Mixing proportion detection control method and system for food preparation

Abstract

The invention discloses a mixing proportion detection control method and system for food preparation, and relates to the technical field of food processing.The mixing proportion detection control method for food preparation comprises the following steps that mixing equipment is used for mixing food raw materials according to a preset proportion, in the mixing process, a multi-camera module is used for carrying out image acquisition; comprehensively evaluating mixing uniformity by combining raw material mixture characteristics and raw material physical properties; making a sampling inspection time window to perform sampling inspection analysis on the raw material mixture; continuously collecting configuration data under different environmental conditions, and generating a plurality of groups of food configuration schemes; extracting an optimal food configuration scheme by using a multi-target optimization algorithm; and constructing a food risk assessment model, and judging whether the optimal food configuration scheme has a food safety risk or not. According to the method, various risk factors possibly existing in the food configuration scheme can be effectively identified and evaluated, so that potential food safety problems can be found as early as possible, and occurrence of food safety accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of food processing, and in particular to a mixing ratio detection and control method and system for food configuration. Background Art

[0002] Food configuration in food production involves mixing and processing different food raw materials according to certain formulas and proportions to prepare specific food products. This process is a key step in producing specialty foods, meeting market demand, and improving product competitiveness. By mixing food raw materials according to specific formulas and proportions, not only can special foods be prepared, but also raw materials can be used to the greatest extent, waste can be reduced, and the efficiency of the production process can be ensured.

[0003] In the process of food preparation, the mixing ratio is one of the key factors affecting food quality. In order to ensure that the mixing ratio of each batch of food meets the standards and formula requirements, it is particularly important to formulate a mixing ratio detection and control method. This can ensure the stability and consistency of food quality, while helping companies reduce production costs, improve production efficiency, and achieve effective cost control. By formulating a detection and control method, the mixing ratio can be kept consistent in each production, ensuring that the final product remains consistent in taste, appearance, and nutritional content, improving brand image and user satisfaction, and optimizing the production process. Improving production efficiency is also one of the purposes of formulating a mixing ratio detection and control method.

[0004] However, after mixing food raw materials in a certain proportion, there may be uneven mixing of food raw materials. This may cause certain areas to contain excessive or insufficient amounts of a certain ingredient, thereby causing safety hazards, such as too high or too low additive content. In addition, uneven mixing may also lead to poor product consistency and inability to ensure the consistency of taste and texture of each batch of products, thereby affecting consumers' trust and loyalty to the brand. In summary, the inability to evaluate the defect of mixing uniformity may lead to unstable food quality and reduce food processing efficiency and quality.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes a mixing ratio detection and control method and system for food configuration to overcome the above technical problems existing in the existing related art.

[0007] To this end, the specific technical solution adopted by the present invention is as follows: According to one aspect of the present invention, a mixing ratio detection control method for food configuration is provided, and the mixing ratio detection control method for food configuration comprises the following steps: S1. Mix the food ingredients according to a preset ratio using a mixing device, and use a multi-camera module to collect images during the mixing process; S2, extracting raw material mixture features from the image set, and comprehensively evaluating the mixing uniformity by combining the raw material mixture features and raw material physical properties; S3. Establish a sampling time window to conduct sampling analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted based on the sampling analysis results; S4, adding new raw materials to the raw material mixture according to the adjusted ratio and performing secondary mixing, while continuously collecting configuration data under different environmental conditions to generate multiple groups of food configuration plans; S5. Compare and analyze multiple groups of food configuration plans, and use a multi-objective optimization algorithm to extract the optimal food configuration plan; S6. Construct a food risk assessment model to determine whether the optimal food configuration plan poses a food safety risk, and formulate targeted preventive measures based on the food safety risk.

[0008] Preferably, extracting raw material mixture features from the image set and comprehensively evaluating the mixing uniformity based on the raw material mixture features and raw material physical properties comprises the following steps: S21, extracting color distribution and texture features of the raw material mixture from the image set, and judging whether the raw material mixture meets the uniformity requirement based on the color distribution and texture features; S22, if the raw material mixture meets the uniformity requirement, execute step S3, if the raw material mixture does not meet the uniformity requirement, execute step S23; S23, estimating the raw material properties of each component in the raw material mixture by using the prior knowledge base, and performing hierarchical weighting on each raw material property to obtain a level characteristic value of each raw material property; S24, calculating the information entropy of each raw material property by using the entropy weight method, and converting the information entropy into a relative entropy value; S25, combining the level characteristic value and the relative entropy value to obtain the characteristic value of the influence of each raw material property on the raw material mixture, sorting the influencing characteristic values, and taking the raw material properties at the top of the sorting position as the main influencing factors; S26. Formulate an adjustment strategy based on the main influencing factors, and continue to mix the raw material mixture that does not meet the uniformity requirements according to the adjustment strategy.

[0009] Preferably, extracting the color distribution and texture features of the raw material mixture from the image set, and judging whether the raw material mixture meets the uniformity requirement based on the color distribution and texture features comprises the following steps: S211, converting each image in the image set into a YCbCr color space, identifying a raw material mixture region according to an image color rule, and then extracting color distribution characteristics in the raw material mixture region; S212, dividing each image into a plurality of image sub-blocks, and using a local binary pattern algorithm to calculate the binary value of each pixel in each image sub-block, and performing histogram statistics on each image sub-block to obtain a plurality of histograms; S213, forming an image feature vector from a plurality of histograms, and describing the texture feature of the image through the image feature vector; S214, using short-range Fourier transform to calculate the frequency domain phase information of each pixel on the image, and using a local phase quantization algorithm to extract texture features through the frequency domain phase information; S215, combining the texture features and color distribution features extracted by the local binary pattern algorithm and the local phase quantization algorithm respectively to form a comprehensive image feature vector, and inputting the comprehensive image feature vector into a support vector machine for uniformity testing; S216. If the color distribution characteristics and the texture characteristics are relatively uniform, it means that the raw material mixture meets the uniformity requirements. If the color distribution characteristics and the texture characteristics are different, it means that the raw material mixture does not meet the uniformity requirements.

[0010] Preferably, estimating the raw material properties of each component in the raw material mixture by using a priori knowledge base, and performing hierarchical weighting on each raw material property to obtain the level characteristic value of each raw material property comprises the following steps: S231, collecting the raw material properties of each food raw material in the raw material mixture, and establishing a priori knowledge base based on the raw material properties, wherein the raw material properties include but are not limited to the particle size, shape, density and moisture content of each food raw material; S232. Classify and combine the raw material properties to form a structured raw material property index system; S233, constructing a judgment matrix based on the raw material property index system, wherein the judgment matrix is ​​composed of the quantitative index values ​​of the raw material properties and the superiority of the qualitative indexes; S234, normalizing the judgment matrix to obtain a normalized judgment matrix, and calculating the raw material physical property elements in the normalized judgment matrix; S235, determining the raw material physical property index value based on the raw material physical property elements, and calculating the relative superiority vector in each layer structure using a multi-level optimization theory algorithm; S236. The relative superiority vector calculated in each layer of the structure is used as the input of the next layer of the structure, and step S235 is repeated to obtain the relative superiority vector of the highest layer of the structure, and use it as the level characteristic value of the raw material property.

[0011] Preferably, comparing and analyzing multiple groups of food configuration schemes and extracting the optimal food configuration scheme using a multi-objective optimization algorithm includes the following steps: S51, obtaining configuration data of each food configuration plan, the configuration data including raw material component ratio, mixing equipment parameters and mixing environment; S52. Determine evaluation indicators for comparative analysis, and use the evaluation indicators to evaluate the performance of each food configuration solution; S53. Use the chaotic sequence generated by chaotic mapping to replace the randomly generated initial population, and screen out the optimal food configuration plan based on population crowding and non-dominated sorting.

[0012] Preferably, using a chaotic sequence generated by a chaotic map to replace the randomly generated initial population, and selecting the optimal food configuration scheme according to the population crowding and non-dominated sorting includes the following steps: S531, taking the number of food configuration solutions as the target number, and using chaotic mapping to generate a chaotic sequence as the initial population of the dynamic optimization strategy; S532, calculating the fitness value of each food configuration scheme in the initial population, and defining the order value and aggregation distance of each food configuration scheme according to the fitness value and the food dominance relationship; S533, calculating the initial population based on the order value and the focal distance to obtain the parent population, and then performing crossover and mutation operations on the parent population to obtain the offspring population; S534, merging the initial population and the parent population to obtain a new population, and using a dynamic optimization strategy to judge all food configuration schemes in the new population, and retaining the food configuration scheme if it satisfies the dominance relationship of the dynamic optimization strategy; S535. Perform non-dominated sorting on all food configuration plans in the new population in combination with the population crowding degree, and select the optimal configuration plan according to the non-dominated sorting result.

[0013] Preferably, building a food risk assessment model, determining whether the optimal food configuration plan has food safety risks, and formulating targeted preventive measures based on food safety risks include the following steps: S61. Carry out food simulation configuration according to the optimal food configuration plan, and build a food risk assessment model based on the food simulation configuration results; S62. Use the food risk assessment model to establish a priority relationship between potential food risk event assessment information and risk level classification thresholds; S63. Determine the risk level of food risk events based on the established priority relationship; S64. Convert the food risk assessment model into a Bayesian model, use the Bayesian model to predict the probability of occurrence of food risk events through forward reasoning, and combine it with reverse reasoning to conduct risk screening; S65. Reversely identify the key sources that trigger food risk events, and formulate targeted preventive measures based on the risk level of food risk events.

[0014] Preferably, converting the food risk assessment model into a Bayesian model, using the Bayesian model to predict the probability of occurrence of food risk events by forward reasoning, and combining reverse reasoning to perform risk screening includes the following steps: S641. Convert the food risk assessment model into a Bayesian model, and use the node transfer and conditional probability in the Bayesian model to calculate the probability of all node state combinations; S642. Dynamically analyze the logical relationship of each food risk event, food safety weaknesses and key causal paths based on the probability of all node status combinations; S643. Predict the probability of occurrence of food risk events using a joint probability distribution algorithm, and traverse the risk node combinations that affect food risk events in a Bayesian model, and calculate the posterior probability of the risk node combinations; S644. Confirm the specific risk nodes based on the posterior probability of the risk node combination, determine the food risk path based on the fault node, and conduct risk investigation on the food risk path in combination with reverse reasoning.

[0015] Preferably, the calculation formula of the raw material physical property elements in the normalized judgment matrix is: ; In the formula, Represents the first i Line j List the physical properties of raw materials; Indicates i Line j Quantitative index values ​​of raw material properties listed; i Indicates the number of rows of the normalized judgment matrix; j Represents the number of columns of the normalized judgment matrix.

[0016] According to another aspect of the present invention, there is also provided a mixing ratio detection and control system for food configuration, the mixing ratio detection and control system for food configuration comprising an image acquisition module, a mixing uniformity evaluation module, a sampling analysis module, a configuration scheme generation module, an optimal configuration scheme selection module and a food risk assessment module; The image acquisition module is connected to the mixing uniformity assessment module, the mixing uniformity assessment module is connected to the sampling analysis module, the sampling analysis module is connected to the configuration scheme generation module, the configuration scheme generation module is connected to the optimal configuration scheme selection module, and the optimal configuration scheme selection module is connected to the food risk assessment module; An image acquisition module, used to mix food ingredients according to a preset ratio using a mixing device, and to acquire images using a multi-camera module during the mixing process; A mixing uniformity assessment module is used to extract the raw material mixture characteristics from the image set and comprehensively assess the mixing uniformity by combining the raw material mixture characteristics and raw material physical properties; The sampling and analysis module is used to set a sampling time window to conduct sampling and analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted based on the sampling and analysis results; A configuration scheme generation module is used to add new raw materials to the raw material mixture according to the adjusted ratio and perform secondary mixing, and continuously collect configuration data under different environmental conditions to generate multiple groups of food configuration schemes; The optimal configuration scheme selection module is used to compare and analyze multiple groups of food configuration schemes and extract the optimal food configuration scheme using a multi-objective optimization algorithm; The food risk assessment module is used to build a food risk assessment model, determine whether the optimal food configuration plan has food safety risks, and formulate targeted preventive measures based on food safety risks.

[0017] The beneficial effects of the present invention are: 1. The present invention can monitor the distribution of raw materials in the mixing process in real time by extracting the characteristics of the raw material mixture from the image set, help to promptly discover problems that may cause non-uniformity, and combine the characteristics of the raw material mixture and the physical properties of the raw materials to achieve a multi-dimensional evaluation of the mixing uniformity, which is helpful to judge the mixing quality more comprehensively and accurately, and improve the control level of the food configuration process. At the same time, by effectively evaluating the mixing uniformity, it is helpful to improve the consistency of food products, ensure that the quality of each batch of finished products is relatively stable, reduce dependence on manual judgment, and thus improve the reliability of food configuration.

[0018] 2. The present invention adds new raw materials to the raw material mixture according to the adjusted ratio and performs secondary mixing, which can flexibly respond to different ratio requirements and improve the flexibility of the production process. At the same time, the configuration data under different environmental conditions are continuously collected, which is helpful to fully understand the configuration performance under different working conditions and improve the robustness of the configuration scheme. A comprehensive comparative analysis of multiple groups of food configuration schemes is carried out through a multi-objective optimization algorithm to help determine the optimal configuration scheme, improve food configuration efficiency and quality, and consider multiple optimization objectives at the same time to ensure that the food configuration meets higher quality standards, thereby helping to improve food production efficiency and food production quality.

[0019] 3. By constructing a food risk assessment model, the present invention can effectively identify and evaluate various risk factors that may exist in food configuration plans, thereby helping to discover potential food safety issues early and reduce the occurrence of food safety accidents. At the same time, by evaluating the food risks in the optimal food configuration plan, it can ensure that the selected optimal configuration plan is also extremely reliable in terms of safety, which can not only improve the overall food quality, but also formulate more effective prevention and control measures based on the risk assessment results, thereby reducing food safety risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 is a flow chart of a mixing ratio detection and control method for food configuration according to an embodiment of the present invention; Figure 2 It is a principle block diagram of a mixing ratio detection control system for food configuration according to an embodiment of the present invention.

[0022] In the figure: 1. Image acquisition module; 2. Mixing uniformity assessment module; 3. Sampling analysis module; 4. Configuration scheme generation module; 5. Optimal configuration scheme selection module; 6. Food risk assessment module. DETAILED DESCRIPTION

[0023] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0024] According to an embodiment of the present invention, a mixing ratio detection control method and system for food configuration are provided.

[0025] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the mixing ratio detection and control method for food configuration according to an embodiment of the present invention, the mixing ratio detection and control method for food configuration includes the following steps: S1. Mix food ingredients according to a preset ratio using a mixing device, and use a multi-camera module to collect images during the mixing process.

[0026] It should be noted that mixing equipment includes mixing barrels, blenders and mixers, etc. The food raw materials are mixed according to preset proportions through the mixing equipment, and multi-camera modules are used to collect images during the mixing process to achieve real-time monitoring and image collection of the food raw material mixing process. The mixing equipment ensures mixing according to preset proportions, and the multi-camera module provides multi-angle and multi-viewpoint image information, which helps to have a more comprehensive understanding of the state of food raw materials during the mixing process.

[0027] S2. Extract the raw material mixture characteristics from the image set, and comprehensively evaluate the mixing uniformity by combining the raw material mixture characteristics and raw material physical properties.

[0028] The steps of extracting the raw material mixture features from the image set and comprehensively evaluating the mixing uniformity by combining the raw material mixture features and the raw material physical properties include: S21. Extracting color distribution and texture features of the raw material mixture from the image set, and judging whether the raw material mixture meets the uniformity requirement based on the color distribution and texture features.

[0029] The steps of extracting the color distribution and texture features of the raw material mixture from the image set and judging whether the raw material mixture meets the uniformity requirements based on the color distribution and texture features include the following steps: S211, converting each image in the image set into a YCbCr color space, identifying a raw material mixture region according to an image color rule, and then extracting color distribution features in the raw material mixture region.

[0030] It should be noted that the YCbCr color space is a color encoding algorithm used to describe color images. It separates the brightness information (Y) and chromaticity information (Cb and Cr) of the image. The Y component (brightness) represents the brightness information of the image, which represents the grayscale level. The higher the value, the brighter the pixel; the Cb and Cr components (chromaticity) represent the difference information of blue chromaticity and red chromaticity respectively. The use of YCbCr color space can independently compress the brightness and chromaticity information, thereby improving the compression efficiency.

[0031] S212, dividing each image into a number of image sub-blocks, and using a local binary pattern algorithm to calculate the binary value of each pixel in each image sub-block, and performing histogram statistics on each image sub-block to obtain a number of histograms.

[0032] S213, forming an image feature vector from a plurality of histograms, and describing the texture features of the image through the image feature vector.

[0033] S214, using short-range Fourier transform to calculate the frequency domain phase information of each pixel on the image, and using a local phase quantization algorithm to extract texture features through the frequency domain phase information.

[0034] It should be noted that the short-range Fourier transform is a method of converting a signal (or image) from the time domain to the frequency domain. For an image, STFT can be used to calculate the frequency domain information of each pixel on the image. STFT divides the signal into small time periods and applies Fourier transform in each time period to obtain frequency information.

[0035] For each image pixel, STFT is used to obtain its frequency domain representation, where phase information is a part of the STFT output. This phase information contains the phase change in the frequency domain and reflects the structural characteristics of the image in the frequency domain. Based on the frequency domain phase information, the local phase quantization algorithm is used to capture the texture characteristics of the image. The algorithm involves dividing the frequency domain phase information into local regions and calculating the statistical characteristics of the phase in these regions, such as local phase patterns or other features that describe the phase information.

[0036] S215, combining the texture features and color distribution features extracted by the local binary pattern algorithm and the local phase quantization algorithm respectively to form a comprehensive image feature vector, and inputting the comprehensive image feature vector into a support vector machine for uniformity testing.

[0037] It should be noted that the local binary pattern algorithm is used to extract texture features. It describes the local texture of an image by comparing the grayscale values ​​of a pixel with its neighboring pixels. For each pixel, a binary number is obtained based on the grayscale values ​​of its surrounding pixels to form a local texture pattern.

[0038] The local phase quantization algorithm is used to extract the phase information of an image. It usually involves decomposing the image into different frequency components and then calculating the phase of each frequency component, which helps capture the detailed information in the image.

[0039] S216. If the color distribution characteristics and the texture characteristics are relatively uniform, it means that the raw material mixture meets the uniformity requirements. If the color distribution characteristics and the texture characteristics are different, it means that the raw material mixture does not meet the uniformity requirements.

[0040] S22. If the raw material mixture meets the uniformity requirement, execute step S3; if the raw material mixture does not meet the uniformity requirement, execute step S23.

[0041] S23. Estimate the raw material properties of each component in the raw material mixture using a priori knowledge base, and perform hierarchical weighting on each raw material property to obtain a level characteristic value for each raw material property.

[0042] Among them, using the prior knowledge base to estimate the raw material properties of each component in the raw material mixture, and hierarchically assigning weights to each raw material property to obtain the level characteristic value of each raw material property includes the following steps: S231. Collect the raw material properties of each food raw material in the raw material mixture, and establish a priori knowledge base based on the raw material properties. The raw material properties include but are not limited to the particle size, shape, density and moisture content of each food raw material.

[0043] S232. Classify and combine the physical properties of raw materials to form a structured raw material physical property index system.

[0044] S233. Construct a judgment matrix based on the raw material property index system. The judgment matrix is ​​composed of the quantitative index values ​​of the raw material properties and the superiority of the qualitative indexes.

[0045] It should be noted that the construction of a judgment matrix based on the raw material property index system, which is composed of the quantitative index values ​​of the raw material properties and the superiority of the qualitative indexes, includes the following steps: Determine the physical properties related to the raw materials. Physical properties can include quantitative indicators (such as temperature, concentration, pH value, etc.) and qualitative indicators (such as color, taste, appearance, etc.).

[0046] For qualitative indicators, the fuzzy set theory in fuzzy mathematics is used to convert qualitative indicators into numerical values.

[0047] The obtained quantitative index values ​​and the superiority of qualitative indexes are combined into a judgment matrix, the rows and columns of which correspond to the raw material physical property indexes, and the matrix elements represent the relative importance or superiority relationship between the corresponding indexes.

[0048] Obtain information on relative importance or superiority / inferiority relationships through literature research or experimental measurements to fill in the elements of the judgment matrix.

[0049] Based on the results of the consistency test, the weight of each indicator is calculated to reflect its relative importance in the judgment.

[0050] S234, normalizing the judgment matrix to obtain a normalized judgment matrix, and calculating the raw material physical property elements in the normalized judgment matrix.

[0051] Among them, the calculation formula of the raw material physical property elements in the normalized judgment matrix is: ; In the formula, Represents the first i Line j List the physical properties of raw materials; Indicates i Line j Quantitative index values ​​of raw material properties listed; i Indicates the number of rows of the normalized judgment matrix; j Represents the number of columns of the normalized judgment matrix.

[0052] S235. Determine the raw material physical property index value based on the raw material physical property elements, and use a multi-level optimization theory algorithm to calculate the relative superiority vector in each layer structure.

[0053] S236. The relative superiority vector calculated in each layer of the structure is used as the input of the next layer of the structure, and step S235 is repeated to obtain the relative superiority vector of the highest layer of the structure, and use it as the level characteristic value of the raw material property.

[0054] S24. Calculate the information entropy of each raw material property using the entropy weight method, and convert the information entropy into a relative entropy value.

[0055] It should be noted that the entropy weight method is used to calculate the information entropy of each raw material property, and the information entropy is converted into a relative entropy value, which includes the following steps: For each raw material property index, first calculate its information entropy, which is a measure of information uncertainty; use the entropy weight method to calculate the weight of each raw material property, and convert the information entropy into a relative entropy value. Through this calculation process, the relative entropy value of each raw material property is obtained, which reflects its relative contribution or importance in the entire physical property index system. This method takes into account the information entropy of each indicator and assigns weights according to the size of the information entropy, so that the contribution of different indicators can be considered more scientifically in decision-making.

[0056] S25. Combining the level characteristic value and the relative entropy value, obtain the characteristic value of the influence of each raw material property on the raw material mixture, sort the influencing characteristic values, and take the raw material properties at the top of the sorting position as the main influencing factors.

[0057] S26. Formulate an adjustment strategy based on the main influencing factors, and continue to mix the raw material mixture that does not meet the uniformity requirements according to the adjustment strategy.

[0058] It should be noted that formulating an adjustment strategy based on the main influencing factors and continuing to mix the raw material mixture that does not meet the uniformity requirements according to the adjustment strategy includes the following steps: To address the problem of mixture inhomogeneity, we first need to identify the main factors that cause the inhomogeneity, which may involve analyzing and evaluating various variables in the mixing process, equipment parameters, raw material characteristics, etc.

[0059] Based on the identified main influencing factors, formulate adjustment strategies to solve the inhomogeneity problem, including adjusting the operating parameters of the mixing equipment (such as stirring speed, time, temperature, etc.), changing the order of raw material addition, or optimizing other key steps in the mixing process.

[0060] The raw material mixture that does not meet the uniformity requirements will be mixed again through the mixing equipment according to the formulated adjustment strategy to ensure that the mixture reaches the required uniformity standard.

[0061] Perform quality checks during the mixing process to ensure that the adjusted mixture meets the expected uniformity requirements. If uniformity issues still exist, it may be necessary to re-evaluate the adjustment strategy and perform further adjustments and mixing operations.

[0062] S3. Establish a sampling time window to conduct sampling analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted based on the sampling analysis results.

[0063] It should be noted that setting a sampling time window to conduct sampling analysis on the raw material mixture, and judging whether the preset ratio needs to be adjusted according to the sampling analysis results includes the following steps: Within a determined time window, samples of the mixture are taken for analysis. The samples should be representative and cover different parts of the mixture to ensure that the sampling results can reflect the uniformity of the entire mixture.

[0064] The quality of the sample is analyzed by physical testing, chemical analysis or other relevant methods. The goal of the analysis is to evaluate the homogeneity of the mixture and whether it meets the predetermined quality standards.

[0065] Compare the actual mixing ratio in the sampling results with the preset ratio. If the actual mixing ratio is significantly different from the preset ratio, it may be necessary to consider adjusting the operating parameters of the mixing equipment or other relevant factors.

[0066] Based on the results of random inspection and analysis, determine whether the preset ratio needs to be adjusted. If the raw material mixture does not meet the standard or deviates from expectations, take corresponding adjustment measures, which include adjusting the parameters of the mixing equipment, changing the amount or order of raw material addition, etc.

[0067] S4. Add new raw materials to the raw material mixture according to the adjusted ratio and perform secondary mixing, while continuously collecting configuration data under different environmental conditions to generate multiple groups of food configuration plans.

[0068] It should be noted that adding new raw materials to the raw material mixture according to the adjusted ratio and performing secondary mixing, while continuously collecting configuration data under different environmental conditions, and generating multiple groups of food configuration plans include the following steps: Determine the new raw material ratio based on the sampling analysis results and the need for adjustment, and ensure the accuracy of the ratio adjustment.

[0069] According to the adjustment ratio, add new raw materials to the mixing equipment to ensure that the new raw materials are evenly mixed into the entire mixture, perform secondary mixing on the mixture to ensure that the newly added raw materials are fully mixed with the existing raw materials, and adjust the operating parameters of the mixing equipment, such as stirring speed, mixing time, etc., to ensure the mixing effect.

[0070] Continuously collect configuration data under different environmental conditions, which may include factors such as temperature, humidity, air pressure, and the operating status of the mixing equipment, to ensure the comprehensiveness and representativeness of the collected data.

[0071] Using the collected configuration data, statistical analysis or machine learning techniques are used to generate multiple groups of food configuration plans, and the generated food configuration plans are evaluated, considering their stability and adaptability under different environmental conditions.

[0072] S5. Compare and analyze multiple groups of food configuration plans, and use a multi-objective optimization algorithm to extract the optimal food configuration plan.

[0073] Among them, the comparative analysis of multiple groups of food configuration schemes and the use of a multi-objective optimization algorithm to extract the optimal food configuration scheme include the following steps: S51. Obtain configuration data of each food configuration plan, where the configuration data includes raw material component ratio, mixing equipment parameters and mixing environment.

[0074] S52. Determine the evaluation indicators for comparative analysis, and use the evaluation indicators to evaluate the performance of each food configuration plan.

[0075] S53. Use the chaotic sequence generated by chaotic mapping to replace the randomly generated initial population, and screen out the optimal food configuration plan based on population crowding and non-dominated sorting.

[0076] It should be noted that the chaotic sequence generated by chaotic mapping is used to replace the randomly generated initial population. Since the chaotic sequence has a pseudo-random nature and the sequence generated by chaotic mapping exhibits the characteristics of deterministic chaos, it can replace the traditional random number generation method in some cases and provide a more complex and richer search space.

[0077] In optimization problems such as genetic algorithms, the initial state of the population has a certain impact on the performance of the algorithm. Traditionally, a randomly generated initial population is used to explore the search space. However, the sequence generated by chaotic mapping has stronger randomness and dispersion, which helps to explore the solution space of the problem more comprehensively.

[0078] The chaotic sequence generated by the chaotic mapping is used to replace the randomly generated initial population, and the optimal food configuration scheme is selected according to the population crowding and non-dominated sorting, which includes the following steps: S531. Taking the number of food configuration solutions as the target number, and using chaotic mapping to generate a chaotic sequence as the initial population of the dynamic optimization strategy.

[0079] S532, calculating the fitness value of each food configuration scheme in the initial population, and defining the order value and aggregation distance of each food configuration scheme according to the fitness value and the food dominance relationship.

[0080] It should be noted that calculating the fitness value of each food configuration scheme in the initial population and defining the order value and clustering distance of each food configuration scheme according to the fitness value and the food dominance relationship includes the following steps: For each food configuration, a fitness function is defined to evaluate its performance. The fitness function can make trade-offs based on factors such as the quality of the configuration, production cost, and market demand.

[0081] For each food configuration scheme in the initial population, the defined fitness function is applied to calculate and obtain the corresponding fitness value. The higher the fitness value, the better the configuration scheme.

[0082] The food dominance relationship is determined according to the fitness value, that is, which configuration schemes dominate other schemes in performance, and the order value is calculated for each configuration scheme based on the food dominance relationship. The order value represents the relative position of a configuration scheme in the population.

[0083] For each configuration scheme, its clustering distance with other schemes that dominate it is calculated. The clustering distance can be measured by considering the difference between the fitness value and the order value, reflecting the concentration of the configuration scheme in the population.

[0084] S533. Calculate the initial population based on the order value and the focal distance to obtain the parent population, and then perform crossover and mutation operations on the parent population to obtain the offspring population.

[0085] S534, merging the initial population and the parent population to obtain a new population, and using a dynamic optimization strategy to judge all food configuration schemes in the new population. If the dominance relationship of the dynamic optimization strategy is satisfied, the food configuration scheme is retained.

[0086] S535. Perform non-dominated sorting on all food configuration plans in the new population in combination with the population crowding degree, and select the optimal configuration plan according to the non-dominated sorting result.

[0087] S6. Construct a food risk assessment model to determine whether the optimal food configuration plan poses a food safety risk, and formulate targeted preventive measures based on the food safety risk.

[0088] Among them, building a food risk assessment model, determining whether the optimal food configuration plan has food safety risks, and formulating targeted preventive measures based on food safety risks include the following steps: S61. Carry out food simulation configuration according to the optimal food configuration plan, and build a food risk assessment model based on the food simulation configuration results.

[0089] S62. Use the food risk assessment model to establish a priority relationship between potential food risk event assessment information and risk level classification thresholds.

[0090] It should be noted that the use of food risk assessment models to establish a priority relationship between potential food risk event assessment information and risk level classification thresholds includes the following steps: Define and establish the assessment information indicators required for the food risk assessment model. These indicators may include but are not limited to food ingredients, production process, raw material sources and supply chain traceability.

[0091] Select or design a food risk assessment model suitable for specific situations and determine the threshold for dividing food risk levels, that is, set standards for different risk levels.

[0092] The relationship between the output results of the assessment model and the set risk level classification thresholds is used to establish priority relationship rules. These rules can be logical judgments based on thresholds or complex combinations of rules to determine the priority level of potential food risk events.

[0093] Consider the importance of different assessment information indicators and assign appropriate weights to them to ensure that the actual situation of the risk can be more accurately reflected when establishing priority relationships.

[0094] Develop response strategies based on the established priority relationship rules, including taking preventive measures, strengthening monitoring and updating production processes to reduce or eliminate potential food risks.

[0095] S63. Determine the risk level of food risk events based on the established priority relationship.

[0096] S64. Convert the food risk assessment model into a Bayesian model, use the Bayesian model to perform forward reasoning to predict the probability of occurrence of food risk events, and combine it with reverse reasoning to conduct risk screening.

[0097] The following steps are involved in converting the food risk assessment model into a Bayesian model, using the Bayesian model to predict the probability of occurrence of food risk events through forward reasoning, and combining reverse reasoning to conduct risk screening: S641. Convert the food risk assessment model into a Bayesian model, and use the node transfer and conditional probability in the Bayesian model to calculate the probability of all node state combinations.

[0098] S642. Dynamically analyze the logical relationship, weak links in food safety and key causal paths of each food risk event based on the probability of all node status combinations.

[0099] It should be noted that the logical relationship of each food risk event, food safety weaknesses and key causal paths based on the probability dynamic analysis of all node status combinations include the following steps: Abstract each node in the food production process into different states, which may include raw material procurement, production, processing, storage, transportation, etc., to ensure that the definition of the state is comprehensive.

[0100] The states of each node are combined to form different state combinations, and based on the probabilistic relationship between each state combination, the logical relationship between each food risk event is analyzed. This can be achieved through probabilistic dynamic analysis methods such as probabilistic graphical models and Markov models.

[0101] Through probabilistic dynamic analysis, we can identify combinations of states with a higher probability of causing food risks. These combinations may constitute weak links in food safety. Weak links refer to the combinations of states under which food risk events are more likely to occur or spread.

[0102] For each food risk event, determine the causal critical paths that lead to its occurrence. These critical paths are the key factors that lead to specific risk events in the state combination. By analyzing them, the main sources and transmission paths of risks can be found.

[0103] Based on the analysis results, improvement measures are formulated to reduce food risks, including measures to improve production processes, strengthen regulatory measures, and enhance raw material testing to reduce high-probability combinations of states and reduce the risks of weak links.

[0104] S643. Use the joint probability distribution algorithm to predict the probability of occurrence of food risk events, and at the same time traverse the risk node combinations that affect food risk events in the Bayesian model, and calculate the posterior probability of the risk node combinations.

[0105] S644. Confirm the specific risk nodes based on the posterior probability of the risk node combination, determine the food risk path based on the fault node, and conduct risk investigation on the food risk path in combination with reverse reasoning.

[0106] S65. Reversely identify the key sources that trigger food risk events, and formulate targeted preventive measures based on the risk level of food risk events.

[0107] like Figure 2As shown, according to another embodiment of the present invention, a mixing ratio detection and control system for food configuration is also provided, and the mixing ratio detection and control system for food configuration includes an image acquisition module 1, a mixing uniformity evaluation module 2, a sampling analysis module 3, a configuration scheme generation module 4, an optimal configuration scheme selection module 5 and a food risk assessment module 6; The image acquisition module 1 is connected to the mixing uniformity assessment module 2, the mixing uniformity assessment module 2 is connected to the sampling analysis module 3, the sampling analysis module 3 is connected to the configuration scheme generation module 4, the configuration scheme generation module 4 is connected to the optimal configuration scheme selection module 5, and the optimal configuration scheme selection module 5 is connected to the food risk assessment module 6; The image acquisition module 1 is used to mix the food raw materials according to a preset ratio using a mixing device, and to acquire images using a multi-camera module during the mixing process; Mixing uniformity assessment module 2, used to extract raw material mixture features from the image set, and comprehensively assess the mixing uniformity based on the raw material mixture features and raw material physical properties; The sampling and analysis module 3 is used to set a sampling time window to perform sampling and analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted according to the sampling and analysis results; Configuration scheme generating module 4, used for adding new raw materials to the raw material mixture according to the adjusted ratio and performing secondary mixing, and continuously collecting configuration data under different environmental conditions to generate multiple groups of food configuration schemes; The optimal configuration scheme selection module 5 is used to compare and analyze multiple groups of food configuration schemes and extract the optimal food configuration scheme using a multi-objective optimization algorithm; Food risk assessment module 6 is used to build a food risk assessment model, determine whether the optimal food configuration plan has food safety risks, and formulate targeted preventive measures based on food safety risks.

[0108] In summary, with the help of the above-mentioned technical scheme of the present invention, the present invention can monitor the distribution of raw materials in the mixing process in real time by extracting the characteristics of the raw material mixture from the image set, help to timely discover problems that may cause unevenness, and combine the characteristics of the raw material mixture and the physical properties of the raw materials to achieve a multi-dimensional evaluation of the mixing uniformity, which is helpful to judge the mixing quality more comprehensively and accurately, improve the control level of the food configuration process, and at the same time, by effectively evaluating the mixing uniformity, it is helpful to improve the consistency of food products, ensure that the quality of each batch of finished products is relatively stable, reduce dependence on manual judgment, and thus improve the reliability of food configuration; the present invention adds new raw materials to the raw material mixture according to the adjusted ratio and performs secondary mixing, which can flexibly respond to different ratio requirements and improve the flexibility of the production process. At the same time, the configuration data under different environmental conditions is continuously collected, which is helpful to fully understand different The configuration performance under working conditions improves the robustness of the configuration scheme, and comprehensively compares and analyzes multiple groups of food configuration schemes through a multi-objective optimization algorithm to help determine the optimal configuration scheme, improve the efficiency and quality of food configuration, and consider multiple optimization goals at the same time to ensure that the food configuration reaches higher quality standards, thereby helping to improve food production efficiency and food production quality; the present invention can effectively identify and evaluate various risk factors that may exist in food configuration schemes by constructing a food risk assessment model, thereby helping to discover potential food safety problems early and reduce the occurrence of food safety accidents. At the same time, by evaluating the food risks in the optimal food configuration scheme, it can ensure that the selected optimal configuration scheme is also extremely reliable in terms of safety, which can not only improve the overall food quality, but also formulate more effective prevention and control measures based on the risk assessment results, thereby reducing food safety risks.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting and controlling a mixing ratio of food preparation, characterized in that: The mixing ratio detection and control method for food preparation comprises the following steps: S1. Mix the food ingredients according to a preset ratio using a mixing device, and use a multi-camera module to collect images during the mixing process; S2, extracting raw material mixture features from the image set, and comprehensively evaluating the mixing uniformity by combining the raw material mixture features and raw material physical properties; S3. Establish a sampling time window to conduct sampling analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted based on the sampling analysis results; S4, adding new raw materials to the raw material mixture according to the adjusted ratio and performing secondary mixing, while continuously collecting configuration data under different environmental conditions to generate multiple groups of food configuration plans; S5. Compare and analyze multiple groups of food configuration plans, and use a multi-objective optimization algorithm to extract the optimal food configuration plan; S6. Construct a food risk assessment model to determine whether the optimal food configuration plan poses a food safety risk, and formulate targeted preventive measures based on the food safety risk.

2. A method for detecting and controlling a mixing ratio of food preparation according to claim 1, characterized in that: The method of extracting the raw material mixture characteristics from the image set and comprehensively evaluating the mixing uniformity by combining the raw material mixture characteristics and the raw material physical properties comprises the following steps: S21, extracting color distribution and texture features of the raw material mixture from the image set, and judging whether the raw material mixture meets the uniformity requirement based on the color distribution and texture features; S22, if the raw material mixture meets the uniformity requirement, execute step S3, if the raw material mixture does not meet the uniformity requirement, execute step S23; S23, estimating the raw material properties of each component in the raw material mixture by using the prior knowledge base, and performing hierarchical weighting on each raw material property to obtain a level characteristic value of each raw material property; S24, calculating the information entropy of each raw material property by using the entropy weight method, and converting the information entropy into a relative entropy value; S25, combining the level characteristic value and the relative entropy value to obtain the characteristic value of the influence of each raw material property on the raw material mixture, sorting the influencing characteristic values, and taking the raw material properties at the top of the sorting position as the main influencing factors; S26. Formulate an adjustment strategy based on the main influencing factors, and continue to mix the raw material mixture that does not meet the uniformity requirements according to the adjustment strategy.

3. A method for detecting and controlling the mixing ratio of food preparation according to claim 2, characterized in that: The method of extracting the color distribution and texture characteristics of the raw material mixture from the image set and judging whether the raw material mixture meets the uniformity requirement based on the color distribution and texture characteristics comprises the following steps: S211, converting each image in the image set into a YCbCr color space, identifying a raw material mixture region according to an image color rule, and then extracting color distribution characteristics in the raw material mixture region; S212, dividing each image into a plurality of image sub-blocks, and using a local binary pattern algorithm to calculate the binary value of each pixel in each image sub-block, and performing histogram statistics on each image sub-block to obtain a plurality of histograms; S213, forming an image feature vector from a plurality of histograms, and describing the texture feature of the image through the image feature vector; S214, using short-range Fourier transform to calculate the frequency domain phase information of each pixel on the image, and using a local phase quantization algorithm to extract texture features through the frequency domain phase information; S215, combining the texture features and color distribution features extracted by the local binary pattern algorithm and the local phase quantization algorithm respectively to form a comprehensive image feature vector, and inputting the comprehensive image feature vector into a support vector machine for uniformity testing.

4. A method for detecting and controlling the mixing ratio of food preparation according to claim 3, characterized in that: The method of estimating the raw material properties of each component in the raw material mixture by using the prior knowledge base and weighting each raw material property in layers to obtain the level characteristic value of each raw material property includes the following steps: S231, collecting the raw material properties of each food raw material in the raw material mixture, and establishing a priori knowledge base according to the raw material properties, wherein the raw material properties include particle size, shape, density and moisture content of each food raw material; S232. Classify and combine the raw material properties to form a structured raw material property index system; S233, constructing a judgment matrix based on the raw material property index system, wherein the judgment matrix is ​​composed of a combination of quantitative index values ​​of raw material properties and superiority of qualitative indexes; S234, normalizing the judgment matrix to obtain a normalized judgment matrix, and calculating the raw material physical property elements in the normalized judgment matrix; S235, determining the raw material physical property index value based on the raw material physical property elements, and calculating the relative superiority vector in each layer structure using a multi-level optimization theory algorithm; S236. The relative superiority vector calculated in each layer of the structure is used as the input of the next layer of the structure, and step S235 is repeated to obtain the relative superiority vector of the highest layer of the structure, and use it as the level characteristic value of the raw material property.

5. A method for detecting and controlling the mixing ratio of food preparation according to claim 4, characterized in that: The comparative analysis of multiple groups of food configuration schemes and the extraction of the optimal food configuration scheme using a multi-objective optimization algorithm include the following steps: S51, obtaining configuration data of each food configuration scheme, wherein the configuration data includes raw material component ratio, mixing equipment parameters and mixing environment; S52. Determine evaluation indicators for comparative analysis, and use the evaluation indicators to evaluate the performance of each food configuration solution; S53. Use the chaotic sequence generated by chaotic mapping to replace the randomly generated initial population, and screen out the optimal food configuration plan based on population crowding and non-dominated sorting.

6. A method for detecting and controlling the mixing ratio of food preparation according to claim 5, characterized in that: The method of replacing the randomly generated initial population with the chaotic sequence generated by the chaotic mapping and selecting the optimal food configuration scheme according to the population crowding degree and non-dominated sorting includes the following steps: S531, taking the number of food configuration solutions as the target number, and using chaotic mapping to generate a chaotic sequence as the initial population of the dynamic optimization strategy; S532, calculating the fitness value of each food configuration scheme in the initial population, and defining the order value and aggregation distance of each food configuration scheme according to the fitness value and the food dominance relationship; S533, calculating the initial population based on the order value and the focal distance to obtain the parent population, and then performing crossover and mutation operations on the parent population to obtain the offspring population; S534, merging the initial population and the parent population to obtain a new population, and using a dynamic optimization strategy to judge all food configuration schemes in the new population, and retaining the food configuration scheme if it satisfies the dominance relationship of the dynamic optimization strategy; S535. Perform non-dominated sorting on all food configuration plans in the new population in combination with the population crowding degree, and select the optimal configuration plan according to the non-dominated sorting result.

7. A method for detecting and controlling the mixing ratio of food preparation according to claim 6, characterized in that: The construction of the food risk assessment model, determining whether the optimal food configuration plan has food safety risks, and formulating targeted preventive measures based on food safety risks includes the following steps: S61. Carry out food simulation configuration according to the optimal food configuration plan, and build a food risk assessment model based on the food simulation configuration results; S62. Use the food risk assessment model to establish a priority relationship between potential food risk event assessment information and risk level classification thresholds; S63. Determine the risk level of food risk events based on the established priority relationship; S64. Convert the food risk assessment model into a Bayesian model, use the Bayesian model to predict the probability of occurrence of food risk events through forward reasoning, and combine it with reverse reasoning to conduct risk screening; S65. Reversely identify the key sources that trigger food risk events, and formulate targeted preventive measures based on the risk level of food risk events.

8. A method for detecting and controlling the mixing ratio of food preparation according to claim 7, characterized in that: The method of converting the food risk assessment model into a Bayesian model, using the Bayesian model to predict the probability of occurrence of food risk events by forward reasoning, and combining reverse reasoning to conduct risk screening includes the following steps: S641. Convert the food risk assessment model into a Bayesian model, and use the node transfer and conditional probability in the Bayesian model to calculate the probability of all node state combinations; S642. Dynamically analyze the logical relationship of each food risk event, food safety weaknesses and key causal paths based on the probability of all node status combinations; S643. Predict the probability of occurrence of food risk events using a joint probability distribution algorithm, and traverse the risk node combinations that affect food risk events in a Bayesian model, and calculate the posterior probability of the risk node combinations; S644. Confirm the specific risk nodes based on the posterior probability of the risk node combination, determine the food risk path based on the fault node, and conduct risk investigation on the food risk path in combination with reverse reasoning.

9. A method for detecting and controlling the mixing ratio of food preparation according to claim 8, characterized in that: The calculation formula of the raw material physical property elements in the normalized judgment matrix is: ； In the formula, Represents the first i Line j List the physical properties of raw materials; Indicates i Line j Quantitative index values ​​of raw material properties listed; i Indicates the number of rows of the normalized judgment matrix; j Represents the number of columns of the normalized judgment matrix.

10. A mixing ratio detection control system for food configuration, used to implement the mixing ratio detection control method for food configuration as claimed in any one of claims 1 to 9, characterized in that: The mixing ratio detection and control system for food configuration includes an image acquisition module, a mixing uniformity assessment module, a sampling analysis module, a configuration scheme generation module, an optimal configuration scheme selection module and a food risk assessment module; The image acquisition module is connected to the mixing uniformity assessment module, the mixing uniformity assessment module is connected to the sampling analysis module, the sampling analysis module is connected to the configuration scheme generation module, the configuration scheme generation module is connected to the optimal configuration scheme selection module, and the optimal configuration scheme selection module is connected to the food risk assessment module; The image acquisition module is used to mix the food ingredients according to a preset ratio using a mixing device, and to acquire images using a multi-camera module during the mixing process; The mixing uniformity assessment module is used to extract the raw material mixture characteristics from the image set, and comprehensively assess the mixing uniformity in combination with the raw material mixture characteristics and the raw material physical properties; The sampling and analysis module is used to set a sampling time window to perform sampling and analysis on the raw material mixture, and determine whether the preset ratio needs to be adjusted according to the sampling and analysis results; The configuration scheme generating module is used to add new raw materials to the raw material mixture according to the adjusted ratio and perform secondary mixing, and continuously collect configuration data under different environmental conditions to generate multiple groups of food configuration schemes; The optimal configuration scheme selection module is used to compare and analyze multiple groups of food configuration schemes and extract the optimal food configuration scheme using a multi-objective optimization algorithm; The food risk assessment module is used to construct a food risk assessment model, determine whether the optimal food configuration plan has food safety risks, and formulate targeted preventive measures based on food safety risks.

Images