Diatomite filter intelligent control method and device based on AI technology, and medium

Abstract

The invention discloses an AI technology-based intelligent control method and device for a diatomite filter, and a medium. The method comprises the following steps: training an AI control model based on historical production data, and generating a curve group of a diatomite addition amount according to the AI control model; obtaining current production data, wherein the AI control model matches a curve of the diatomite addition amount corresponding to the current production batch according to a matching rule and the current production data; the AI control model predicts a comprehensive evaluation result of the current production batch and controls the opening degree of an adding pump of the diatomite filter according to the comprehensive evaluation result and the curve of the diatomite adding amount corresponding to the current production batch, and the opening degree of the adding pump is used for controlling the diatomite adding amount. The diatomite coating amount can be optimized, the single-batch filtering amount is increased, the diatomite dismantling frequency is reduced, and the purpose of saving diatomite is achieved.

Description

Technical Field

[0001] The present invention relates to the field of winemaking equipment control, and in particular to an AI-based intelligent control method, device and medium for a diatomaceous earth filter. Background Art

[0002] Diatomaceous earth filter aid has good microporous structure, adsorption performance and compression resistance. It can not only make the filtered liquid obtain a better flow rate ratio, but also filter out fine suspended matter, ensuring the clarity of the filtered material. It is used in production processes such as beer, white wine, rice wine, and grape wine.

[0003] The diatomaceous earth addition strategy in the prior art only relies on the set values ​​of the filtered wine flow rate and the diatomaceous earth addition flow rate, and is therefore limited to preset rules and experience, lacking accuracy and completeness; it is unable to provide the optimal strategy for complex problems such as multivariable nonlinearity, or unknown new operating conditions. Summary of the invention

[0004] In view of this, the present invention provides an intelligent control method, device, and medium for a diatomaceous earth filter based on AI technology, which can optimize the diatomaceous earth coating amount and better provide a control strategy for the diatomaceous earth filter.

[0005] The first aspect of the present invention provides an intelligent control method for a diatomaceous earth filter based on AI technology, comprising:

[0006] An AI control model is trained based on historical production data, and a curve group of diatomaceous earth addition amount is generated according to the AI ​​control model, wherein the curve group of diatomaceous earth addition amount includes multiple curves of diatomaceous earth addition amount; current production data is obtained, and the AI ​​control model matches the curve of diatomaceous earth addition amount corresponding to the current production batch according to matching rules and current production data; the AI ​​control model predicts a comprehensive evaluation result of the current production batch, and controls the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of diatomaceous earth addition amount corresponding to the current production batch, and the opening of the addition pump is used to control the diatomaceous earth addition amount.

[0007] In some embodiments, the training of the AI ​​control model based on historical production data and the generation of a curve group of diatomaceous earth addition amount according to the AI ​​control model include:

[0008] Training an AI control model based on historical production data, generating production curve groups corresponding to different production batches according to the AI ​​control model, wherein the historically generated data includes turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data;

[0009] The production curve group is evaluated comprehensively according to historical production data so as to optimize the score of the production curve of each batch in the production curve group.

[0010] In some embodiments, the comprehensive evaluation of the production curve group based on historical production data to optimize the score of the production curve of each batch in the production curve group includes:

[0011] The production curve of each batch is determined according to the comprehensive evaluation score of turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data; and the production curve is evaluated according to the following formula:

[0012]

[0013] Among them, Score batch Indicates the batch comprehensive evaluation score of the brand batch or filtered batch; S i represents the evaluation score of historical production data; w i Represents the weight of historical production data, and the sum of the weights is 1.

[0014] In some embodiments,

[0015] The production curve of each batch is determined according to the comprehensive evaluation score of turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data, including: calculating the data score by the following formula:

[0016] Turbidity score deviation = (1-|turbidity target value – turbidity actual value / turbidity target value|) × 100;

[0017] Pressure difference stability score = (1-flow standard deviation / average flow) × 100;

[0018] Microbial reduction rate score before and after filtration = (count before filtration – count after filtration) / count before filtration × 100;

[0019] Soil efficiency score = (1-|target soil usage value – actual soil usage value| / target soil usage value) × 100;

[0020] Wine filtration efficiency score = (actual single batch filtration time / planned single batch filtration time) × (actual filtration volume / planned filtration volume) × 100;

[0021] The physical and chemical data include dissolved oxygen content, pH stability, chromaticity change, and bitterness retention rate;

[0022] Dissolved oxygen content deviation score = (1-|oxygen content target value – oxygen content actual value| / oxygen content target value) × 100;

[0023] pH stability score = (1-pH standard deviation / average pH) × 100;

[0024] Calculation formula: Chroma change rate score = (chroma before filtering – chroma after filtering) / chroma before filtering × 100;

[0025] Bitterness value retention score=[1−(bitterness value before filtration−bitterness value after filtration) / bitterness value before filtration]×100.

[0026] In some embodiments, the obtaining of current production data, and the matching of the AI ​​control model with a curve of the amount of diatomaceous earth added corresponding to the current production batch according to the current production data, include:

[0027] Obtain the indicator characteristics and state characteristics in the current production data, match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the state characteristics, and use the indicator characteristics for the AI ​​control model to adjust the curve of the diatomaceous earth addition amount corresponding to the current production batch to control the opening of the addition pump of the diatomaceous earth filter; wherein the state characteristics include brand, step number, and planned wine filtration volume.

[0028] In some embodiments, the indicator characteristics include turbidity and pressure difference.

[0029] In some embodiments, controlling the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch includes:

[0030] The indicator characteristics are monitored in real time, and the AI ​​control model predicts the target indicator characteristics of a preset time period in the future. When the indicator characteristics monitored in real time deviate from the target indicator characteristics, the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the opening of the addition pump of the diatomaceous earth filter.

[0031] In some embodiments, the training of the AI ​​control model based on historical production data includes:

[0032] The AI ​​control model is trained using XGBoost algorithm, LightGBM algorithm, or NeuralProphet algorithm and historical production data.

[0033] The second aspect of the present invention provides an intelligent control device for a diatomaceous earth filter based on AI technology, comprising:

[0034] A first control module is used to train an AI control model based on historical production data, and generate a curve group of diatomite addition amount according to the AI ​​control model, wherein the curve group of diatomite addition amount includes a plurality of curves of diatomite addition amount;

[0035] The second control module is used to obtain the current production data and match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the matching rules and the current production data;

[0036] The third control module is used to predict the comprehensive evaluation result of the current production batch based on the AI ​​control model, and control the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch, and the opening of the addition pump is used to control the diatomaceous earth addition amount.

[0037] According to a third aspect of the present invention, there is provided an electronic device, comprising:

[0038] At least one processor; and at least one memory in communication with the processor, wherein: the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the diatomaceous earth filter intelligent control method based on AI technology as described in the first aspect of the embodiment of the present invention.

[0039] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a computer, the intelligent control method for a diatomaceous earth filter based on AI technology as described in the first aspect of the embodiment of the present invention is executed.

[0040] The present invention obtains production curves of different batches of diatomite addition amounts based on an AI control model, and controls the opening of a diatomite filter addition pump according to the production curve of the diatomite addition amount, thereby realizing control of the diatomite addition amount, being able to optimize the diatomite coating amount, increase the single batch filtration amount, and reduce the number of diatomite dismantling times, thereby achieving the purpose of diatomite saving; at the same time, being able to improve filtration efficiency and save water and electricity consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in one or more embodiments of the present application or related technologies, the drawings required for use in the embodiments or related technology descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in one or more embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0042] Figure 1 The figure is a flow chart of an intelligent control method of a diatomaceous earth filter based on AI technology in an embodiment of the present invention.

[0043] Figure 2 The schematic diagram is a schematic diagram of an intelligent control method for a diatomaceous earth filter based on AI technology in an embodiment of the present invention.

[0044] Figure 3 Schematic diagram of another intelligent control method for a diatomaceous earth filter based on AI technology in an embodiment of the present invention.

[0045] Figure 4 Schematic diagram of an intelligent control method for a diatomaceous earth filter based on pressure difference and turbidity characteristic prediction in an embodiment of the present invention.

[0046] Figure 5 Schematic diagram of an intelligent control device for a diatomaceous earth filter based on AI technology in an embodiment of the present invention.

[0047] Figure 6 Schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0049] It should be understood that although the terms first, second, third, etc. may be used to describe various information in the disclosed embodiments, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the disclosed embodiments, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0050] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the claims, the singular forms of "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in the specification of the present invention and the claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0051] See also Figure 1 , Figure 1 FIG. 1 is a flow chart showing an intelligent control method of a diatomaceous earth filter based on AI technology according to an embodiment of the present disclosure. Figure 1 As shown, the AI-based diatomaceous earth filter intelligent control method provided by the present invention includes the following steps.

[0052] S101: Training an AI control model based on historical production data, and generating a curve group of diatomaceous earth addition amount according to the AI ​​control model, wherein the curve group of diatomaceous earth addition amount includes a plurality of curves of diatomaceous earth addition amount.

[0053] In some embodiments, the AI ​​control model is trained using XGBoost algorithm, LightGBM algorithm, or NeuralProphet algorithm and historical production data.

[0054] Among them, XGBoost (eXtreme Gradient Boosting) is an efficient machine learning algorithm, which is an improved version of the gradient boosting framework; LightGBM is an efficient gradient boosting framework that uses a tree-based algorithm for learning, and is particularly suitable for machine learning tasks such as classification and regression of large-scale data sets; NeuralProphet is a time series prediction tool based on a neural network that can be used to predict production indicators in the future.

[0055] The historical production data can be obtained by collecting various process parameters and diatomaceous earth addition data generated during the production of each brand of wine. It should be understood that the data on various process parameters and diatomaceous earth addition should be stored in chronological order to facilitate the use of AI control models to train key indicators and dimensional features related to diatomaceous earth addition.

[0056] In other embodiments, the historical production data can also be obtained by collecting process parameters (such as turbidity, pressure difference, oxygen content, temperature, etc.) and diatomaceous earth addition amount data generated during the production process, such as the step number or planned wine filtration volume in the production process.

[0057] That is, the amount of diatomaceous earth added to the production batches of different types of beer is different, and the control strategy of the amount of diatomaceous earth added to the production batches of different types of beer can be obtained according to the model training. For example, when producing a fixed brand of beer, the amount of diatomaceous earth added is adaptively added or reduced according to the different process parameters; for another example, when producing according to a certain production step number, the amount of diatomaceous earth added is adaptively added or reduced according to the different process parameters.

[0058] Based on the above, after the AI ​​control model is obtained through training, a curve group of diatomaceous earth addition amount can be generated according to the AI ​​control model, which respectively corresponds to the curves of diatomaceous earth addition amount of different brands, or different step numbers, or different planned wine filtering amounts. Multiple curves of diatomaceous earth addition amount constitute a curve group of diatomaceous earth addition amount.

[0059] In some embodiments, corresponding AI control models can be trained for production requirements of different brands, different step numbers, or different planned wine filtration volumes. Of course, an AI control model can also be trained to be generalized and applied to wine production.

[0060] The AI ​​control model continuously learns from data and optimizes its own control strategy through machine learning and deep learning algorithms; it can autonomously adapt to different working conditions and environmental changes, and can improve the efficiency and quality of the production process. The AI ​​control model will analyze and understand multivariate, nonlinear relationships and unpredictable factors from historical batches, and can respond to complex filtering production processes and process control beyond manual experience.

[0061] S102: Acquire current production data, and the AI ​​control model matches the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the matching rules and the current production data.

[0062] Taking beer production as an example, the production indicators of beer production, the parameters set by the production equipment, the operating parameters of the production equipment, etc. can be obtained through various sensors or production equipment.

[0063] In some embodiments, the production data includes: turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth usage data, wine filtration efficiency data, etc. The exemplary production process involves turbidity, oxygen content, pressure difference, temperature, wine flow rate, total operation time, duration of each step, flow addition amount, wine filtration amount, soil ratio, inlet temperature, initial pressure difference, etc., which can be used as production data.

[0064] The AI ​​control model uses the matching rule to adaptively recommend the curve of the diatomite addition amount that matches the current production batch as the basic data of the control strategy. The diatomite addition amount is the diatomite addition flow rate. Since the diatomite addition flow rate is linearly related to the opening of the diatomite addition pump, the diatomite addition flow rate can be controlled by providing a diatomite addition pump opening strategy for the diatomite filter.

[0065] In some embodiments, the AI ​​control model can identify the production scenarios such as which brand of production or which process step of production the current production process belongs to by identifying the initial process parameters in the production data, thereby matching the curve of the diatomaceous earth addition amount corresponding to the current production batch.

[0066] The AI ​​control model monitors production data in real time. When production data deviates, the addition pump opening strategy can be fine-tuned according to the diatomaceous earth addition amount curve to achieve optimal control of the entire production process.

[0067] S103: The AI ​​control model predicts the comprehensive evaluation result of the current production batch, and controls the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch, and the opening of the addition pump is used to control the diatomaceous earth addition amount.

[0068] The AI ​​control model can control the opening of the addition pump of the diatomaceous earth filter in real time according to the curve of the diatomaceous earth addition amount. The AI ​​control model makes decisions and adjustments quickly through real-time monitoring and analysis of data to ensure the stability and efficiency of the production process.

[0069] In some embodiments, when the production data deviates, the AI ​​control model can timely adjust the amount of diatomaceous earth added to avoid failures or abnormal operating conditions.

[0070] In some embodiments, the AI ​​control model can monitor key indicators in production parameters in real time, and adjust the amount of diatomaceous earth added in a timely manner when the key indicators change.

[0071] In some embodiments, the AI ​​control model can monitor key indicators in production parameters in real time and predict key indicators in future time periods. When the key indicators in production parameters deviate from the predicted key indicators, the amount of diatomaceous earth added is adjusted in a timely manner.

[0072] See also Figure 2 , Figure 2 This is the schematic diagram of the intelligent control method provided by the present invention. In the diatomite filtration process production, the construction of the AI ​​control model is based on the process program and the automation control logic. Its system interaction mode combines the business process and automation control theory of the existing production process. The AI ​​control system uses the database PGDB and basic data collection of the IOT base as the interaction middleware with the OT side, including a two-way link for data input and strategy output.

[0073] The base IOT collects data from the local system or IOT through the OPC-UA protocol, through the gateway, and transfers it to PGDB. During the data flow, the data is converted and calculated based on the agreed structure and format. Then, it is passed through the AI ​​control model ( Figure 2 In steps 1 to 8), the final control strategy instruction is sent to the PLC of the diatomaceous earth filter through the sending interface of the IOT base.

[0074] Figure 2In the process, all historical batch data are obtained from the PGDB of IOT. After data alignment, the basic information of the brand / filter batch, including batch number, brand, start and end time, filter time, filter volume, flow addition, etc., is obtained. The historical original point indicators are scored in percentage, including turbidity, pressure difference, oxygen content, temperature, etc. The evaluation results of the above original point indicators are weighted to form a comprehensive evaluation of the batch. According to the characteristics of brand, step, filter volume and initial state, multiple historical batch clusters are generated. Based on the above batch clusters, the corresponding optimal strategy curve group is generated, including the diatomaceous earth addition pump opening strategy curve for each process / step. For the actual data and information of the online production batch, rule matching is performed to match the model of the same cluster (same brand, same step and same planned output), and the opening production curve of the optimal batch matching the current production batch is adaptively recommended as the control benchmark strategy. During the production process, important production indicators are monitored in real time. When deviations occur, the control strategy is fine-tuned to achieve optimal control of the entire production process. For example, if the turbidity is predicted to be poor in the next 10 minutes, the flow pump opening will be increased and the strategy will be fine-tuned. The final AI control strategy will be stored in PGDB, and the command will be sent to the PLC, and also displayed on the UI interface.

[0075] First, historical data is obtained and preprocessed to form prior knowledge, and the best optimization strategy is constructed to control the opening of the addition pump of the diatomaceous earth filter; then the data is used to train the model to obtain the AI ​​control model. The AI ​​control model can obtain the curve of the diatomaceous earth addition amount through training, which is used as the main line strategy to control the opening of the addition pump.

[0076] During production, the AI ​​control model obtains online production data and performs model inference based on the production data to obtain key indicators for beer production in future time periods; it monitors key indicators in production parameters in real time and determines whether there are any abnormalities. If there are any abnormalities, the main line strategy is used to correct the opening of the addition pump.

[0077] The present invention obtains the production curves of different batches of diatomaceous earth addition based on the AI ​​control model, and controls the opening of the addition pump of the diatomaceous earth filter according to the production curve of the diatomaceous earth addition, thereby realizing the control of the diatomaceous earth addition, and can optimize the diatomaceous earth coating amount, increase the single batch filtration amount, and reduce the number of diatomaceous earth removal times, thereby achieving the purpose of diatomaceous earth saving. Since the diatomaceous earth addition amount can be controlled in real time, the filtration efficiency can be improved, and the water and electricity consumption can also be saved. In addition, the refined control of the diatomaceous earth addition amount can achieve the effect of reducing the turbidity of sake, thereby improving the freshness of the wine and the physical, chemical and flavor evaluation.

[0078] Furthermore, when training an AI control model based on historical production data, the historical production data can be characterized to meet the model training requirements.

[0079] As mentioned above, production data include: turbidity, oxygen content, pressure difference, temperature, wine flow rate, total operation time, duration of each step, flow addition amount, filtered wine amount, soil ratio, inlet temperature, initial pressure difference, microorganisms, wine filtration efficiency, dissolved oxygen content, pH stability, color change, bitterness retention rate, etc.

[0080] In one embodiment, an AI control model is trained based on historical production data, and a production curve group corresponding to different production batches is generated according to the AI ​​control model. At least turbidity data, pressure difference data, microbial data, physical and chemical data (including dissolved oxygen content, pH stability, color change, bitterness retention rate), diatomaceous earth dosage data, and wine filtration efficiency data are extracted from the historically generated data. And other data; then the production curve group is comprehensively evaluated based on the historical production data to optimize the score of the production curve of each batch in the production curve group. As a model constraint condition, constraints can be imposed from the different dimensions to ensure good quality, appropriate cost and high efficiency of beer production. The different dimensions may include quality dimension, cost dimension, and efficiency dimension. That is, the opening of the addition pump is controlled by predicting the comprehensive evaluation score of the current production batch.

[0081] For example, the production curve of each batch is determined based on the comprehensive evaluation scores of turbidity data, differential pressure data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data. Specifically, the comprehensive evaluation of historical production data can be calculated according to the formula, that is, the comprehensive evaluation of the current production number and the comprehensive evaluation of the production batch before prediction within a preset time period can also be calculated based on the current production data.

[0082] The calculation formula is:

[0083] Among them, Score batch Indicates the batch comprehensive evaluation score of the brand batch or filtered batch; S i represents the evaluation score of historical production data; w i Represents the weight of historical production data, and the sum of the weights is 1.

[0084] The scoring calculation process for turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data is as follows.

[0085] Turbidity score deviation=(1-|turbidity target value−turbidity actual value / turbidity target value|)×100.

[0086] Differential pressure stability score = (1-flow standard deviation / average flow) × 100.

[0087] Microbial reduction rate score before and after filtration = (count before filtration – count after filtration) / count before filtration × 100.

[0088] Soil efficiency score = (1-|target soil usage value – actual soil usage value| / target soil usage value) × 100.

[0089] Wine filtration efficiency score = (actual single batch filtration time / planned single batch filtration time) × (actual filtration volume / planned filtration volume) × 100.

[0090] Dissolved oxygen content deviation score = (1-|oxygen content target value - oxygen content actual value| / oxygen content target value) × 100.

[0091] pH stability score = (1-pH standard deviation / average pH) × 100.

[0092] Calculation formula: Chroma change rate score = (chroma before filtering – chroma after filtering) / chroma before filtering × 100.

[0093] Bitterness value retention score=[1−(bitterness value before filtration−bitterness value after filtration) / bitterness value before filtration]×100.

[0094] The above method ensures that the comprehensive evaluation score of the beer in the production process is maximized, so the corresponding production curve is optimal. The diatomaceous earth filter is controlled according to the production curve to produce the best beer.

[0095] Of course, in some other embodiments, the evaluation may be performed in combination with other dimensions, such as the failure dimension, to reduce the failure rate.

[0096] In another embodiment, the opening control of the addition pump of the diatomite filter can be fine-tuned according to specific index characteristics in the production data. In some embodiments, the index characteristics can be index characteristics and state characteristics. The AI ​​control model controls the opening of the addition pump by predicting the index characteristics.

[0097] The indicator characteristics are mainly used to characterize the process environment during beer production and the key indicators that affect beer production. Among them, the indicator characteristics include or pressure difference. The state characteristics are mainly used to characterize production batches, such as brand batches and filtration batches, etc. The state characteristics include brand, step number, and planned filtration volume. The AI ​​control model can judge the production batch based on the state characteristics. When training the AI ​​control model, all historical batches are classified according to the above characteristics as matching labels for online model reasoning.

[0098] When the AI ​​control model monitors indicator characteristics, it mainly monitors the values ​​of the indicator characteristics. When the values ​​of the indicator characteristics change, it is necessary to determine whether it is abnormal.

[0099] When training the AI ​​control model, it is necessary to ensure that the generated curve of diatomaceous earth addition is the optimal control of the entire production process. Therefore, when training the AI ​​control model, the production batch can be evaluated from different dimensions, and the optimal control strategy can be obtained by comprehensively evaluating the beer scores of the production batches.

[0100] See also Figure 3 The present invention also provides an intelligent control method for a diatomaceous earth filter based on AI technology, which includes the following steps.

[0101] S310: Training an AI control model based on historical production data, and generating a curve group of diatomaceous earth addition amount according to the AI ​​control model.

[0102] S320: Obtaining the index characteristics and state characteristics in the current production data, matching the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the state characteristics, and using the index characteristics for the AI ​​control model to adjust the curve of the diatomaceous earth addition amount corresponding to the current production batch to control the opening of the addition pump of the diatomaceous earth filter.

[0103] In some embodiments, the AI ​​control model obtains key indicators affecting beer production according to the indicator characteristics; the AI ​​control model obtains brand batches and filtration batches, etc. according to the state characteristics, and the AI ​​control model matches the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the state characteristics.

[0104] S330: The indicator characteristics are monitored in real time, and the AI ​​control model predicts the target indicator characteristics of a preset time period in the future. When the indicator characteristics monitored in real time deviate from the target indicator characteristics, the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the opening of the addition pump of the diatomaceous earth filter.

[0105] In one embodiment, the turbidity or pressure difference in the indicator feature is predicted and monitored. Based on the NeuralProphet algorithm, the turbidity of beer during production in a preset time period in the future and the pressure difference or turbidity in the tank can be predicted. When the turbidity or pressure difference predicted and monitored by the AI ​​control model based on the current production parameters deviates from the target indicator feature (i.e., the predicted turbidity or pressure difference), the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the opening of the addition pump of the diatomaceous earth filter.

[0106] like Figure 4As shown, the AI ​​control model outputs the main strategy for controlling the opening of the diatomite filter's dosing pump. The AI ​​control model predicts the current pressure difference trend. When the pressure difference threshold or trend may be abnormal, the strategy for controlling the diatomite dosing pump opening will be fine-tuned. If it is a normal working condition, no adjustment is required and the smoothed main strategy is output. The AI ​​control model predicts the current turbidity trend (here mainly 90° turbidity). When the turbidity threshold or trend may be abnormal, the strategy for controlling the diatomite dosing pump opening will be fine-tuned. If it is a normal working condition, no adjustment is required and the smoothed main strategy is output.

[0107] For example, common turbidity indicators include 90° turbidity, 25° turbidity and 11° turbidity, and the turbidity value is compared with the target value (0.5EBC-0.8EBC). The pressure difference indicator can be defined according to the type of wine actually produced. The pressure difference is the same as the turbidity, and is compared with the set target value.

[0108] For example, the NeuralProphet algorithm predicts the turbidity or pressure difference in the next 10 minutes based on the real-time turbidity and pressure difference data values. When the real-time monitored turbidity or pressure difference ten minutes later deviates from the predicted turbidity or pressure difference in the next 10 minutes, the addition pump opening of the diatomaceous earth filter is adjusted.

[0109] The present invention obtains production curves of different batches of diatomite addition amounts based on an AI control model, and controls the opening of a diatomite filter addition pump according to the production curve of the diatomite addition amount, thereby realizing control of the diatomite addition amount, being able to optimize the diatomite coating amount, increase the single batch filtration amount, and reduce the number of diatomite dismantling times, thereby achieving the purpose of diatomite saving; at the same time, being able to improve filtration efficiency and save water and electricity consumption.

[0110] See also Figure 5 , Figure 5 Schematic diagram of an AI-based diatomite filter intelligent control device in an embodiment of the present invention. The AI-based diatomite filter intelligent control device provided by the present invention comprises:

[0111] The first control module 51 is used to train the AI ​​control model based on historical production data, and match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the matching rules and the current production data.

[0112] The second control module 52 is used to obtain the current production data and match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the matching rules and the current production data.

[0113] The third control module 53 is used to predict the comprehensive evaluation result of the current production batch based on the AI ​​control model, and control the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch, and the opening of the addition pump is used to control the diatomaceous earth addition amount.

[0114] Furthermore, when the AI ​​control model is trained based on historical production data, the production curve group is evaluated based on the comprehensive evaluation of the historical production data so that the score of the production curve of each batch in the production curve group is optimal. For example, the production curve of each batch is determined based on the comprehensive evaluation score of turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data.

[0115] In some embodiments, the AI ​​control model monitors the index characteristics in real time and predicts the target index characteristics of a preset time period in the future. When the index characteristics monitored in real time deviate from the target index characteristics, the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the addition pump opening of the diatomaceous earth filter. For example, when the turbidity or pressure difference monitored in real time deviates from the target index characteristics, the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the addition pump opening of the diatomaceous earth filter.

[0116] For the description of the intelligent control device for diatomaceous earth filter based on AI technology, please refer to the intelligent control method for diatomaceous earth filter based on AI technology, which will not be repeated here.

[0117] like Figure 6 As shown, the present invention also provides an electronic device, including:

[0118] At least one processor; and at least one memory in communication with the processor, wherein: the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the above-mentioned diatomaceous earth filter intelligent control method based on AI technology.

[0119] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned AI-based diatomaceous earth filter intelligent control method is implemented.

[0120] Computer-readable storage media may include: any entity or device capable of carrying a computer program, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc. A computer program includes a computer program code. The computer program code may be in source code form, object code form, an executable file, or some intermediate form, etc. A computer-readable storage medium may include: any entity or device capable of carrying a computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc.

[0121] In certain embodiments of the present invention, the electronic device may include a controller or a processor, and the controller is a single-chip microcomputer chip that integrates a processor, a memory, a communication module, etc. The processor may refer to the processor contained in the controller. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.

[0122] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

[0123] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0124] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent control method for diatomaceous earth filter based on AI technology, characterized in that: include: Training an AI control model based on historical production data, generating a curve group of diatomite addition amount according to the AI ​​control model, wherein the curve group of diatomite addition amount includes a plurality of curves of diatomite addition amount; Acquire current production data, and the AI ​​control model matches a curve of diatomaceous earth addition amount corresponding to the current production batch according to matching rules and current production data; The AI ​​control model predicts the comprehensive evaluation result of the current production batch, and controls the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch. The opening of the addition pump is used to control the diatomaceous earth addition amount.

2. The AI-based diatomaceous earth filter intelligent control method according to claim 1 is characterized in that: The AI ​​control model is trained based on historical production data, and a curve group of diatomaceous earth addition amount is generated according to the AI ​​control model, including: Training an AI control model based on historical production data, generating production curve groups corresponding to different production batches according to the AI ​​control model, wherein the historically generated data includes turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data; The production curve group is evaluated comprehensively according to historical production data so as to optimize the score of the production curve of each batch in the production curve group.

3. The intelligent control method of diatomaceous earth filter based on AI technology according to claim 2 is characterized in that: The comprehensive evaluation of the production curve group according to the historical production data to optimize the score of the production curve of each batch in the production curve group includes: The production curve of each batch is determined according to the comprehensive evaluation score of turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data; and the production curve is evaluated according to the following formula: Among them, Score batch Indicates the batch comprehensive evaluation score of the brand batch or filtered batch; S i represents the evaluation score of historical production data; w i Represents the weight of historical production data, and the sum of the weights is 1.

4. The AI-based diatomaceous earth filter intelligent control method according to claim 3 is characterized in that: The production curve of each batch is determined based on the comprehensive evaluation score of turbidity data, pressure difference data, microbial data, physical and chemical data, diatomaceous earth dosage data, and wine filtration efficiency data, including: calculating the data score by the following formula: Turbidity score deviation = (1-|turbidity target value – turbidity actual value / turbidity target value|) × 100; Pressure difference stability score = (1-flow standard deviation / average flow) × 100; Microbial reduction rate score before and after filtration = (count before filtration – count after filtration) / count before filtration × 100; Soil efficiency score = (1-|target soil usage value – actual soil usage value| / target soil usage value) × 100; Wine filtration efficiency score = (actual single batch filtration time / planned single batch filtration time) × (actual filtration volume / planned filtration volume) × 100; The physical and chemical data include dissolved oxygen content, pH stability, chromaticity change, and bitterness retention rate; Dissolved oxygen content deviation score = (1-|oxygen content target value – oxygen content actual value| / oxygen content target value) × 100; pH stability score = (1-pH standard deviation / average pH) × 100; Calculation formula: Chroma change rate score = (chroma before filtering – chroma after filtering) / chroma before filtering × 100; Bitterness value retention score=[1−(bitterness value before filtration−bitterness value after filtration) / bitterness value before filtration]×100.

5. The intelligent control method of diatomaceous earth filter based on AI technology according to claim 1 is characterized in that: The obtaining of current production data, and the matching of the AI ​​control model with a curve of the amount of diatomaceous earth added corresponding to the current production batch according to the current production data, include: Obtain the indicator characteristics and state characteristics in the current production data, match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the state characteristics, and use the indicator characteristics for the AI ​​control model to adjust the curve of the diatomaceous earth addition amount corresponding to the current production batch to control the opening of the addition pump of the diatomaceous earth filter; wherein the state characteristics include brand, step number, and planned wine filtration volume.

6. The AI-based diatomaceous earth filter intelligent control method according to claim 5 is characterized in that: The index characteristics include turbidity and pressure difference.

7. The AI-based diatomaceous earth filter intelligent control method according to claim 5 is characterized in that: Controlling the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch includes: The indicator characteristics are monitored in real time, and the AI ​​control model predicts the target indicator characteristics of a preset time period in the future. When the indicator characteristics monitored in real time deviate from the target indicator characteristics, the curve of the diatomaceous earth addition amount corresponding to the current production batch is adjusted to control the opening of the addition pump of the diatomaceous earth filter.

8. The AI-based diatomaceous earth filter intelligent control method according to claim 1 is characterized in that: The training of the AI ​​control model based on historical production data includes: The AI ​​control model is trained using XGBoost algorithm, LightGBM algorithm, or NeuralProphet algorithm and historical production data.

9. An intelligent control device for diatomaceous earth filter based on AI technology, characterized in that: include: A first control module is used to train an AI control model based on historical production data, and generate a curve group of diatomite addition amount according to the AI ​​control model, wherein the curve group of diatomite addition amount includes a plurality of curves of diatomite addition amount; The second control module is used to obtain the current production data and match the curve of the diatomaceous earth addition amount corresponding to the current production batch according to the matching rules and the current production data; The third control module is used to predict the comprehensive evaluation result of the current production batch based on the AI ​​control model, and control the opening of the addition pump of the diatomaceous earth filter according to the comprehensive evaluation result and the curve of the diatomaceous earth addition amount corresponding to the current production batch, and the opening of the addition pump is used to control the diatomaceous earth addition amount.

10. An electronic device, characterized in that: include: at least one processor; And at least one memory communicatively connected to the processor, wherein: the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the diatomaceous earth filter intelligent control method based on AI technology as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is run by a computer, the intelligent control method for a diatomaceous earth filter based on AI technology as claimed in any one of claims 1 to 8 is executed.

Images