An intelligent control system and method for ultra-high temperature instantaneous steam sterilization equipment

Abstract

This invention belongs to the field of equipment control technology and discloses an intelligent control system and method for an ultra-high temperature instantaneous steam sterilization equipment. The method includes: receiving food raw material data collected by an analysis terminal; analyzing the food raw material data to obtain the food surface area; optimizing the equipment parameters of the sterilization equipment based on the food raw material data and food surface area to obtain the optimal equipment parameters; collecting parameter influence data; calculating the parameter deviation based on the parameter influence data; correcting the optimal equipment parameters based on the parameter deviation; controlling the sterilization equipment to operate according to the corrected optimal equipment parameters; evaluating the sterilization effect of the sterilized food; optimizing the preset feed rate based on the sterilization effect; and sending the optimized feed rate to the analysis terminal. This invention realizes automated control of the sterilization equipment, avoids wasting excessive energy on ineffective sterilization, and improves the sterilization effect and energy utilization efficiency.

Description

Technical Field

[0001] This invention relates to the field of equipment control technology, and more specifically, to an intelligent control system and method for an ultra-high temperature instantaneous steam sterilization device. Background Technology

[0002] Chinese Patent CN212087934U discloses an ultra-high temperature instantaneous food material sterilization machine, comprising: a high-temperature sterilization tank, a heat exchanger, and heating pipes; the high-temperature sterilization tank has a cylindrical structure; the heat exchanger is located on one side of the high-temperature sterilization tank and is connected to the high-temperature sterilization tank via a discharge transition pipe and a feed transition pipe; the heating pipes are located inside the high-temperature sterilization tank and are connected to the high-temperature sterilization tank by welding; the steam input pipe is located on one side of the high-temperature sterilization tank and is connected to the high-temperature sterilization tank by welding; the steam output pipe is located at the top of the high-temperature sterilization tank; this device has the advantages of simple structure, convenient maintenance, fast heating speed, and high working efficiency.

[0003] While the aforementioned technologies can be used in food sterilization processes, they cannot adjust various parameters of the sterilization equipment in real time, such as steam temperature and feed rate. In food production lines, factors such as the characteristics of the raw materials (e.g., surface area, moisture content), environmental conditions (e.g., temperature and humidity), and feed rate can all change. Therefore, the actual requirements for various parameters of the sterilization equipment to achieve the best sterilization effect vary during the sterilization process. If these parameters cannot be adjusted in real time, it is difficult to guarantee the stability of food sterilization quality and will result in a certain degree of energy waste, thereby reducing the energy efficiency of the sterilization equipment.

[0004] In view of this, the present invention proposes an intelligent control system and method for an ultra-high temperature instantaneous steam sterilization device to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and achieve the above objectives, the present invention provides the following technical solution:

[0006] A smart control method for an ultra-high temperature instantaneous steam sterilization device includes:

[0007] Receive and analyze food ingredient data collected by the analysis unit;

[0008] Analyze food ingredient data to obtain the food surface area;

[0009] Based on food raw material data and food surface area, the equipment parameters of the sterilization equipment are optimized to obtain the best equipment parameters;

[0010] The parameters collected affect the data;

[0011] Based on the parameter influence data, calculate the parameter deviation, correct the optimal equipment parameters according to the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters;

[0012] The sterilization effect of the food after sterilization is evaluated, the preset feed rate is optimized based on the sterilization effect, and the optimized feed rate is sent to the analysis terminal.

[0013] Furthermore, the food ingredient data includes microbial contamination levels, moisture content, and food images;

[0014] The method for obtaining the surface area of ​​food includes:

[0015] The food image in the food raw material data is converted to grayscale to obtain the grayscale values ​​of a pixels, where a is the total number of pixels in the food image.

[0016] A preset grayscale threshold is set. The grayscale values ​​of a pixels are compared and analyzed with the grayscale threshold. Pixels with grayscale values ​​less than or equal to the grayscale threshold are marked as food pixels, while pixels with grayscale values ​​greater than the grayscale threshold are not marked.

[0017] Extract the attributes of the food image, obtain the resolution of the food image based on the attributes, and then obtain the area of ​​a pixel based on the resolution of the food image; preset the scaling factor, count the number of food points, and multiply the number of food points, the scaling factor, and the area of ​​a pixel in sequence to obtain the surface area of ​​the food.

[0018] Furthermore, the equipment parameters include steam temperature and feed rate;

[0019] Methods for obtaining optimal device parameters include:

[0020] Initial temperature T max Minimum temperature T min Cooling coefficient δ, maximum number of iterations The system consists of a set of m parameters, where m is an integer greater than 1. Available solutions are randomly generated, and a fitness function is defined. New available solutions χ′ are generated by iteratively perturbing the neighborhood. The fitness difference determines whether to accept a new available solution (better solutions are directly adopted, and worse solutions are accepted based on probability). After one round of iteration, the current temperature is lowered, and the maximum number of iterations is reset. This process is repeated until the temperature drops to the minimum temperature T. min Finally, the set of parameters corresponding to the available solution with the best fitness is output, which serves as the set of parameters corresponding to the best equipment.

[0021] Furthermore, in the method for obtaining optimal equipment parameters, a preset parameter range is defined, which includes a steam temperature range and a feed rate range; a value is randomly selected from the steam temperature range and the feed rate range respectively, and a parameter set is constructed, for a total of m parameter sets, and all m parameter sets are different;

[0022] The fitness difference f″ is calculated by comparing the fitness f′ corresponding to the new available solution χ′ with the fitness f corresponding to the available solution χ.

[0023] Furthermore, in the method for obtaining optimal equipment parameters, the fitness function value is the residual contamination level;

[0024] Residual contamination level refers to the average degree of microbial contamination remaining in a batch of food that has undergone sterilization; methods for obtaining residual contamination level include:

[0025] Obtain the sterilization distance, which is the distance between the inlet and outlet of the sterilization equipment; divide the sterilization distance by the feed rate corresponding to the available scheme χ to obtain the sterilization time; use the analysis data, sterilization time and steam temperature as test data, input the test data into the trained contamination prediction model to predict the corresponding residual contamination level; the analysis data includes microbial contamination level, moisture content and food surface area.

[0026] The training process for the pollution prediction model includes:

[0027] b sets of test data are collected in advance, and a corresponding residual contamination level is set for each set of test data, where b is an integer greater than 1. The test data and the corresponding residual contamination level are converted into a set of feature vectors. Each set of feature vectors is used as the input to the contamination prediction model. The contamination prediction model takes the predicted residual contamination level corresponding to each set of test data as the output and the actual residual contamination level corresponding to each set of test data as the prediction target. The actual residual contamination level is the residual contamination level corresponding to the test data collected in advance. The training objective is to minimize the sum of prediction errors of all test data. The contamination prediction model is trained until the sum of prediction errors converges and then training stops. The contamination prediction model is a deep neural network model.

[0028] Furthermore, the parameter impact data includes environmental impact data and power difference; the environmental impact data includes ambient temperature and ambient humidity.

[0029] The method for obtaining the power difference is as follows: continuously collect historical power differences at c time points, where c is an integer greater than 1; train a difference prediction model based on the c historical power differences; input the c historical power differences into the trained difference prediction model to predict the power difference at the current time point.

[0030] Furthermore, the method for calculating the parameter deviation includes:

[0031] The parameter influence data is input into the trained deviation analysis model to calculate the corresponding parameter deviation. The training process of the deviation analysis model is the same as that of the pollution prediction model, and both are deep neural network models. The parameter deviation is the steam temperature deviation. The parameter deviation is added to the steam temperature in the optimal equipment parameters to obtain the steam temperature correction. The steam temperature in the optimal equipment parameters is then corrected to the steam temperature correction.

[0032] Furthermore, the method for evaluating the sterilization effect of the food after sterilization includes:

[0033] Collect the microbial contamination level of the food after sterilization and mark it as the residual contamination level; divide the residual contamination level by the microbial contamination level to obtain the sterilization effect;

[0034] The method for optimizing the preset feed rate based on the sterilization effect includes:

[0035] The feed amount is the weight of the food products in the same batch that are being sterilized; a preset effect threshold is used to compare the sterilization effect with the effect threshold; if the sterilization effect is less than the effect threshold, an optimization instruction is generated; if the sterilization effect is greater than or equal to the effect threshold, no optimization instruction is generated; a preset ratio coefficient is used; if an optimization instruction is generated, the ratio coefficient is multiplied by the sterilization effect to obtain the optimization ratio, and the preset feed amount is multiplied by the optimization ratio to obtain the optimized feed amount.

[0036] Furthermore, it also includes: collecting the distance to be sterilized and calculating the heating time, and adjusting the feeding speed according to the distance to be sterilized and the heating time;

[0037] The sterilization distance is the distance between the next batch of food to be sterilized and the inlet of the sterilization equipment;

[0038] Methods for calculating heating time include:

[0039] Obtain the steam mass and specific heat capacity; collect the current steam temperature; subtract the current steam temperature from the optimal equipment parameters to obtain the steam temperature to be raised; obtain the actual heating power at the current time point and mark it as the current power; multiply the steam mass, specific heat capacity, and steam temperature to be raised in sequence, and then divide by the current power to obtain the heating time;

[0040] Divide the distance to be sterilized by the heating time to obtain the optimal feeding speed; control the sterilization equipment to operate according to the optimal feeding speed.

[0041] An intelligent control system for an ultra-high temperature instantaneous steam sterilization device, comprising the following intelligent control method for the ultra-high temperature instantaneous steam sterilization device:

[0042] The data receiving module is used to receive food raw material data collected by the analysis terminal;

[0043] The data analysis module is used to analyze food raw material data and obtain the surface area of ​​the food.

[0044] The parameter optimization module is used to optimize the equipment parameters of the sterilization equipment based on food raw material data and food surface area to obtain the best equipment parameters;

[0045] The data acquisition module is used to collect data on the effects of parameters.

[0046] The parameter correction module is used to calculate the parameter deviation based on the parameter influence data, correct the optimal equipment parameters based on the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters.

[0047] The effect evaluation module is used to evaluate the sterilization effect of the food after sterilization, optimize the preset feed amount based on the sterilization effect, and send the optimized feed amount to the analysis terminal.

[0048] The technical effects and advantages of the intelligent control system and method for ultra-high temperature instantaneous steam sterilization equipment of the present invention are as follows:

[0049] 1. By collecting and analyzing food raw material data in real time, the surface area of ​​the food is accurately obtained. Based on the raw material data and surface area, a simulated annealing algorithm is used to dynamically optimize the parameters. At the same time, the system collects data on the influence of parameters and uses deep learning algorithms to correct the parameters in real time, thereby achieving intelligent control of the sterilization equipment parameters throughout the process. In addition, the feed rate is optimized based on the evaluation results of the sterilization effect, further improving the control parameters of the sterilization equipment. It has the advantages of instantaneous response and automatic adaptation to different raw material characteristics and influencing parameters, thereby achieving automated control of the sterilization equipment, avoiding excessive energy waste in ineffective sterilization, thus ensuring the quality of food sterilization, improving sterilization effect and energy utilization efficiency.

[0050] 2. Based on the distance to be sterilized and the time required for heating steam before different batches of food enter the sterilization equipment, the optimal feeding speed is dynamically calculated. This prevents the food from entering the sterilization equipment before the steam reaches the optimal temperature, thus affecting the sterilization effect, while also avoiding the impact of a slow feeding speed on production efficiency. Through optimized control of the feeding speed, precise control of the sterilization temperature and intelligent optimization of the process are achieved, which not only improves the sterilization effect but also effectively saves energy consumption. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the intelligent control system of an ultra-high temperature instantaneous steam sterilization device according to Embodiment 1 of the present invention;

[0052] Figure 2 This is a schematic diagram showing the positions of the analysis end and the sterilization device in Embodiment 1 of the present invention;

[0053] Figure 3 This is a schematic diagram of the intelligent control system of an ultra-high temperature instantaneous steam sterilization device according to Embodiment 2 of the present invention;

[0054] Figure 4 This is a flowchart of an intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to Embodiment 3 of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1

[0057] Please see Figure 1 As shown in this embodiment, the intelligent control system of the ultra-high temperature instantaneous steam sterilization equipment includes a data receiving module, a data analysis module, a parameter optimization module, a data acquisition module, a parameter correction module, and an effect evaluation module; each module is connected via wired and / or wireless means to realize data transmission between modules.

[0058] The data receiving module is used to receive food raw material data collected by the analysis terminal, including food products such as fruits, vegetables, and gastrodia elata.

[0059] The analysis unit includes feeding equipment and analytical equipment; the feeding equipment is used to transport food from the storage silo to the production line, and the analytical equipment is used to analyze the data of the food raw materials; the positional relationship between the analysis unit and the sterilization equipment is as follows: Figure 2 As shown, the sterilization equipment is an ultra-high temperature instantaneous steam sterilization device; the analytical equipment includes a microbial contamination detector, an infrared moisture analyzer, and an image sensor.

[0060] Food ingredient data includes microbial contamination level, moisture content, and food images; microbial contamination level is the average degree of microbial contamination in the same batch of sterilized food; moisture content is the average moisture content of the same batch of sterilized food; microbial contamination level is obtained by a microbial contamination level detector, moisture content is obtained by an infrared moisture meter, and food images are obtained by an image sensor.

[0061] A higher level of microbial contamination indicates more microbial contamination in the same batch of food. Therefore, it is necessary to increase the steam temperature to completely kill the existing microorganisms. At the same time, it is necessary to reduce the feeding speed to extend the sterilization time of the same batch of food and ensure that the same batch of food is fully sterilized at ultra-high temperature. A higher moisture content means more free water molecules inside the same batch of food, which will reduce the heat conduction efficiency. Therefore, it is necessary to increase the steam temperature to ensure that the inside of the same batch of food can reach a sufficient sterilization temperature. At the same time, it is necessary to slow down the feeding speed to extend the sterilization time and ensure that the same batch of food is fully heated at ultra-high temperature.

[0062] It should be noted that, since multiple foods can be sterilized simultaneously during each sterilization process in a food production line to improve sterilization efficiency, the foods sterilized at the same time are considered to be from the same batch. Therefore, the microbial contamination level and moisture content in the food raw material data are averages of multiple foods to accurately reflect the overall condition of the food sterilized in the same batch.

[0063] The data analysis module is used to analyze food raw material data and obtain the surface area of ​​the food.

[0064] Methods for obtaining the surface area of ​​food include:

[0065] The food image in the food raw material data is converted to grayscale to obtain the grayscale values ​​of a pixels, where a is the total number of pixels in the food image.

[0066] A preset grayscale threshold is set. The grayscale values ​​of each of the 'a' pixels are compared and analyzed with the grayscale threshold. Pixels with grayscale values ​​less than or equal to the grayscale threshold are marked as food points, while pixels with grayscale values ​​greater than the grayscale threshold are not marked.

[0067] It should be noted that the grayscale threshold is determined by those skilled in the art during the historical food sterilization process by collecting multiple food images and performing grayscale processing, taking the grayscale values ​​of the corresponding pixels of the food in a food image as a grayscale value set; obtaining the grayscale value with the largest value in each grayscale value set and marking it as the maximum grayscale value; and taking the average of multiple maximum grayscale values ​​as the grayscale threshold.

[0068] It should be understood that because food itself is light in color, usually yellow or grayish-brown, while production line equipment is usually made of metal and is darker in color, such as black or dark gray; after grayscale processing, the darker colors correspond to higher grayscale values, while the lighter colors correspond to lower grayscale values. Therefore, the pixel value of the food in the food image will be lower than the grayscale value of the production line equipment.

[0069] The process involves extracting attributes from food images, determining the resolution of the food image based on these attributes, and then calculating the area of ​​a single pixel based on the resolution. A scaling factor is preset, and the number of food points is counted. The food surface area is obtained by multiplying the number of food points, the scaling factor, and the area of ​​a single pixel sequentially. The scaling factor is determined by a person skilled in the art by measuring the surface area of ​​multiple food images and the actual surface area of ​​the food when determining the grayscale threshold. The food image surface area is the measured surface area in the food image, and the actual surface area is the measured surface area on-site. The actual surface area is divided by the food image surface area to obtain the surface area ratio, and the average of multiple surface area ratios is used as the scaling factor.

[0070] The expression for the surface area of ​​food is: M = TS × TM × λ;

[0071] In the formula, M is the surface area of ​​the food, TS is the number of food dots, TM is the area of ​​a single pixel, and λ is the scaling factor.

[0072] The larger the surface area of ​​the food, the larger the area that steam can contact and transfer heat to. This means that with the same heat input, the food can be heated to the required sterilization temperature more quickly. Therefore, the required steam temperature is lower, while the feeding speed is higher, in order to reduce sterilization time, improve sterilization efficiency, and ensure sterilization quality.

[0073] The parameter optimization module is used to optimize the equipment parameters of the sterilization equipment based on food raw material data and food surface area to obtain the optimal equipment parameters.

[0074] Equipment parameters include steam temperature and feed rate;

[0075] Steam temperature refers to the temperature of the steam used to sterilize food. Steam temperature directly affects the sterilization effect of food. If the steam temperature is too low, the sterilization temperature cannot be reached, thus reducing the sterilization effect. If the temperature is too high, it will cause energy waste, reduce energy utilization efficiency, and may even damage the quality of food.

[0076] The feeding speed is the speed at which food enters the sterilization equipment. The feeding speed determines the residence time of the food in the sterilization equipment, that is, the actual sterilization time. If the feeding speed is too fast, the food cannot fully absorb the required heat, thereby reducing the sterilization effect. If the feeding speed is too slow, it will reduce production efficiency and increase energy consumption, thus reducing energy utilization efficiency.

[0077] Methods for obtaining optimal device parameters include:

[0078] Initial temperature T max Minimum temperature T min Cooling coefficient δ, maximum number of iterations The system consists of a set of m parameters, where m is an integer greater than 1. Available solutions are randomly generated, and a fitness function is defined. New available solutions χ′ are generated by iteratively perturbing the neighborhood. The fitness difference determines whether to accept a new available solution (better solutions are directly adopted, and worse solutions are accepted based on probability). After one round of iteration, the current temperature is lowered, and the maximum number of iterations is reset. This iterative process is repeated until the temperature drops to the minimum temperature T. min Finally, the set of parameters corresponding to the available solution with the best fitness is output as the set of optimal device parameters; the specific steps are as follows:

[0079] Step a: Preset initial temperature T max Minimum temperature T min Cooling coefficient δ, maximum number of iterations And a set of m parameters, and let the current temperature T = T max m is an integer greater than 1;

[0080] Step b: Randomly set an available solution χ, which is the parameter set, and the range of available solutions χ is the set of m parameters;

[0081] Step c: Determine the fitness function;

[0082] Step d: Calculate the fitness f corresponding to the available solution χ; take the available solution χ as the current point, perform random perturbation in the neighborhood of the current point to obtain a new available solution χ′, and calculate the fitness f′ corresponding to the new available solution χ′;

[0083] Step e: Calculate the fitness difference f″; if the fitness difference f″ > 0, then let χ = ​​χ′; if the fitness difference f″ ≤ 0, then calculate the probability p′, and let χ = ​​χ′ based on the probability p′;

[0084] Step f: Repeat steps d through e until the maximum number of iterations is reached. When the loop ends, proceed to step g;

[0085] Step g: Let the current temperature T = T × δ, that is, cool down the current temperature from step a, and assign the cooled value to the current temperature; set the maximum number of iterations. The reduced maximum number of iterations is assigned to the maximum number of iterations; if the reduced maximum number of iterations is not an integer, the reduced maximum number of iterations is rounded up to make it an integer.

[0086] Step h: Repeat steps d to g until the current temperature T < T min When the loop ends, the set of parameters corresponding to the available scheme χ is obtained and used as the set of parameters corresponding to the optimal device.

[0087] It should be noted that the initial temperature T is... max Minimum temperature T min Cooling coefficient δ and maximum number of iterations As preset parameters, these parameters are obtained by those skilled in the art during historical food sterilization processes by acquiring multiple sets of different analytical data, including microbial contamination levels, moisture content, and food surface area. For a batch of food with identical analytical data, multiple sets of different preset parameters are sequentially pre-set, and simulated annealing algorithms are used sequentially to obtain parameter sets. Based on the acquired parameter sets, the corresponding fitness is obtained. The preset parameters corresponding to the parameter set with the highest fitness are used as the preset parameters corresponding to that set of test data. This process is repeated to obtain preset parameters corresponding to multiple sets of different test data. The average value of multiple preset parameters (i.e., the average initial temperature, the average minimum temperature, the average cooling coefficient, and the average maximum number of iterations) is used as the preset initial temperature T in step a. max Minimum temperature T min Cooling coefficient δ and maximum number of iterations

[0088] Step a above includes a preset parameter range, which includes a steam temperature range and a feeding speed range. The parameter range is set by a person skilled in the art through multiple data collections of corresponding steam temperature and feeding speed during historical food sterilization processes. The steam temperature range is set based on the minimum and maximum steam temperature values ​​collected, and the feeding speed range is set based on the minimum and maximum feeding speed values ​​collected. A value is randomly selected from the steam temperature range and the feeding speed range respectively, and a parameter set is constructed. That is, a parameter set includes a steam temperature and a feeding speed. A total of m parameter sets are constructed, and the m parameter sets are all different.

[0089] In step c above, the fitness function is expressed as: f = ww;

[0090] In the formula, f is the fitness and ww is the residual contamination level;

[0091] The residual contamination level is the average degree of contamination of microorganisms remaining after sterilization in the same batch of food.

[0092] Methods for obtaining residual contamination levels include:

[0093] The sterilization distance is obtained, which is the distance between the inlet and outlet of the sterilization equipment. The sterilization distance is obtained by measuring the sterilization equipment by a person skilled in the art. The sterilization distance is divided by the feed rate corresponding to the available scheme χ to obtain the sterilization time. The analysis data, sterilization time and steam temperature are used as test data. The test data are input into the trained pollution prediction model to predict the corresponding residual pollution level.

[0094] The specific training process for the pollution prediction model includes:

[0095] Pre-collect b sets of test data, and set corresponding residual contamination levels for each b set of test data, where b is an integer greater than 1. Convert the test data and corresponding residual contamination levels into a set of corresponding feature vectors. The residual contamination levels corresponding to the test data are collected by those skilled in the art during historical food sterilization processes. Under the analysis data conditions in each set of test data, a sterilization device with steam temperature adjusted to the steam temperature in the test data is used. After the sterilization process of the sterilization time in the test data, a microbial contamination level detector is used to collect the microbial contamination level again, which is then used as the residual contamination level. Set the corresponding residual contamination levels for the b set of test data sequentially.

[0096] Each set of feature vectors is used as input to the pollution prediction model. The pollution prediction model outputs the predicted residual pollution level corresponding to each set of test data and uses the actual residual pollution level corresponding to each set of test data as the prediction target. The actual residual pollution level is the residual pollution level corresponding to the test data collected in advance. The training objective is to minimize the sum of prediction errors of all test data. The prediction error is calculated using the formula η. K =(β) K -ε K ) 2 , where η K Let K be the prediction error, K be the group number of the feature vector corresponding to the test data, and β be the prediction error. K Let ε be the predicted residual contamination level corresponding to the Kth test data group. K Let K be the actual residual pollution level corresponding to the Kth test data; train the pollution prediction model until the sum of prediction errors converges and then stop training.

[0097] It should be noted that deep neural networks can learn complex nonlinear relationships and accurately capture the impact of sterilization time and steam temperature on residual contamination levels under the analyzed data conditions, thereby improving prediction accuracy. Furthermore, deep neural networks can automatically learn effective feature representations from raw data, reducing modeling difficulty. At the same time, deep neural network models have a certain degree of fault tolerance, and can still provide reliable prediction results in the presence of noise or missing data, improving the prediction accuracy of food residual contamination levels in actual production lines. In addition, as more data is collected and fed back, deep neural network models can be continuously retrained and optimized, continuously improving prediction performance.

[0098] In step e above, the fitness difference f″ is expressed as f″ = f′ - f; let χ = ​​χ′, that is, assign the value of the new available solution χ′ to the available solution χ; the probability p′ is expressed as: In the formula, e is the natural constant; let p′ be the probability that χ=χ′, i.e., χ=χ′ is p′; the operation used in step e is roulette wheel selection, which determines the probability of selection based on the fitness of each available solution, thereby retaining excellent available solutions and eliminating poor available solutions; and it does not directly eliminate available solutions with lower fitness, avoiding premature convergence of the algorithm, and can continue to explore new solution spaces, increasing the chance of finding the global optimum.

[0099] It should be understood that the reason for using the simulated annealing algorithm to obtain the optimal equipment parameters is that the simulated annealing algorithm can automatically optimize equipment parameters to obtain the best sterilization effect and improve the automation level of the equipment; and the simulated annealing algorithm can gradually converge to the optimal solution on a certain degree of randomness, with good global search capability and avoid getting trapped in local optima; at the same time, it can flexibly set algorithm parameters such as initial temperature and cooling coefficient to adapt to the characteristics of different sterilization equipment and food raw materials, thereby improving the adaptability and robustness of the simulated annealing algorithm.

[0100] The data acquisition module is used to collect data on the effects of parameters.

[0101] Parameter impact data includes environmental impact data and power difference;

[0102] Environmental impact data includes ambient temperature and ambient humidity; ambient temperature is obtained by a temperature sensor installed on the sterilization equipment; ambient humidity is obtained by a humidity sensor installed on the sterilization equipment; the lower the ambient temperature, the more heat is lost during steam transmission, and therefore the lower the steam temperature, and vice versa; when the ambient humidity is high, steam is more likely to condense during transmission, and steam condensation will cause the steam temperature to drop, therefore the steam temperature is lower.

[0103] The power difference is the difference between the set heating power and the actual heating power of the heating system in the sterilization equipment. The method for obtaining the power difference is as follows: historical power differences at c time points are continuously collected, where c is an integer greater than 1. The historical power differences at c time points are obtained by those skilled in the art by subtracting the corresponding actual heating power collected at the set heating power at c time points during historical food sterilization processes. The time difference between each two time points is a preset time point, which is preset by those skilled in the art according to the actual situation. A difference prediction model is trained based on the c historical power differences. The c historical power differences are input into the trained difference prediction model to predict the power difference at the current time point. The larger the power difference, the more it indicates that the performance of the heating system in the sterilization equipment has decreased, the temperature rise is hindered, and thus the actual steam temperature cannot reach the set steam temperature, resulting in a decrease in steam temperature.

[0104] The training method for the differential prediction model employs a dynamic time series analysis strategy, and the specific steps are as follows:

[0105] Data preprocessing stage:

[0106] A basic training dataset is constructed based on the historical power differences at c consecutive time points, forming an input matrix with time series characteristics.

[0107] Sample generation strategy:

[0108] A sliding window technique (window length W, step size L) is used to dynamically segment the power differences within the basic training dataset, generating training sample groups containing contextual information. Each sample contains the power differences at the first W time points as input features, and the power differences at the Lth step as the prediction target.

[0109] Model architecture design:

[0110] A recurrent neural network (RNN) is used as the core prediction model, leveraging its ability to remember time-series data to capture the dynamic changes in power differences. The input layer dimension matches the window length, and the output layer dimension corresponds to the single-step prediction target.

[0111] Training optimization process:

[0112] With prediction accuracy as the optimization objective, the mean absolute percentage error (MAPE) is used as the model performance evaluation metric. Network parameters are optimized using the backpropagation algorithm, and training is terminated when the measured MAPE value falls below a preset threshold.

[0113] Model validation mechanism:

[0114] After training, a differential prediction model with time series prediction capabilities is generated, which can output the power differential prediction results for the next L steps based on the input historical differential sequence.

[0115] This method enhances data utilization through a dynamic window mechanism and combines the time-series modeling advantages of RNNs to achieve accurate prediction of power differences.

[0116] It should be noted that as the sterilization equipment is used for longer periods, the internal components will gradually age and wear down, and the internal control system will also gradually drift, resulting in an increase in power difference as the sterilization equipment is used for longer periods. RNN models can effectively capture dynamic patterns in time series data and are more suitable for handling issues such as equipment aging that change over time. Therefore, it is necessary to use an RNN neural network model to accurately predict the power difference at future time points, which will facilitate subsequent calculation of parameter deviations.

[0117] The parameter correction module is used to calculate the parameter deviation based on the parameter influence data, correct the optimal equipment parameters based on the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters.

[0118] Methods for calculating parameter deviations include:

[0119] The parameter influence data is input into the trained deviation analysis model to calculate the corresponding parameter deviation. The training process of the deviation analysis model is the same as that of the pollution prediction model, and both are deep neural network models. The parameter deviation is the steam temperature deviation. The parameter deviation is added to the steam temperature in the optimal equipment parameters to obtain the steam temperature correction. The steam temperature in the optimal equipment parameters is then corrected to the steam temperature correction.

[0120] The effect evaluation module is used to evaluate the sterilization effect of the food after sterilization, optimize the preset feed amount based on the sterilization effect, and send the optimized feed amount to the analysis terminal.

[0121] Methods for evaluating the sterilization effect of food after sterilization include:

[0122] The microbial contamination level of the food after sterilization is collected and marked as the residual contamination level. The residual contamination level is obtained by a microbial contamination level detector installed behind the outlet of the sterilization equipment. The sterilization effect is obtained by dividing the residual contamination level by the microbial contamination level.

[0123] Methods for optimizing the preset feed rate based on sterilization effect include:

[0124] The feed amount is the weight of the food to be sterilized in the same batch; the feed amount is preset by those skilled in the art based on the actual situation of the sterilization equipment; a preset effect threshold is used to compare the sterilization effect with the effect threshold; if the sterilization effect is less than the effect threshold, an optimization instruction is generated; if the sterilization effect is greater than or equal to the effect threshold, no optimization instruction is generated; the effect threshold is preset by those skilled in the art based on the required sterilization precision requirements; a preset proportional coefficient is used; if an optimization instruction is generated, the proportional coefficient is multiplied by the sterilization effect to obtain the optimization ratio, and the preset feed amount is multiplied by the optimization ratio to obtain the optimized feed amount; the proportional coefficient is obtained by those skilled in the art through multiple adjustments and sterilization processes during historical food sterilization when an optimization instruction is generated, analyzing the corresponding sterilization effect, obtaining the adjusted feed amount corresponding to the sterilization effect with the largest value, and marking it as the adjusted feed amount, dividing the adjusted feed amount by the feed amount to obtain the adjustment coefficient; and so on to obtain the adjustment coefficients corresponding to multiple optimization instruction generation, and using the average of multiple adjustment coefficients as the proportional coefficient.

[0125] The amount of feed affects the sterilization effect of food. The larger the feed, the thicker the food will accumulate in the sterilization equipment. Excessively thick food will hinder the penetration of ultra-high temperature steam and its full contact with the food, resulting in poor local sterilization effect. It will also affect the heat transfer and temperature uniformity, causing some food to be too cold and unable to reach the required sterilization temperature, thus affecting the overall sterilization effect. Therefore, dynamically optimizing the feed amount based on the sterilization effect feedback after each sterilization can control the feed amount to an optimal value, thereby improving the overall sterilization effect of food.

[0126] This embodiment accurately obtains the surface area of ​​food by collecting and analyzing food raw material data in real time. Based on the raw material data and surface area, a simulated annealing algorithm is used to dynamically optimize the parameters. Simultaneously, parameter influence data is collected and parameters are corrected in real time using a deep learning algorithm, thereby achieving intelligent control of the sterilization equipment parameters throughout the process. In addition, the feed rate is optimized based on the sterilization effect evaluation results, further improving the control parameters of the sterilization equipment. It has the advantages of instantaneous response and automatic adaptation to different raw material characteristics and influencing parameters, thereby realizing automated control of the sterilization equipment, avoiding excessive energy waste for ineffective sterilization, thus ensuring the quality of food sterilization, improving sterilization effect and energy utilization efficiency.

[0127] Example 2

[0128] like Figure 3 As shown, after a batch of food is sterilized, the steam inside the sterilization equipment needs to be reheated to ensure that the steam temperature reaches the sterilization temperature required for the next batch of food. That is, the steam inside the sterilization equipment needs to be heated to the steam temperature in the optimal equipment parameters before the next batch of food enters the sterilization equipment. Therefore, this embodiment provides an intelligent control system for an ultra-high temperature instantaneous steam sterilization equipment, which also includes a speed regulation module.

[0129] The speed adjustment module is used to collect the distance to be sterilized and calculate the heating time, and adjust the feeding speed according to the distance to be sterilized and the heating time.

[0130] The sterilization distance is the distance between the next batch of food to be sterilized and the inlet of the sterilization equipment; the sterilization distance is obtained by a laser rangefinder installed on the sterilization equipment, and the laser rangefinder is parallel to the conveyor belt on the food production line.

[0131] Methods for calculating heating time include:

[0132] The process involves obtaining the steam mass and specific heat capacity. The steam mass is measured by a person skilled in the art, and the specific heat capacity is obtained from a steam property handbook. The current steam temperature is acquired by a temperature sensor such as a thermocouple or RTD (thermal resistance temperature detector) installed inside the sterilization equipment. The current steam temperature is subtracted from the optimal equipment parameters to obtain the desired steam temperature. The actual heating power at the current time point is obtained and marked as the current power. The current power is obtained by the control system inside the sterilization equipment. The heating time is obtained by multiplying the steam mass, specific heat capacity, and desired steam temperature in sequence and then dividing by the current power.

[0133] Divide the distance to be sterilized by the heating time to obtain the optimal feeding speed; control the sterilization equipment to operate according to the optimal feeding speed.

[0134] This embodiment dynamically calculates the optimal feeding speed based on the distance to be sterilized and the time required for heating steam before different batches of food enter the sterilization equipment. This prevents the food from entering the sterilization equipment before the steam reaches the optimal temperature, thus affecting the sterilization effect, while also avoiding the impact of a slow feeding speed on production efficiency. Through optimized control of the feeding speed, precise control of the sterilization temperature and intelligent optimization of the process are achieved, which not only improves the sterilization effect but also effectively saves energy consumption.

[0135] Example 3

[0136] Please see Figure 4 As shown, the parts not described in detail in this embodiment are described in Embodiments 1 and 2. An intelligent control method for an ultra-high temperature instantaneous steam sterilization device is provided, the method comprising:

[0137] Receive and analyze food ingredient data collected by the analysis unit;

[0138] Analyze food ingredient data to obtain the food surface area;

[0139] Based on food raw material data and food surface area, the equipment parameters of the sterilization equipment are optimized to obtain the best equipment parameters;

[0140] The parameters collected affect the data;

[0141] Based on the parameter influence data, calculate the parameter deviation, correct the optimal equipment parameters according to the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters;

[0142] The sterilization effect of the food after sterilization is evaluated, the preset feed rate is optimized based on the sterilization effect, and the optimized feed rate is sent to the analysis terminal.

[0143] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0144] In conclusion, 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 principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent control method for an ultra-high temperature instantaneous steam sterilization device, characterized in that, include: The system receives and analyzes food ingredient data, which includes microbial contamination levels, moisture content, and food images. The surface area of ​​food is obtained by analyzing food ingredient data; the method for obtaining the surface area of ​​food includes: The food image in the food raw material data is converted to grayscale to obtain the grayscale values ​​of a pixels, where a is the total number of pixels in the food image. A preset grayscale threshold is set. The grayscale values ​​of a pixels are compared and analyzed with the grayscale threshold. Pixels with grayscale values ​​less than or equal to the grayscale threshold are marked as food pixels, while pixels with grayscale values ​​greater than the grayscale threshold are not marked. Extract the attributes of the food image, obtain the resolution of the food image based on the attributes, and then obtain the area of ​​a pixel based on the resolution of the food image; preset the scaling factor, count the number of food points, and multiply the number of food points, the scaling factor, and the area of ​​a pixel in sequence to obtain the surface area of ​​the food. Based on food raw material data and food surface area, the equipment parameters of the sterilization equipment are optimized to obtain the best equipment parameters; Collect parameter impact data; the parameter impact data includes environmental impact data and power difference; the environmental impact data includes ambient temperature and ambient humidity; The method for obtaining the power difference is as follows: continuously collect historical power differences at c time points, where c is an integer greater than 1; train a difference prediction model based on the c historical power differences; input the c historical power differences into the trained difference prediction model to predict the power difference at the current time point. Based on the parameter influence data, calculate the parameter deviation, correct the optimal equipment parameters according to the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters; The sterilization effect of the food after sterilization is evaluated, the preset feed rate is optimized based on the sterilization effect, and the optimized feed rate is sent to the analysis terminal.

2. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 1, characterized in that, The equipment parameters include steam temperature and feed rate; methods for obtaining optimal equipment parameters include: Initial temperature Minimum temperature Cooling coefficient Maximum number of iterations And a set of m parameters, where m is an integer greater than 1, randomly generate available solutions and define a fitness function; generate new available solutions by iteratively perturbing the neighborhood. Based on the fitness difference, a decision is made on whether to accept a new available solution. After completing one round of iteration, the current temperature is lowered and the maximum number of iterations is reset. The iteration process is repeated until the temperature drops to the minimum. Finally, the set of parameters corresponding to the available solution with the best fitness is output, which serves as the set of parameters corresponding to the best equipment.

3. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 2, characterized in that, In the method for obtaining optimal equipment parameters, a preset parameter range is defined, which includes a steam temperature range and a feed rate range. A value is randomly selected from the steam temperature range and the feed rate range respectively, and a parameter set is constructed. A total of m parameter sets are constructed, and each of the m parameter sets is different. The fitness difference By calculating new available solutions Corresponding fitness With available solutions Corresponding fitness The difference is obtained by calculation.

4. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 3, characterized in that, In the method for obtaining optimal equipment parameters, the fitness function value is the residual contamination level. Residual contamination level refers to the average degree of microbial contamination remaining in a batch of food that has undergone sterilization; methods for obtaining residual contamination level include: Obtain the sterilization distance, which is the distance between the inlet and outlet of the sterilization equipment; divide the sterilization distance by the available solutions. The corresponding feeding speed is used to obtain the sterilization time; the analysis data, sterilization time and steam temperature are used as test data, and the test data are input into the trained contamination prediction model to predict the corresponding residual contamination level; the analysis data includes microbial contamination level, moisture content and food surface area.

5. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 1, characterized in that, The method for calculating the parameter deviation includes: Input the parameter influence data into the trained deviation analysis model to calculate the corresponding parameter deviation; the deviation analysis model is a deep neural network model; the parameter deviation is the steam temperature deviation; add the parameter deviation to the steam temperature in the optimal equipment parameters to obtain the steam temperature correction; and correct the steam temperature in the optimal equipment parameters to obtain the steam temperature correction.

6. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 5, characterized in that, The method for evaluating the sterilization effect of food after sterilization includes: Collect the microbial contamination level of the food after sterilization and mark it as the residual contamination level; divide the residual contamination level by the microbial contamination level to obtain the sterilization effect; The method for optimizing the preset feed rate based on the sterilization effect includes: The feed amount is the weight of the food products in the same batch that are being sterilized; a preset effect threshold is used to compare the sterilization effect with the effect threshold; if the sterilization effect is less than the effect threshold, an optimization instruction is generated; if the sterilization effect is greater than or equal to the effect threshold, no optimization instruction is generated; a preset ratio coefficient is used; if an optimization instruction is generated, the ratio coefficient is multiplied by the sterilization effect to obtain the optimization ratio, and the preset feed amount is multiplied by the optimization ratio to obtain the optimized feed amount.

7. The intelligent control method for an ultra-high temperature instantaneous steam sterilization device according to claim 6, characterized in that, Also includes: Collect the distance to be sterilized and calculate the heating time, then adjust the feeding speed according to the distance to be sterilized and the heating time; The sterilization distance is the distance between the next batch of food to be sterilized and the inlet of the sterilization equipment; Methods for calculating heating time include: Obtain the steam mass and specific heat capacity; collect the current steam temperature; subtract the current steam temperature from the optimal equipment parameters to obtain the steam temperature to be raised; obtain the actual heating power at the current time point and mark it as the current power; multiply the steam mass, specific heat capacity, and steam temperature to be raised in sequence, and then divide by the current power to obtain the heating time; Divide the distance to be sterilized by the heating time to obtain the optimal feeding speed; control the sterilization equipment to operate according to the optimal feeding speed.

8. An intelligent control system for an ultra-high temperature instantaneous steam sterilization device, implementing the intelligent control method for an ultra-high temperature instantaneous steam sterilization device as described in any one of claims 1-7, characterized in that, include: The data receiving module is used to receive food raw material data collected by the analysis terminal; the food raw material data includes microbial contamination level, moisture content, and food images; The data analysis module is used to analyze food raw material data and obtain the food surface area; the method for obtaining the food surface area includes: The food image in the food raw material data is converted to grayscale to obtain the grayscale values ​​of a pixels, where a is the total number of pixels in the food image. A preset grayscale threshold is set. The grayscale values ​​of a pixels are compared and analyzed with the grayscale threshold. Pixels with grayscale values ​​less than or equal to the grayscale threshold are marked as food pixels, while pixels with grayscale values ​​greater than the grayscale threshold are not marked. Extract the attributes of the food image, obtain the resolution of the food image based on the attributes, and then obtain the area of ​​a pixel based on the resolution of the food image; preset the scaling factor, count the number of food points, and multiply the number of food points, the scaling factor, and the area of ​​a pixel in sequence to obtain the surface area of ​​the food. The parameter optimization module is used to optimize the equipment parameters of the sterilization equipment based on food raw material data and food surface area to obtain the best equipment parameters; The data acquisition module is used to collect parameter impact data; the parameter impact data includes environmental impact data and power difference; the environmental impact data includes ambient temperature and ambient humidity. The method for obtaining the power difference is as follows: continuously collect historical power differences at c time points, where c is an integer greater than 1; train a difference prediction model based on the c historical power differences; input the c historical power differences into the trained difference prediction model to predict the power difference at the current time point. The parameter correction module is used to calculate the parameter deviation based on the parameter influence data, correct the optimal equipment parameters based on the parameter deviation, and control the sterilization equipment to operate according to the corrected optimal equipment parameters. The effect evaluation module is used to evaluate the sterilization effect of the food after sterilization, optimize the preset feed amount based on the sterilization effect, and send the optimized feed amount to the analysis terminal.

Images