process parameters and process variables of the sample data of the sauce and brine process, such as calculating the mean, standard deviation, coefficient of variation, etc., to understand the central trend, degree of dispersion and relative variation level of each parameter. Then, inferential statistical methods such as hypothesis testing and regression analysis are used to examine the correlation and significant influence between each process parameter and the flavor quality of the sauce and brine, and preliminarily screen out the key process parameters that have a decisive role in the formation of the flavor of the sauce and brine. Next, dimensionality reduction methods such as principal component analysis and factor analysis are used to extract a few principal components or common factors that can best explain the variation of flavor quality from the high-dimensional parameter space of the sample data of the sauce and brine process, so as to achieve dimensionality reduction compression and feature extraction of the sauce and brine process data. By analyzing the score distribution of each sample in the principal component space or factor space, the samples with the scores closest to the overall mean level and the most representative of the general production process characteristics are selected as the reference samples of the sauce and brine process. The combination of production process parameters corresponding to these reference samples represents the optimal and most stable operation mode of sauce and brine production.

[0040] S102, selecting a sensor based on flavor reference data to obtain a target sensor; Specifically, after obtaining the flavor reference data, the method of the present application performs the step of selecting a sensor based on the flavor reference data to obtain a target sensor. The flavor reference data reflects the key flavor characteristics of the sauce and marinade products, such as the type and content of flavoring substances. The selection of sensors needs to match these key flavor characteristics in order to achieve effective perception and recognition of the sauce and marinade flavor.

[0041] Specifically, the dominant components of the flavor of the sauce and marinade, such as amino acids, organic acids, sugars, etc., are determined based on the flavor reference data. Different types of sensors have different response sensitivities and selectivities to these components. By analyzing the correlation between the flavor reference data and the sensor response characteristics, the sensor with the highest matching degree with the dominant flavor component is selected as the candidate sensor.

[0042] Among the candidate sensors, the sensor's stability, repeatability, anti-interference ability and other performance indicators are further examined, and the sensor with the best performance is selected as the target sensor. This optimization process often requires multiple rounds of experiments and data analysis to balance the sensitivity, selectivity and stability of the sensor. If necessary, it is also necessary to develop new sensing materials and optimize the sensor structure to meet the specific needs of sauce and marinade flavor detection.

[0043] By selecting sensors based on flavor reference data, the target sensors obtained can accurately perceive and capture the key characteristics of sauce and marinade flavors, significantly improving the pertinence and accuracy of flavor detection. At the same time, the selected target sensors are both stable and reliable, able to adapt to the complex and changeable environment of sauce and marinade production, and ensure the consistency and repeatability of detection data.

[0044] Based on the above embodiment, as an optional embodiment, selecting a sensor based on flavor reference data to obtain a target sensor includes: According to the flavor reference data, the key flavor components are determined, and based on the key flavor components, sensors that are sensitive to the key flavor components are screened to obtain target sensors.

[0045] Specifically, in the process of selecting the target sensor, the present application method first determines the key flavor components based on the flavor reference data. The flavor reference data identifies the core material composition of the sauce flavor, such as glutamic acid, inosinic acid, flavor peptides, etc., through chemical analysis, chromatography-mass spectrometry, etc. These key flavor components are the material basis for the unique taste and flavor characteristics of the sauce, and need to be considered when selecting the sensor.

[0046] After determining the key flavor components, further screening sensors that are sensitive to the key flavor components based on the physicochemical properties of these components, thereby obtaining the target sensor. Specifically, according to the polarity, charge properties, molecular weight, etc. of the key flavor components, select the matching sensitive materials and transduction principles. For example, for negatively charged glutamate, a cationic electrochemical sensor can be selected; for hydrophobic flavor peptides, a piezoelectric sensor based on hydrophobic interaction can be selected.

[0047] During the screening process, the focus was on evaluating the sensor's response sensitivity, selectivity, and linear range to key flavor components. Response sensitivity reflects the sensor's response capability and detection limit to the substance being tested, selectivity reflects the sensor's specific recognition capability to the target substance, and linear range determines the sensor's dynamic detection range. By comprehensively considering these indicators, the sensor with the most sensitive response to key flavor components, the best selectivity, and the most matching linear range was selected as the target sensor.

[0048] The significance of screening target sensors is to achieve accurate matching of sensor functions with the flavor to be tested. By focusing on the key components of flavor and selecting sensors that are highly compatible with them, the sensitivity and specificity of the sensor to the flavor of sauce and marinade are significantly improved. Compared with conventional empirical selection or broad-spectrum response sensors, the target sensors selected by the method of this application can better capture the key attributes of the flavor of sauce and marinade, the correlation between the sensor response and the flavor characteristics is closer, and the pertinence and accuracy of flavor recognition are greatly improved.

[0049] S103, constructing a target sensor array based on the target sensor type and the flavor reference data; Specifically, after obtaining the target sensor, the method of the present application further constructs a target sensor array based on the target sensor type and flavor reference data. The flavor of soy sauce and marinade is the result of the interaction and synergistic perception of multiple flavoring substances. It is difficult for a single sensor to fully reflect the complexity and diversity of the flavor of soy sauce and marinade. Constructing a target sensor array can realize multi-parameter and multi-dimensional perception of the flavor of soy sauce and marinade, and improve the comprehensiveness and accuracy of flavor detection.

[0050] The process of constructing the target sensor array requires comprehensive consideration of the type, quantity, and arrangement of the target sensors. First, the main characteristic dimensions of the sauce flavor, such as salty and fresh taste, meaty flavor, and aftertaste, are determined based on the flavor reference data. Then, target sensors matching each characteristic dimension are selected to form a sensor subarray. Certain complementary and redundant designs can be adopted between different subarrays to enhance the robustness and reliability of the array.

[0051] After determining the sensor subarray, further optimize the spatial arrangement of the sensors and the signal modulation method. Reasonable arrangement can reduce mutual interference between sensors and improve the signal-to-noise ratio and resolution of the array. The choice of signal modulation method must take into account both detection speed and data quality. Common methods include frequency domain modulation and time domain modulation.

[0052] The constructed target sensor array is like an "electronic nose" that can perceive and identify the flavor of sauces and marinades in three dimensions. Each sensor subarray is responsible for sensing specific flavor characteristics, and the response data of multiple subarrays can be fused to form a comprehensive and accurate flavor "fingerprint".

[0053] Based on the above embodiment, as an optional embodiment, a target sensor array is constructed based on the target sensor type and flavor reference data, including: S201, determining the number of each target sensor type based on flavor reference data; Specifically, after determining the target sensor type, the method of this application further determines the number of each target sensor type based on the flavor reference data, laying the foundation for building a target sensor array. This is because the contribution and importance of different types of sensors to the flavor of the sauce and marinade are different, and the number ratio of each type of sensor needs to be reasonably configured to achieve a comprehensive and balanced perception of the flavor of the sauce and marinade by the sensor array.

[0054] The process of determining the number of sensors requires an in-depth analysis of the correlation pattern between flavor reference data and sensor response. First, the flavor reference data is subjected to principal component analysis to identify the key components that contribute most to the flavor of the marinated sauce, such as flavor amino acids, flavor nucleotides, aroma compounds, etc. Then, the number of types, concentration distribution, flavor activity and other indicators of each key component are counted, and the theoretical ratio of each target sensor is preliminarily estimated based on this.

[0055] On this basis, we further combined the sensitivity coefficient, linear range and other parameters of the target sensor, and used mathematical modeling and simulation optimization methods to finely determine the optimal number of each type of sensor. Common optimization criteria include maximizing the amount of flavor information collected and minimizing the redundancy of the sensor array. By constructing a multi-objective optimization model and using intelligent algorithms to solve it, we can obtain the optimal configuration of each type of target sensor and achieve a balance between the scale and efficiency of the sensor array.

[0056] Reasonable determination of the number of target sensors of various types is crucial to building a high-performance, low-redundancy sensor array. The optimal configuration of the number of sensors enables the sensor array to achieve all-round, multi-level, three-dimensional perception of the flavor of sauces and marinades at the smallest scale. Each key flavor can be perceived and identified in a targeted manner, and the information supplementation and mutual verification of various sensors significantly improve the comprehensiveness and reliability of flavor perception.

[0057] S202: Arrange the target sensor array according to a preset arrangement array based on the quantity of each target sensor type.

[0058] Specifically, after determining the optimal number of each type of target sensor, the method of the present application uses a preset arrangement array to spatially arrange the target sensors, thereby obtaining the final target sensor array. The arrangement of sensors directly affects the topological structure and signal transmission characteristics of the array, and requires careful design and optimization to give full play to the synergistic effect of various sensors and achieve efficient collection and fusion of sauce and marinade flavor information.

[0059] The preset array arrangement is usually based on the optimization theory and empirical model of the sensor array, taking into account the coupling between sensors, the spatial resolution of the signal, the symmetry and coverage of the array, etc. Common arrangement modes include rectangular array, circular array, cross array, etc. According to the specific needs of the flavor perception of the sauce and marinade, the optimal arrangement mode is selected, and the specific geometric dimensions of the arrangement array are determined according to the number ratio of various sensors, the size of the sensitive area and other parameters.

[0060] During the arrangement process, it is necessary to strictly follow the preset arrangement array rules and accurately place various target sensors at the designated positions of the array. On the one hand, it is necessary to ensure that the spatial distribution of various sensors is uniform to achieve all-round and no-dead-angle collection of sauce and marinade flavor signals; on the other hand, it is necessary to control the spacing and orientation between adjacent sensors to minimize crosstalk and signal coupling between sensors and improve the spatial resolution of the array. At the same time, attention should also be paid to optimizing the integration process of the sensor array, and adopting advanced manufacturing technologies such as micro-electromechanical systems (MEMS) and integrated circuits (IC) to achieve miniaturization, integration and mass production of sensor arrays.

[0061] Building a target sensor array based on a preset array arrangement is another key measure to achieve accurate perception of sauce flavor. By optimizing the design of sensor position and topology, the synergistic and complementary effects of various sensors are maximized to achieve multi-dimensional and high-resolution acquisition of flavor signals. The optimized sensor array is like a highly coordinated "orchestra". Various sensors are precisely positioned and work well together to achieve real-time, synchronous and three-dimensional perception of the dynamic evolution of sauce flavor.

[0062] S104, determining flavor detection reference data based on the sauce and marinade process data; Specifically, after constructing the target sensor array, the method of this application needs to further determine the flavor detection reference data based on the sauce and marinade process data. The sauce and marinade process data reflects the process parameters and operating conditions of the entire production process of the sauce and marinade products, such as the raw material ratio, marinating temperature, marinating time, etc. These parameters and conditions have a decisive influence on the formation of the sauce and marinade flavor. Therefore, the flavor detection and evaluation must be based on the sauce and marinade process data to establish an objective and accurate reference standard.

[0063] The process of determining flavor detection reference data requires an in-depth analysis of the intrinsic relationship between the sauce and marinade process data and flavor formation. First, the sauce and marinade process data is preprocessed and feature extracted to screen out the key parameters with the highest correlation with flavor quality, such as marinade concentration, pH value, heating temperature curve, etc. Then, a data mining algorithm is used to establish a quantitative relationship model between key process parameters and flavor indicators. Commonly used modeling methods include multiple linear regression, partial least squares regression, support vector machine, etc.

[0064] Through modeling analysis, the optimal range of values ​​for each key process parameter and the control threshold for flavor formation are determined. On this basis, combined with product standards and quality requirements, a set of scientific and rigorous flavor detection reference data is formulated. This reference data stipulates the conditions and evaluation criteria for determining the qualified flavor of sauce and marinade products, forming a set of clear and executable flavor quality control specifications.

[0065] The determination of flavor detection reference data makes the flavor detection and evaluation of sauce and marinade more systematic and avoids the randomness and uncertainty of traditional experience judgment. The reference data is directly derived from production practice, which can truly reflect the objective laws of the formation of sauce and marinade flavor and ensure the scientificity and accuracy of flavor judgment. At the same time, the establishment of reference data also provides a clear direction for process optimization, and continuously improves the flavor quality and stability of sauce and marinade products through data feedback and iterative optimization.

[0066] S105, performing flavor detection on the sauce-marinated product to be detected based on the target sensor array, and obtaining flavor detection data and detection environment data of each marinating process of the sauce-marinated product to be detected; Specifically, during the marinating process of the marinated products, the target sensor array continuously collects flavor data and environmental data. Each sensor subarray responds to different characteristics of the marinated flavor, such as saltiness, umami, meat flavor, etc. After amplification, filtering, A / D conversion and other processing, the sensor signal forms a series of flavor detection digital data. At the same time, the environmental sensor records the temperature and humidity curves during the marinating process and stores them synchronously with the flavor detection data.

[0067] As the marinating process progresses, the flavor of the marinated meat shows a dynamic evolution. The sensor array can capture the key nodes of flavor change, such as changes in marinade concentration and meat texture, to form a flavor-time change curve. By comparing and analyzing flavor detection data from different batches and different time periods, the dynamic laws of marinated meat flavor formation can be understood, providing data support for process optimization.

[0068] The significance of continuously collecting flavor detection data and detection environment data is to comprehensively record the entire process of flavor formation and evolution of sauce and marinade products, rather than just sampling and testing the final product. This online continuous detection mode can timely detect abnormal flavor fluctuations and achieve early warning and process control of flavor quality. At the same time, collecting environmental data also helps to analyze the impact of environmental factors on flavor and optimize the environmental adaptability of flavor detection.

[0069] The target sensor array is used to collect flavor detection data and detection environment data, and the information and digitalization level of flavor detection is significantly improved. Massive detection data provides rich original materials for flavor quality analysis. Through big data analysis technology, it is possible to deeply explore the hidden laws of flavor formation and realize the traceability and prediction of flavor quality. At the same time, the collection of environmental data also provides a data basis for the adaptive calibration and compensation of the sensor array, improving the environmental stability of the detection results.

[0070] S106, calibrating the flavor detection data based on the detection environment data to obtain target flavor detection data, and evaluating the flavor of the sauce-marinated product to be detected based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data.

[0071] Specifically, the calibration process first establishes a quantitative relationship model between flavor detection data and environmental factors. Multiple regression analysis and other methods are used to characterize the influence of environmental parameter changes on sensor response. The calibration coefficients under different environmental conditions are calculated according to the model, and the flavor detection data is normalized to eliminate the influence of environmental fluctuations and obtain standardized target flavor detection data.

[0072] After obtaining the target flavor detection data, it is compared and analyzed with the predetermined flavor detection reference data to comprehensively evaluate the flavor quality of the sauce and marinade product to be tested. Specifically, the target flavor detection data is mapped to the high-dimensional feature space of flavor quality using intelligent algorithms such as pattern recognition and machine learning, and flavor characteristic indicators such as flavor intensity and flavor balance are extracted. Then, these indicators are compared with the standard values ​​specified in the flavor detection reference data to determine the flavor grade and quality level of the sauce and marinade product to be tested, and a flavor evaluation report is formed.

[0073] During the flavor evaluation process, not only should the flavor be judged as qualified or not, but also the flavor quality quantitative score and improvement suggestions should be generated. This requires comprehensive consideration of the weighted contributions of multiple flavor indicators to form a scientific flavor evaluation index system and scoring rules. At the same time, for products with unqualified flavor or deviation from the optimal state, the flavor evaluation report should also specifically point out the causes of flavor defects and improvement measures, such as adjusting the marinating process parameters, optimizing the raw material ratio, etc., to provide practical data feedback for production control and quality optimization.

[0074] Obtaining target flavor detection data through environmental calibration and evaluating flavor quality based on target data and reference data is the last step in realizing intelligent flavor perception of sauce and marinade products. This process integrates a number of key technologies such as environmental calibration, feature extraction, and pattern recognition, converting the results of flavor perception from qualitative descriptions to quantitative scores, and from simple judgments to comprehensive evaluations. Flavor evaluation data reflects the flavor quality status of sauce and marinade products in an intuitive and quantitative manner, providing an objective basis for production management and quality control, and helping to establish a closed-loop feedback mechanism for flavor perception, quality evaluation, and process optimization.

[0075] Based on the above embodiment, as an optional embodiment, the flavor detection data is calibrated based on the detection environment data to obtain the target flavor detection data, including: S401, preprocessing the detection environment data to obtain preprocessed detection environment data, and preprocessing the flavor detection data to obtain preprocessed detection environment data; Specifically, after obtaining the detection environment data and flavor detection data, the method of the present application first preprocesses the two types of data respectively to obtain the preprocessed detection environment data and the preprocessed flavor detection data. This is because the original detection data usually has different degrees of noise interference, data missing, inconsistent format and other problems. Directly using it for subsequent analysis and calibration will not only affect the calibration accuracy, but also may introduce errors and distortion. In order to ensure the quality and reliability of the detection data, it must be preprocessed and standardized as necessary.

[0076] The preprocessing of the detection environment data is mainly aimed at the environmental condition parameters during the detection of sauce and marinade flavor, such as temperature, humidity, air pressure, etc. First, the environmental parameter data is detected and eliminated for outliers to eliminate obvious abnormal data points caused by sensor failure or other factors. Then, digital filtering, wavelet transform and other methods are used to smooth and denoise the data to eliminate high-frequency noise and random interference. Next, the environmental parameter data from different sources and dimensions are normalized and mapped to a unified dimensionless scale to eliminate the impact of dimensional differences. In addition, the missing environmental parameter data needs to be interpolated and repaired to ensure the integrity and continuity of the data.

[0077] The preprocessing of flavor detection data is mainly aimed at the sauce and marinade flavor signal data collected by sensors such as electronic noses or electronic tongues. Similar to environmental parameter data, it is first necessary to detect and eliminate outliers to exclude abnormal signals that obviously deviate from the normal range. Then, digital filtering, baseline drift correction and other methods are used to reduce noise and smooth the flavor signal data to eliminate the effects of sensor drift and background interference. Next, the flavor signal data collected by multiple sensors are standardized to unify them to the same amplitude range and baseline level to eliminate the response differences between sensors. In addition, data synchronization and alignment are also required to ensure the consistency of flavor signal data collected from different batches and at different times on the timeline.

[0078] Through the above preprocessing steps, the original test environment data and flavor test data are converted into standardized, high-quality data representations, laying a good data foundation for subsequent calibration analysis. The preprocessed test environment data accurately reflects the environmental background conditions during the sauce flavor detection, while the preprocessed flavor test data truly depicts the flavor characteristic information of the sauce sample. The effective integration of these two types of data is the key to achieving environmental factor calibration and improving the accuracy of flavor detection.

[0079] S402, outputting the preprocessed detection environment data as target flavor detection data.

[0080] Specifically, after obtaining the pre-processed detection environment data and the pre-processed flavor detection data, the method of the present application inputs these two types of data into a preset calibration model, and after model calculation and correction, finally outputs the target flavor detection data. This is because the flavor detection of sauces and brisket is often carried out in a complex and changeable environment, and environmental factors such as temperature, humidity, and air pressure have a significant impact on the flavor detection results. In order to eliminate environmental interference and extract the intrinsic flavor characteristics of the sauce and brisket samples, the flavor detection data must be adaptively corrected with the help of a calibration model.

[0081] The calibration model is obtained by training a large amount of historical flavor detection data and corresponding environmental parameter data, and can characterize the intrinsic relationship between environmental factors and flavor detection results. Common calibration models include multivariate linear regression models, partial least squares regression models, neural network models, etc. These models use environmental parameters as independent variables and flavor detection data as dependent variables, and establish a quantitative mapping relationship between environmental factors and flavor detection results through mathematical function fitting or machine learning algorithms. Model parameters are trained and tuned through optimization algorithms such as least squares estimation and gradient descent, so that the model can fit the sample data to the greatest extent and has good environmental adaptive correction capabilities.

[0082] During the flavor detection process, the preprocessed detection environment data and flavor detection data are input into the preset calibration model. The model will automatically identify the current environmental parameter status and dynamically adjust the flavor detection data according to the internal parameters. Specifically, the model first extracts key parameters such as temperature, humidity, and air pressure from the detection environment data and uses them as input variables of the calibration model. At the same time, the model extracts various indicators of the flavor detection data, such as response value, response curve, characteristic peak position, etc., as the target output variables of the calibration model. Then, according to the value of the environmental parameters, the model adaptively adjusts the internal function coefficients or weight parameters, transforms and corrects the flavor detection data, and finally outputs the calibrated target flavor detection data.

[0083] The calibrated target flavor detection data has effectively deducted the influence of environmental factors and truly reflects the intrinsic flavor characteristics of the sauce and marinade samples. Regardless of the environmental conditions under which the test is conducted, the flavor detection data can maintain a high degree of consistency and comparability after adaptive adjustment of the calibration model. This data-driven adaptive calibration method overcomes the limitations of traditional empirical calibration, significantly improves the environmental adaptability and result accuracy of flavor detection, and provides reliable data support for subsequent flavor characteristic analysis and quality judgment.

[0084] Based on the above embodiment, as an optional embodiment, before the pre-processed detection environment data is input into a preset calibration model and the target flavor detection data is output, the method further includes: S501, extracting the sauce and marinade flavor sample data to obtain detection environment sample data and flavor detection sample data; Specifically, before the preprocessed detection environment data and the preprocessed flavor detection data are input into the preset calibration model, the method of the present application also needs to extract data from the sauce and marinade flavor sample data to obtain the detection environment sample data and the flavor detection sample data. This is because the training and optimization of the calibration model depends on a large amount of historical detection data, and these data are contained in the massive amount of sauce and marinade flavor sample data. In order to mine valuable environmental parameters and flavor detection indicators from the sauce and marinade flavor sample data and construct a high-quality calibration model training set, targeted data extraction and feature engineering must be performed on the sauce and marinade flavor sample data.

[0085] The first step of data extraction is to preprocess and clean the sauce and marinade flavor sample data. Sauce and marinade flavor sample data usually comes from various sensors and detection equipment on the production line. The data format is diverse and mixed with interference factors such as noise and missing values. Therefore, it is necessary to perform preprocessing operations such as format conversion, noise filtering, and missing value filling on the sample data to normalize it into a unified data representation. At the same time, data cleaning is also required to eliminate obviously abnormal or invalid sample records to ensure data quality and reliability.

[0086] After completing data preprocessing and cleaning, the detection environment sample data and flavor detection sample data are extracted from the sauce and marinade flavor sample data. The detection environment sample data mainly includes various environmental parameters for sauce and marinade flavor detection, such as temperature, humidity, air pressure, etc. By extracting and reorganizing the environmental parameter fields in the sample data, a structured detection environment sample data set is obtained. The flavor detection sample data mainly includes sauce and marinade flavor signal indicators collected by various sensors, such as electronic nose response value, electronic tongue characteristic peak position, etc. By parsing and extracting the flavor detection fields in the sample data, a structured flavor detection sample data set is obtained.

[0087] On the basis of extracting the test environment sample data and flavor test sample data, feature selection and feature construction are also required. Through feature selection algorithms such as variance analysis and correlation analysis, the environmental parameter features and flavor signal features that have the most significant impact on the flavor detection results are screened out to reduce the data dimension and cut redundant information. At the same time, based on domain knowledge and expert experience, some high-order interactive features are constructed, such as the ratio between environmental parameters and the difference between flavor signals, to explore the complex relationship between environmental factors and flavor characteristics, and further enhance the data representation ability.

[0088] Through the above data extraction steps, the sauce and marinade flavor sample data are converted into standardized and structured test environment sample data and flavor test sample data, and are optimized into information-concentrated and closely related feature subsets through feature selection and feature construction, providing high-quality data support for the training and optimization of the calibration model. These extracted sample data not only retain the key environmental background and flavor characteristics of sauce and marinade flavor detection, but also reveal the intrinsic connection between environmental factors and flavor indicators, which are the core elements for building a high-performance calibration model.

[0089] S502, calculating the Pearson correlation coefficient matrix between the detection environment sample data and the flavor detection sample data, and screening out a subset of key environmental parameters significantly correlated with each flavor according to the Pearson correlation coefficient matrix; Specifically, after obtaining the test environment sample data and the flavor test sample data, the method of the present application needs to further calculate the Pearson correlation coefficient matrix between the two types of data, and screen out the key environmental parameter subsets that are significantly correlated with each flavor based on the matrix results. This is because among the many environmental parameters, not all parameters are closely related to the flavor characteristics of the sauce and marinade. The effects of some parameters are minimal, and some parameters may even introduce noise interference. In order to optimize the input of the calibration model and improve the generalization performance of the model, it is necessary to identify the key parameter subset that contributes most to the flavor characteristics from the environmental parameters.

[0090] The Pearson correlation coefficient is a statistic that measures the degree of linear correlation between two variables, and its value range is between -1 and 1. The larger the absolute value of the coefficient, the stronger the correlation between the two variables, and the smaller the absolute value, the weaker the correlation. By calculating the Pearson correlation coefficient between each environmental parameter in the test environment sample data and each flavor index in the flavor test sample data, the strength of the association between environmental factors and the flavor characteristics of the sauce and marinade can be quantitatively characterized.

[0091] The specific calculation process is as follows: First, align the test environment sample data and the flavor test sample data according to the sample number to form paired observations of environmental parameters and flavor indicators. Then, for each environmental parameter and each flavor indicator, calculate the covariance and standard deviation of their paired observations. The covariance measures the overall error of the two variables, and the standard deviation describes the degree of dispersion of the variables. Finally, divide the covariance by the product of the standard deviations of the two variables to obtain the Pearson correlation coefficient. Repeat the above calculation process to obtain a Pearson correlation coefficient matrix of environmental parameters and flavor indicators.

[0092] According to the Pearson correlation coefficient matrix, select a subset of key environmental parameters that are significantly correlated with each flavor index. Usually, a correlation threshold can be set, such as 0.6 or 0.7, to retain environmental parameters with correlation coefficients higher than the threshold, and eliminate parameters with weaker correlation. In this way, a group of parameters that are most critical to the flavor characteristics of the sauce and marinade can be screened out from a large number of environmental parameters, highlighting strong correlation factors and weakening weak correlation factors. At the same time, attention should also be paid to the collinearity problem between environmental parameters. If the correlation coefficient of two environmental parameters is very high, it means that they contain a lot of repeated information and their effects on the flavor characteristics tend to be consistent. You can consider eliminating one of the parameters to reduce the complexity of the model.

[0093] By calculating the Pearson correlation coefficient matrix and screening key environmental parameters, the test environment sample data is condensed into the most influential and informative parameter subsets, effectively reducing the input dimension of the calibration model and reducing data redundancy and noise. This not only improves the efficiency of model training, but also enhances the model's generalization and anti-interference capabilities. Because the impact of key environmental parameters on the flavor characteristics of sauces and marinades is more stable and significant, the model can more accurately capture the inherent laws between environmental factors and flavor indicators, thereby achieving more accurate and robust calibration of flavor detection results.

[0094] S503, constructing a preliminary calibration model based on the key environmental parameter subset, the detection environment sample data, and the flavor detection sample data; Specifically, after screening out a subset of key environmental parameters, the method of the present application constructs a preliminary calibration model based on the parameter subset, the detection environment sample data, and the flavor detection sample data. This is because the preliminary calibration model is the core of the adaptive environmental calibration. It reveals the quantitative relationship between the sauce flavor detection results and environmental factors, and is an indispensable mathematical tool for flavor detection data calibration. In order to train a calibration model with superior performance, it is necessary to make full use of the screened key environmental parameters, while taking into account the overall distribution characteristics of environmental parameters and flavor indicators, and summarize the environment-flavor response law from massive historical sample data.

[0095] The construction of the preliminary calibration model mainly includes the steps of model selection, sample preparation, parameter estimation, etc. First of all, it is necessary to select the appropriate calibration model type according to the characteristics and requirements of the sauce flavor detection. Common calibration models include multiple linear regression, partial least squares regression, support vector machine, neural network, etc. Different models have their own advantages and disadvantages in terms of expression ability, computational efficiency, generalization performance, etc., which need to be comprehensively weighed. This application prefers multiple linear regression or partial least squares regression models because they are simple in structure, fast in operation, and can well characterize the linear relationship between environmental parameters and flavor indicators.

[0096] After the model selection is completed, training samples and test samples need to be prepared. The test environment sample data and flavor test sample data are randomly divided into training sets and test sets according to a certain ratio. The training set is used for model parameter estimation, and the test set is used for performance evaluation. For the training samples and test samples, the key environmental parameters obtained by screening are extracted to form the input matrix of the model. At the same time, the detection values ​​of each flavor index are extracted to form the output vector of the model.

[0097] Using the input and output data of the training samples, the parameters of the preliminary calibration model are estimated through optimization algorithms such as least squares estimation and gradient descent. Taking the multivariate linear regression model as an example, the optimization goal is to minimize the sum of squared errors between the weighted combination of environmental parameters and flavor indicators, and solve the regression coefficient to minimize the fitting error. During the model training process, cross-validation, regularization and other techniques can also be used to improve the stability and robustness of parameter estimation. Through repeated iterative optimization, the preliminary calibration model gradually converges, and the intrinsic connection between environmental parameters and flavor indicators is quantified as a set of regression equations or weight vectors.

[0098] After the preliminary calibration model is built, the calibration performance of the model is evaluated on the test samples. The key environmental parameters of the test samples are input into the preliminary calibration model to obtain the predicted values ​​of the flavor indicators, which are compared with the actual test values, and the error statistics such as mean square error and determination coefficient are calculated. By analyzing the size and distribution characteristics of the prediction error, the goodness of fit and generalization ability of the preliminary calibration model are judged, providing a basis for subsequent model optimization.

[0099] S504: training a preliminary calibration model based on the detection environment sample data and the flavor detection sample data to obtain a preset calibration model.

[0100] Specifically, after constructing the preliminary calibration model, the method of the present application further trains the preliminary calibration model based on the detection environment sample data and the flavor detection sample data to obtain a preset calibration model. This is because although the preliminary calibration model depicts the basic connection between environmental factors and the flavor characteristics of the sauce, due to the limitations of the training samples and the constraints of the model itself, its calibration and generalization performance still needs to be improved. In order to further optimize the calibration model so that it can adapt to environmental changes more accurately and robustly, it is necessary to make full use of the environmental-flavor historical sample data and adopt advanced machine learning methods to retrain and iteratively optimize the preliminary model, and finally obtain a preset calibration model with superior performance.

[0101] The specific training process is as follows: First, on the basis of the original training set, more test environment sample data and flavor test sample data are added to expand the sample size and enrich the sample distribution to improve the adequacy of data support. When supplementing samples, attention should be paid to the representativeness and balance of the data, covering various typical environmental conditions and sauce flavor types as much as possible, and avoiding model distortion caused by sample skewness. At the same time, necessary preprocessing and cleaning are performed on the supplemented sample data to remove outliers and outliers to ensure data quality.

[0102] Secondly, the input and output variables of the preliminary calibration model, namely the key environmental parameters and flavor index detection values, are extracted from the sample data and divided into training set and validation set according to a certain ratio. The training set is used for parameter optimization and model update, and the validation set is used for performance evaluation and hyperparameter selection. When dividing the data set, stratified sampling and other methods should be used to ensure the consistency of the distribution of environmental factors and flavor characteristics between the training set and the validation set to avoid overfitting and underfitting.

[0103] Then, the preliminary calibration model is retrained using the training set data through a supervised learning algorithm. Taking the multivariate linear regression model as an example, based on the original regression coefficients, the least squares estimation or gradient descent optimization algorithm is used to minimize the sum of square errors between the environmental parameter combination and the flavor index, and the regression coefficients are fine-tuned and updated to make the calibration model fit better on new samples. Regularization terms can also be introduced during the training process to control the complexity of the model and improve the model's generalization and noise resistance. Through multiple rounds of iterative training, the calibration model is gradually optimized, and the response relationship between the environment and flavor is further corrected and strengthened.

[0104] While the model is being trained, the validation set data is used to evaluate the performance of the calibration model. The validation set samples are input into the optimized calibration model to obtain the predicted values ​​of the flavor index, which are compared with the actual test values, and various performance indicators such as mean square error, determination coefficient, and mean absolute percentage error are calculated. By analyzing the prediction performance of the calibration model on the validation set, the effectiveness of the model optimization is judged, and the model's hyperparameters such as regularization coefficient and learning rate are further adjusted to find the best balance between performance and generalization ability. After multiple rounds of evaluation and tuning, a preset calibration model with better parameters and stronger generalization performance is obtained.

[0105] Finally, the performance of the preset calibration model is verified on the test set. The key environmental parameters of the test samples are input into the preset model to obtain the calibrated predicted values ​​of the sauce flavor index, which are compared with the measured values ​​to calculate comprehensive performance indicators such as the coefficient of determination and root mean square error. By analyzing the calibration effect on the test set, the practical application potential of the preset model is judged, and a performance report is generated to provide a reference for the deployment and maintenance of the model.

[0106] Based on the above embodiment, as an optional embodiment, the flavor of the sauce-marinated product to be detected is evaluated based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data, including: S601, determining a flavor deviation value based on a data difference between the target flavor detection data and the flavor detection reference data and a preset flavor difference threshold; Specifically, after obtaining the target flavor detection data and the flavor detection reference data, the method of the present application determines the flavor deviation value based on the data difference between the two types of data and the preset flavor difference threshold. This is because the flavor deviation value is a key indicator for measuring the flavor quality of the sauce-cured product to be tested, and it quantitatively describes the degree of difference between the actual flavor characteristics of the product and the standard flavor characteristics. In order to accurately evaluate the flavor state of the sauce-cured product, it is necessary to calculate the deviation amplitude of each flavor index based on the target detection data and the reference data, and combine the preset difference threshold to determine whether the flavor deviation is within an acceptable range, and then derive a comprehensive flavor score.

[0107] The specific implementation process is as follows: First, extract the values ​​of various flavor indicators in the target flavor detection data and the flavor detection reference data, including color, aroma, taste, texture state, etc. Pair the target detection value and the reference value of the same flavor indicator to form a series of actual value-standard value tuples of flavor indicators.

[0108] Secondly, for each flavor index's actual value-standard value tuple, calculate the difference between the actual value and the standard value to obtain the deviation value of the flavor index. The deviation value can be an absolute difference, that is, the absolute value of the actual value minus the standard value; or it can be a relative difference, that is, the absolute difference divided by the percentage of the standard value. The calculation method of the deviation value needs to be flexibly selected according to the characteristics of the flavor index and the evaluation requirements. By calculating the deviation values ​​of all flavor indicators, a set of flavor deviation vectors is obtained, which comprehensively characterizes the degree of deviation of the flavor characteristics of various aspects of the sauce and marinade products.

[0109] Then, the deviation value of each flavor index is compared with the preset flavor difference threshold to determine whether the flavor index exceeds the allowable deviation range. The flavor difference threshold is the upper limit set for the acceptable deviation range of each flavor index based on the quality standards of the sauce and marinated products and consumer preferences. If the deviation value is less than or equal to the difference threshold, it means that the flavor index is within the normal fluctuation range and the flavor meets the standard requirements; conversely, if the deviation value is greater than the difference threshold, it means that the flavor index deviates too much and there is a risk of exceeding the standard, which requires attention.

[0110] Finally, the flavor deviation value of the sauce-marinated product is calculated by comprehensively considering the deviation values ​​and exceeding the standard of each flavor index. The flavor deviation value can be the weighted average of the deviation values, and the weight coefficient reflects the degree of influence of different flavor indexes on the overall quality of the product; it can also be the sum of the deviation values ​​of the flavor indexes exceeding the standard, highlighting the severity of the flavor defects. The smaller the flavor deviation value, the closer the flavor characteristics of the sauce-marinated product are to the standard level and the better the quality; conversely, the larger the flavor deviation value, the farther the product flavor is from the standard reference value and the greater the quality risk.

[0111] S602, evaluating the flavor of the sauce-marinated product to be tested based on the flavor deviation value to obtain flavor evaluation data.

[0112] See also Figure 2 , Figure 2 The following is an architecture diagram of a sauce flavor detection and evaluation system provided in an embodiment of the present application. The sauce flavor detection and evaluation system may include: A data acquisition module is used to acquire the sauce and marinade flavor sample data of the sauce and marinade product to be tested, and to extract flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; A sensor selection module, used for selecting a sensor based on flavor reference data to obtain a target sensor; A sensor array building module, used for building a target sensor array based on a target sensor type and flavor reference data; A flavor detection module is used to determine flavor detection reference data based on the sauce and marinade process data; A flavor detection calibration module is used to perform flavor detection on the sauce and marinade product to be detected based on the target sensor array, and obtain flavor detection data and detection environment data of each marinating process of the sauce and marinade product to be detected; The flavor evaluation module is used to calibrate the flavor detection data based on the detection environment data to obtain target flavor detection data, and to evaluate the flavor of the sauce-marinated product to be detected based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data.

[0113] It should be noted that: when the system provided in the above embodiment realizes its functions, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0114] Please refer to Figure 3 The application also discloses an electronic device. Figure 3 The electronic device 300 may include: at least one processor 301, at least one network interface 304, a user interface 303, a memory 305, and at least one communication bus 302 or end-to-end wireless communication.

[0115] The communication bus 302 is used to realize the connection and communication between these components.

[0116] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0117] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0118] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field~Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0119] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read~Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally be at least one storage system located away from the aforementioned processor 301. Refer to Figure 3 , the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a sauce flavor detection and evaluation method.

[0120] exist Figure 3In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the application program storing the sauce and marinade flavor detection and evaluation method in the memory 305, and when executed by one or more processors 301, the electronic device 300 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited to the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application. In the above embodiments, the descriptions of each embodiment have their own emphasis. For the parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0121] In the several implementations provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are only schematic, such as the division of modules, which is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, indirect coupling or communication connection of systems or modules, which can be electrical or other forms.

[0122] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0123] The present application also provides a computer storage medium that can store multiple instructions, which are suitable for being loaded and executed by a processor as described above. Figure 1 The method for detecting and evaluating the flavor of the sauce and marinade in the embodiment shown in the figure can be found in the specific implementation process. Figure 1 The specific description of the illustrated embodiment will not be repeated here.

[0124] In addition, each functional module in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software functional modules.

[0125] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0126] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0127] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for detecting and evaluating the flavor of soy sauce and marinade, characterized in that: The method comprises: Acquire sauce and marinade flavor sample data of the sauce and marinade product to be tested, and extract flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; Selecting a sensor based on the flavor reference data to obtain a target sensor; constructing a target sensor array based on the target sensor type and the flavor reference data; Based on the sauce and marinade process data, determining flavor detection reference data; Performing flavor detection on the to-be-detected marinated product based on the target sensor array to obtain flavor detection data and detection environment data of each marinating process of the to-be-detected marinated product; The flavor detection data is calibrated based on the detection environment data to obtain target flavor detection data, and the flavor of the to-be-detected marinated product is evaluated based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data.

2. The method according to claim 1, characterized in that The step of selecting a sensor based on the flavor reference data to obtain a target sensor comprises: According to the flavor reference data, a flavor key component is determined, and according to the flavor key component, a sensor that is sensitive to the flavor key component is screened to obtain the target sensor.

3. The method according to claim 1, characterized in that: The step of constructing a target sensor array based on the target sensor type and the flavor reference data includes: determining a quantity of each of the target sensor types based on the flavor reference data; The target sensor array is obtained by arranging the target sensors in a preset arrangement array based on the quantity of each target sensor type.

4. The method according to claim 1, characterized in that: The extracting of flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data includes: Preprocessing the sauce and marinade flavor sample data to obtain preprocessed sauce and marinade flavor sample data, and performing feature extraction based on the preprocessed sauce and marinade flavor sample data to obtain flavor sample data and sauce and marinade process sample data; Performing statistical screening on the flavor sample data to obtain the flavor reference data; The sample data of the sauce and marinade process are statistically screened to obtain the sauce and marinade process data.

5. The method according to claim 1, characterized in that The step of calibrating the flavor detection data based on the detection environment data to obtain target flavor detection data includes: Preprocessing the detection environment data to obtain preprocessed detection environment data, and preprocessing the flavor detection data to obtain preprocessed detection environment data; The preprocessed detection environment data is input into a preset calibration model, and the target flavor detection data is output.

6. The method according to claim 5, characterized in that Before inputting the preprocessed detection environment data into a preset calibration model and outputting the target flavor detection data, the method further includes: Extracting data from the sauce and marinade flavor sample data to obtain detection environment sample data and flavor detection sample data; Calculating a Pearson correlation coefficient matrix between the detection environment sample data and the flavor detection sample data, and screening out a subset of key environmental parameters significantly correlated with each flavor based on the Pearson correlation coefficient matrix; constructing a preliminary calibration model based on the key environmental parameter subset, the detection environment sample data, and the flavor detection sample data; The preliminary calibration model is trained based on the detection environment sample data and the flavor detection sample data to obtain the preset calibration model.

7. The method according to claim 1, characterized in that The flavor of the to-be-detected marinated product is evaluated based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data, including: determining a flavor deviation value based on a data difference between the target flavor detection data and the flavor detection reference data and a preset flavor difference threshold; The flavor of the sauce-marinated product to be tested is evaluated based on the flavor deviation value to obtain the flavor evaluation data.

8. A sauce and marinade flavor detection and evaluation system, characterized in that: The system comprises: A data acquisition module, used to acquire sauce and marinade flavor sample data of the sauce and marinade product to be tested, and extract flavor reference data and sauce and marinade process data based on the sauce and marinade flavor sample data; A sensor selection module, configured to select a sensor based on the flavor reference data to obtain a target sensor; A sensor array construction module, used to construct a target sensor array based on the target sensor type and the flavor reference data; A flavor detection module, used to determine flavor detection reference data based on the sauce and marinade process data; A flavor detection calibration module, used to perform flavor detection on the sauce-marinated product to be detected based on the target sensor array, and obtain flavor detection data and detection environment data of each marinating process of the sauce-marinated product to be detected; The flavor evaluation module is used to calibrate the flavor detection data based on the detection environment data to obtain target flavor detection data, and to evaluate the flavor of the to-be-detected marinated product based on the target flavor detection data and the flavor detection reference data to obtain flavor evaluation data.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes a processor, a memory and a transceiver, the memory is used to store instructions, the transceiver is used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

Images