Veterinary disinfectant production line monitoring method, device and equipment based on artificial intelligence

Abstract

The invention discloses a veterinary disinfectant production line monitoring method, device and equipment based on artificial intelligence, and relates to the technical field of production line monitoring. According to the veterinary disinfectant production line monitoring method based on artificial intelligence, effective raw materials are screened by acquiring expected standard data (including seven key indexes such as skin mucous membrane irritation of livestock and poultry) of a veterinary disinfectant, and a raw material ratio is determined by utilizing a random forest decision; data are monitored in three stages of initial mixing, stable reaction and homogeneous curing to judge the mixing effect, equipment parameters in the corresponding stages are adjusted based on a Q value strategy when the mixing effect does not reach the standard, full-process intelligent monitoring is realized by means of an artificial intelligence technology, and the stages from raw materials to mixing are dynamically optimized, so that the product is ensured to meet the expected standard, and the product quality is improved. The production efficiency and the product quality stability are improved, and the problems that in the production process of the veterinary disinfectant, raw material screening is not accurate, the proportion is unreasonable, and the mixing process is difficult to control, so that the product quality does not reach the standard, and the stability is poor are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of production line monitoring, in particular to a veterinary disinfectant production line monitoring method, device and equipment based on artificial intelligence. BACKGROUND

[0002] The purpose of disinfection in a farm is to eliminate pathogenic microorganisms that spread in the external environment, cut off the transmission route, and prevent the introduction or spread of epidemic diseases, so as to achieve the purpose of preventing, controlling and eliminating infectious diseases. All farms should establish a practical and feasible epidemic prevention and disinfection system and carry out disinfection work regularly. An important measure. When disinfecting a large-scale pig farm, the disinfection object should be clearly defined, and appropriate disinfectant and corresponding disinfection method should be selected for comprehensive and thorough disinfection.

[0003] The Chinese patent application file with publication number CN111189980A discloses a method for detecting and monitoring the type and concentration of disinfectant, which includes the following steps: measuring the PH values of different types of disinfectant mother liquor and after dilution in different proportions, establishing a data model of the change of PH value of each disinfectant and the corresponding PH value change range according to the measured PH value, and establishing a database of disinfectant types and concentrations; using the established database to determine the type and concentration of disinfectant and configure the disinfectant solution.

[0004] However, the existing veterinary disinfectant production process has the problems of inaccurate raw material screening, unreasonable proportioning, difficult control of the mixing process, resulting in substandard product quality and poor stability. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a veterinary disinfectant production line monitoring method, device and equipment based on artificial intelligence, which solves the problems of inaccurate raw material screening, unreasonable proportioning, difficult control of the mixing process, resulting in substandard product quality and poor stability in the production of veterinary disinfectants.

[0006] To achieve the above purpose, the present application is implemented by the following technical solutions: a veterinary disinfectant production line monitoring method based on artificial intelligence, comprising the following steps: obtaining veterinary disinfectant expected standard data, effectively screening veterinary disinfectant raw materials based on the veterinary disinfectant expected standard data, and determining effective veterinary disinfectant raw materials, the veterinary disinfectant expected standard data including livestock and poultry skin mucous membrane irritation index, drug-resistant bacteria gene induction risk value, raw material mutagenicity risk value, compatibility index with veterinary vaccines, organic impurity degradation rate, targeted bactericidal efficiency and thermal stability coefficient;

[0007] After determining the effective raw materials of the veterinary disinfectant, the random forest decision is used to process the expected standard data of the veterinary disinfectant, and the proportion of the effective raw materials of the veterinary disinfectant is determined; based on the determined proportion of the effective raw materials of the veterinary disinfectant, the effective raw materials of the veterinary disinfectant are mixed, and the disinfectant monitoring data of each mixing stage is obtained in the mixing process, the mixing stage including the initial mixing stage, the reaction stabilization stage and the homogenization curing stage; based on the disinfectant monitoring data, it is determined whether the disinfectant mixing effect of each mixing stage meets the standard:

[0008] If it meets the standard, the mixing continues until the mixing is completed.

[0009] If it does not meet the standard, the initial equipment parameter adjustment is performed on the mixing equipment based on the Q value strategy, and the initial equipment parameters include the initial parameters of the equipment in the initial mixing stage, the initial parameters of the equipment in the reaction stabilization stage and the initial parameters of the equipment in the homogenization curing stage.

[0010] Further, the effective raw materials of the veterinary disinfectant are screened based on the expected standard data of the veterinary disinfectant, and the effective raw materials of the veterinary disinfectant are determined, including the following steps: obtaining the historical characteristic spectrum matrix of each batch of various effective raw materials of the veterinary disinfectant, and simultaneously obtaining the corresponding veterinary disinfectant historical actual standard data, raw material historical proportion, initial mixing stage equipment historical parameter, reaction stabilization stage equipment historical parameter, homogenization curing stage equipment historical parameter, initial mixing stage disinfectant historical monitoring data, reaction stabilization stage disinfectant historical monitoring data and homogenization curing stage disinfectant historical monitoring data of the same batch of effective raw materials of the veterinary disinfectant; based on the block chain, the historical characteristic spectrum matrix and the corresponding veterinary disinfectant historical actual standard data, raw material historical proportion, initial mixing stage equipment historical parameter, reaction stabilization stage equipment historical parameter, homogenization curing stage equipment historical parameter, initial mixing stage disinfectant historical monitoring data, reaction stabilization stage disinfectant historical monitoring data and homogenization curing stage disinfectant historical monitoring data are stored;

[0011] The similarity analysis is performed on the veterinary disinfectant expected standard data and the veterinary disinfectant historical actual standard data stored in the block chain, the veterinary disinfectant actual standard data most similar to the veterinary disinfectant expected standard data is determined, and then the historical characteristic spectrum matrix of the corresponding various effective raw materials of the veterinary disinfectant is determined; the characteristic spectrum matrix of each veterinary disinfectant raw material is obtained, and the cosine similarity analysis is performed on the characteristic spectrum matrix of each veterinary disinfectant raw material and the historical characteristic spectrum matrix of the corresponding effective veterinary disinfectant raw material, to obtain the cosine similarity value corresponding to each veterinary disinfectant raw material; if the cosine similarity value is greater than the set similarity threshold value, the corresponding veterinary disinfectant raw material is the effective veterinary disinfectant raw material; if the cosine similarity value is not greater than the set similarity threshold value, the corresponding veterinary disinfectant raw material is not the effective veterinary disinfectant raw material, and is replaced.

[0012] Further, the effective veterinary disinfectant raw material ratio is determined, including the following steps:

[0013] The veterinary disinfectant actual standard data of each batch of veterinary disinfectant is fused and processed to obtain a historical disinfection efficiency reference value S ：

[0014]

[0015] Wherein, F1 and F2 are transfer functions, max(F1+F2) is the maximum value of the function value of the transfer functions F1 and F2, LPSI is the normalized livestock and poultry skin mucosa irritation index, DRG is the normalized drug-resistant bacteria gene induction risk value, RMM is the normalized raw material mutagenicity risk value, CIV is the normalized compatibility index with veterinary vaccine, OID is the normalized organic impurity degradation rate, oid is the normalized organic impurity degradation rate reference value, TBE is the normalized targeted bactericidal efficiency, TSC is the normalized thermal stability coefficient, and α1, α2, α3, α4, β1, β2 and β3 are weight coefficients;

[0016] The effective veterinary disinfectant raw material historical ratio of each batch of veterinary disinfectant is obtained, the effective veterinary disinfectant raw material historical ratio and the corresponding disinfection efficiency reference value are taken as the input of the random forest model, and the reference value-ratio formula is output, the reference value-ratio formula:

[0017] S=ω1*a1+ω2*a2+...+ω n *a n ;

[0018] Wherein, ω1, ω2, …, ω n are weight coefficients, a1, a2, …, a n The closure is the ratio of each effective veterinary disinfectant raw material, that is, the ratio of the mass of each effective veterinary disinfectant raw material to the total mass.

[0019] The expected standard data of the veterinary disinfectant is fused and processed to obtain an expected disinfection efficiency reference value;

[0020] Based on the reference value-ratio formula, the effective veterinary disinfectant raw material ratio solution set corresponding to the expected disinfection efficiency reference value is determined, if there are multiple solutions in the effective veterinary disinfectant raw material ratio solution set, the optimal solution is determined as the effective veterinary disinfectant raw material ratio, and if there is only one solution, it is directly output as the effective veterinary disinfectant raw material ratio.

[0021] Further, the optimal solution is determined, including the following steps: determining the effective veterinary disinfectant raw material to be determined for each solution, performing similarity analysis on the effective veterinary disinfectant raw material to be determined and the raw material historical ratio, determining the raw material historical ratio corresponding to each solution, and further determining the corresponding veterinary disinfectant historical actual standard data; calculating the sum C of the raw material cost corresponding to each solution; calculating the cosine similarity value B of the veterinary disinfectant historical actual standard data corresponding to each solution and the veterinary disinfectant expected standard data; calculating the evaluation value of the solution: Wherein, sC is the sum of the raw material cost reference value, e is a natural constant, λ1 and λ2 are weight factors; the solution corresponding to the maximum evaluation value is taken as the optimal solution.

[0022] Further, based on the disinfectant monitoring data, it is determined whether the disinfectant mixing effect of each mixing stage meets the standard, including the following steps: performing K-means clustering analysis on the veterinary disinfectant historical actual standard data to determine a plurality of standard data clustering centers, and based on the blockchain, obtaining the initial mixing stage disinfectant historical monitoring data, the reaction stable stage disinfectant historical monitoring data and the homogeneous curing stage disinfectant historical monitoring data corresponding to the veterinary disinfectant historical actual standard data in each standard data clustering center; performing mean processing and normalization on the veterinary disinfectant historical actual standard data corresponding to each standard data clustering center to obtain veterinary disinfectant historical actual standard feature data;

[0023] The initial mixing stage disinfectant historical monitoring data, the reaction stable stage disinfectant historical monitoring data and the homogeneous curing stage disinfectant historical monitoring data corresponding to each standard data clustering center are processed and normalized by mean processing to obtain historical monitoring feature data, including initial mixing stage disinfectant historical monitoring feature data, reaction stable stage disinfectant historical monitoring feature data and homogeneous curing stage disinfectant historical monitoring feature data; after the veterinary disinfectant expected standard data is normalized, the similarity comparison is performed between the veterinary disinfectant expected standard data and the veterinary disinfectant historical actual standard feature data corresponding to each standard data clustering center, and the standard data clustering center corresponding to the most similar veterinary disinfectant historical actual standard feature data is determined.

[0024] The disinfectant monitoring data of the effective veterinary disinfectant raw material in the mixing stage is obtained and normalized, and cosine similarity analysis is performed on the historical monitoring feature data corresponding to the mixing stage to obtain a monitoring data similarity value; if the monitoring data similarity value is greater than a set monitoring similarity threshold, the disinfectant mixing effect of the corresponding mixing stage meets the standard; if the monitoring data similarity value is not greater than the set monitoring similarity threshold, the disinfectant mixing effect of the corresponding mixing stage does not meet the standard.

[0025] Further, the initial device parameter acquisition process is as follows: based on the most similar veterinary disinfectant historical actual compliance feature data corresponding to the compliance data clustering center, the corresponding initial mixing stage device historical parameters, reaction stable stage device historical parameters and homogenization curing stage device historical parameters are obtained; the initial mixing stage device historical parameters, reaction stable stage device historical parameters and homogenization curing stage device historical parameters are statistically processed to determine the initial mixing stage device historical parameter interval, reaction stable stage device historical parameter interval and homogenization curing stage device historical parameter interval; the initial mixing stage device historical parameters are subjected to K-means clustering analysis to determine the initial mixing stage device parameter clustering center corresponding to the initial mixing stage device historical parameters, and the initial mixing stage device parameter clustering center corresponding to the most samples is determined;

[0026] The reaction stable stage device historical parameters are subjected to K-means clustering analysis to determine the reaction stable stage device parameter clustering center corresponding to the reaction stable stage device historical parameters, and the reaction stable stage device parameter clustering center corresponding to the most samples is determined; the homogenization curing stage device historical parameters are subjected to K-means clustering analysis to determine the homogenization curing stage device parameter clustering center corresponding to the homogenization curing stage device historical parameters, and the homogenization curing stage device parameter clustering center corresponding to the most samples is determined; the samples corresponding to the initial mixing stage device parameter clustering center corresponding to the most samples, the reaction stable stage device parameter clustering center corresponding to the most samples and the homogenization curing stage device parameter clustering center corresponding to the most samples are subjected to mean value processing respectively to obtain the initial mixing stage device initial parameters, reaction stable stage device initial parameters and homogenization curing stage device initial parameters.

[0027] Furthermore, the equipment parameters of the mixing equipment are adjusted based on the Q-value strategy, including the following steps: obtaining the disinfectant monitoring data of the current mixing stage, which is defined as a state space; obtaining an action space set stored in a database, wherein each action in the action space set corresponds to an equipment parameter adjustment amount of a group of mixing equipment, and each action also corresponds to a stored Q value; determining the action corresponding to the largest Q value, and adjusting the initial equipment parameters of the current mixing stage based on the equipment parameter adjustment amount corresponding to the action to obtain the adjusted equipment parameters; obtaining the equipment parameter limit interval obtained from the database, and verifying the adjusted equipment parameters based on the equipment parameter limit interval, the adjusted equipment parameters and the equipment parameter adjustment amount: if the verification passes, the mixing equipment is adjusted based on the adjusted equipment parameters; if the verification fails, the Q value is reduced based on the reduction value set in the database, and the action corresponding to the largest Q value is re-determined, and the adjusted equipment parameters are verified again; after adjusting the mixing equipment based on the adjusted equipment parameters, the disinfectant monitoring data after the set time is obtained, and the state space is updated; based on the updated disinfectant monitoring data, the disinfectant mixing effect is evaluated again to determine whether to adjust the equipment parameters again, and the Q value is updated.

[0028] Furthermore, the disinfectant mixing effect is evaluated again based on the updated disinfectant monitoring data to determine whether the equipment parameters are adjusted again and the Q value is updated, including the following steps: obtaining the disinfectant monitoring data after the set time after the equipment adjustment, and obtaining the short-term evaluation effect based on the disinfectant monitoring data after the set time; inputting the disinfectant monitoring data after the set time and the current adjusted equipment parameters into the trained LSTM model to obtain the disinfectant monitoring data after the end of the current mixing stage, and obtaining the long-term evaluation effect based on the disinfectant monitoring data after the end of the current mixing stage; inputting the disinfectant monitoring data after the end of the current mixing stage and the initial equipment parameters of the next mixing stage into the trained LSTM model to obtain the disinfectant monitoring data after the end of the next mixing stage, and obtaining the auxiliary long-term evaluation effect based on the disinfectant monitoring data after the end of the next mixing stage; based on the short-term evaluation effect Long-term evaluation effect and assist in long-term evaluation of the effect Get the predicted effect of disinfectant mixing Ψ ：

[0029]

[0030] Among them, k is the auxiliary long-term evaluation effect The number of , e is a natural constant;

[0031] If the disinfectant mixing prediction effect is less than the set prediction effect evaluation value, no device parameter adjustment is performed, and the Q value is increased based on the increase value stored in the database for Q value updating; if the disinfectant mixing prediction effect is not less than the set prediction effect evaluation value, device parameter adjustment is performed, and the Q value is reduced based on the reduction value stored in the database for Q value updating.

[0032] The AI-based veterinary disinfectant production line monitoring method and device, comprising an effective raw material determination module, a ratio determination module and a mixing adjustment module, wherein: the effective raw material determination module is used to obtain veterinary disinfectant expected standard data, screen effective veterinary disinfectant raw materials based on the veterinary disinfectant expected standard data, and determine effective veterinary disinfectant raw materials; the veterinary disinfectant expected standard data includes livestock skin mucous membrane irritation index, drug-resistant bacteria gene induction risk value, raw material mutagenicity risk value, compatibility index with veterinary vaccines, organic impurity degradation rate, targeted bactericidal efficiency and thermal stability coefficient; the ratio determination module is used to determine the effective veterinary disinfectant raw material ratio after determining the effective veterinary disinfectant raw material by using random forest decision to process the veterinary disinfectant expected standard data; the mixing adjustment module is used to mix each effective veterinary disinfectant raw material based on the determined effective veterinary disinfectant raw material ratio, and obtain disinfectant monitoring data at each mixing stage in the mixing process; the mixing stages include an initial mixing stage, a reaction stabilization stage and a homogenization curing stage; whether the disinfectant mixing effect at each mixing stage meets the standard is determined based on the disinfectant monitoring data; if it meets the standard, the mixing continues until the mixing is completed; if it does not meet the standard, the initial device parameter adjustment is performed on the mixing equipment based on the Q value strategy; the initial device parameters include initial parameters of the initial mixing stage equipment, initial parameters of the reaction stabilization stage equipment and initial parameters of the homogenization curing stage equipment.

[0033] An electronic device, comprising: a processor; a computer readable storage medium; and a memory having stored therein computer program instructions, which, when executed by the processor, cause the processor to perform the AI-based veterinary disinfectant production line monitoring method as described above; and a computer readable storage medium for storing a program, which, when executed by a processor, implements the AI-based veterinary disinfectant production line monitoring method as described above.

[0034] The present application has the following beneficial effects:

[0035] The artificial intelligence-based veterinary disinfectant production line monitoring method screens effective raw materials by obtaining veterinary disinfectant expected standard data (covering seven key indicators such as livestock and poultry skin mucous membrane irritation), determines the raw material ratio by using random forest decision, and monitors data in the initial mixing, reaction stabilization and homogenization curing stages to judge the mixing effect. When the standard is not met, the corresponding stage equipment parameters are adjusted based on the Q value strategy, and the whole process is intelligently monitored by means of artificial intelligence technology, dynamically optimizing each stage from raw materials to mixing, ensuring that the product meets the expected standard, improving production efficiency and product quality stability, and solving the problems of inaccurate raw material screening, unreasonable ratio, difficult control of the mixing process, resulting in substandard product quality and poor stability in veterinary disinfectant production.

[0036] Of course, implementing any product of the present application does not necessarily require all the advantages described above to be achieved at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 The flowchart of the artificial intelligence-based veterinary disinfectant production line monitoring method of the present application.

[0038] Figure 2 The flowchart of the artificial intelligence-based veterinary disinfectant production line monitoring device of the present application. DETAILED DESCRIPTION

[0039] Please refer to Figure 1 The technical scheme provided by the embodiments of the present application is: an artificial intelligence-based veterinary disinfectant production line monitoring method, comprising the following steps: obtaining veterinary disinfectant expected standard data, screening effective raw materials for veterinary disinfectant raw materials based on the veterinary disinfectant expected standard data, and determining effective veterinary disinfectant raw materials. The veterinary disinfectant expected standard data includes livestock and poultry skin mucous membrane irritation index, drug-resistant bacteria gene induction risk value, raw material mutagenicity risk value, vaccine compatibility index, organic impurity degradation rate, targeted bactericidal efficiency and thermal stability coefficient, which comprehensively covers the core performance and safety requirements of veterinary disinfectants: the irritation index ensures the safety of livestock and poultry use, the drug-resistant bacteria induction risk and mutagenicity risk are related to public health and ecological safety, the vaccine compatibility index avoids affecting the epidemic prevention effect, the organic impurity degradation rate relates to environmental friendliness, the targeted bactericidal efficiency ensures the disinfection effect, and the thermal stability coefficient ensures the performance stability in storage and use. The combination of these parameters can accurately define the quality standard of disinfectants, provide scientific basis for raw material screening, ratio optimization and production process monitoring, and has the advantages of constructing an evaluation system from multiple dimensions such as safety, efficiency, compatibility and stability to ensure that the produced veterinary disinfectant is efficient and reliable, and meets the actual needs of the livestock industry. The acquisition methods are as follows:

[0040] Skin mucous membrane irritation index of livestock and poultry, 9-day-old chicken embryos were taken, and after the down chorioallantoic membrane was peeled off, the embryos were exposed to the diluted disinfectant sample (concentration was 1.5 times of the actual use concentration) for continuous observation of the membrane bleeding, dissolution, coagulation reaction within 5 minutes. Quantification was performed according to the standard scoring table (0-10 points), 0 points for no irritation, 10 points for strong irritation, and the final index was the average score of 3 parallel tests.

[0041] Drug-resistant bacteria gene induction risk value, a target drug-resistant gene (such as the blaCTX-M gene of Escherichia coli) was selected, the disinfectant sample was co-cultured with a sensitive strain for 24 hours, and then bacterial DNA was extracted for real-time fluorescent quantitative PCR. The relative expression of the gene was calculated by the Ct value (cycle threshold value): when the Ct value was > 35, the risk value was 0.1 (low risk); when 25 < Ct value ≤ 35, the risk value was 0.5 (medium risk); when Ct value ≤ 25, the risk value was 0.9 (high risk), and the larger the value, the higher the induction risk.

[0042] Raw material mutagenicity risk value, the extract of the disinfectant raw material was mixed with a histidine-deficient strain of Salmonella typhimurium, and was cultured in a histidine-free medium for 48 hours. The number of revertant bacteria was counted. The threshold was twice the number of spontaneous mutant bacteria, and the risk value was calculated as (measured number of bacteria / number of spontaneous bacteria) / 2, with two decimal places.

[0043] Compatibility index with veterinary vaccines, commonly used veterinary vaccines (such as swine fever vaccine, avian influenza vaccine) were selected, the vaccine was mixed with the disinfectant at a ratio of 1:10, and then was placed at 37°C for 1 hour. The retention rate of the vaccine antigen titer was detected by ELISA. The calculation formula was: compatibility index = (titer after mixing / titer without mixing) × 100%, and the integer was retained.

[0044] Degradation rate of organic impurities, the disinfectant was added to a solution containing standard organic impurities (such as bovine serum albumin, concentration 10 g / L), and the TOC values at 0 hour and 24 hours were measured by a total organic carbon (TOC) analyzer. Degradation rate = (initial TOC value - 24-hour TOC value) / initial TOC value × 100% / 24h, the result was in the unit of "% / h" (range 0-5% / h, the higher the value, the stronger the degradation ability).

[0045] Targeted bactericidal efficacy, for target pathogenic bacteria (such as Staphylococcus aureus, porcine circovirus), a suspension quantitative bactericidal test was used: the disinfectant was mixed with the bacterial suspension (concentration 10 6 CFU / mL) at a certain ratio, and then was inoculated on a culture dish after being acted for 10 minutes, and the number of surviving bacteria was counted. Bactericidal efficacy = (initial number of bacteria - number of surviving bacteria) / initial number of bacteria × 100%, and the integer was retained.

[0046] Thermal stability coefficient, after sealing the disinfectant in a 54°C incubator for accelerated aging for 14 days, the content of active ingredients before and after aging was measured (such as potassium hydrogen persulfate was measured by iodometric method). Thermal stability coefficient = (content after aging / initial content) x 100%, keep the integer.

[0047] The historical characteristic spectrum matrix of each batch of various effective veterinary disinfectant raw materials is obtained, and the corresponding veterinary disinfectant historical actual standard data, raw material historical proportion, initial mixing stage equipment historical parameter, reaction stabilization stage equipment historical parameter, homogenization and curing stage equipment historical parameter, initial mixing stage disinfectant historical monitoring data, reaction stabilization stage disinfectant historical monitoring data and homogenization and curing stage disinfectant historical monitoring data of the same batch of effective veterinary disinfectant raw materials are obtained; provide comprehensive and related historical reference for subsequent raw material screening, ensure that the analysis is based on complete production chain information, and improve the integrity and relevance of data support.

[0048] The historical characteristic spectrum matrix and the corresponding veterinary disinfectant historical actual standard data, raw material historical proportion, initial mixing stage equipment historical parameter, reaction stabilization stage equipment historical parameter, homogenization and curing stage equipment historical parameter, initial mixing stage disinfectant historical monitoring data, reaction stabilization stage disinfectant historical monitoring data and homogenization and curing stage disinfectant historical monitoring data are stored based on the blockchain; the above historical data is stored based on the blockchain, which has the advantages of using the characteristics of the blockchain that cannot be tampered with and can be traced back to ensure the authenticity and security of the historical data, and provides a reliable data basis for data reuse, similarity analysis, etc., avoiding the influence of data tampering or loss on the accuracy of subsequent processes.

[0049] The veterinary disinfectant expected standard data is subjected to similarity analysis with each veterinary disinfectant historical actual standard data stored in the blockchain to determine the veterinary disinfectant actual standard data most similar to the veterinary disinfectant expected standard data, and then determine the corresponding historical characteristic spectrum matrix of various effective veterinary disinfectant raw materials; by correlating the expected standard with the historical standard case, the adaptive raw material characteristic reference is quickly located, which provides a precise comparison benchmark for raw material effectiveness judgment.

[0050] The characteristic spectrum matrix of various veterinary disinfectant raw materials is obtained, and the characteristic spectrum matrix of various veterinary disinfectant raw materials is subjected to cosine similarity analysis with the corresponding historical characteristic spectrum matrix of effective veterinary disinfectant raw materials, respectively, to obtain the cosine similarity value corresponding to various veterinary disinfectant raw materials; by virtue of the characteristics of cosine similarity energy quantization feature matching degree, the consistency of the current raw material and the effective historical raw material is scientifically judged, and the objectivity and accuracy of raw material screening are improved.

[0051] If the cosine similarity value is greater than the set similarity threshold value, the corresponding veterinary disinfectant raw material is an effective veterinary disinfectant raw material; if the cosine similarity value is not greater than the set similarity threshold value, the corresponding veterinary disinfectant raw material is not an effective veterinary disinfectant raw material, and is replaced. The clear threshold value realizes clear definition of the effectiveness of the raw material, ensures that only raw materials meeting the characteristic requirements enter production, guarantees product quality from the source, and reduces the impact of unqualified raw materials on production.

[0052] After determining the effective veterinary disinfectant raw material, the random forest decision is used to process the veterinary disinfectant expected compliance data to determine the effective veterinary disinfectant raw material ratio; the random forest can process multi-dimensional data and has strong anti-overfitting ability, and can mine the complex correlation between compliance data and ratio based on historical data, and improve the scientificity of ratio determination.

[0053] The veterinary disinfectant actual compliance data of each batch of historical veterinary disinfectant is fused and processed to obtain a historical disinfection efficiency reference value. By fusing multi-dimensional actual compliance data, the dispersed indicators are integrated into a unified reference value, which is convenient for subsequent corresponding analysis with expected data and simplifies the data processing dimension. The historical disinfection efficiency reference value S is:

[0054]

[0055] Wherein, F1 and F2 are transfer functions, max(F1+F2) is the maximum value of the function value of the transfer functions F1 and F2, LPSI is the normalized livestock and poultry skin mucous membrane irritation index, DRG is the normalized drug-resistant bacteria gene induction risk value, RMM is the normalized raw material mutagenicity risk value, CIV is the normalized compatibility index with veterinary vaccines, OID is the normalized organic impurity degradation rate, oid is the normalized organic impurity degradation rate reference value, TBE is the normalized targeted bactericidal efficiency, TSC is the normalized thermal stability coefficient, and α1, α2, α3, α4, β1, β2 and β3 are weight coefficients.

[0056] The historical disinfection efficiency reference value formula can normalize indicators of different dimensions through transfer functions, normalization processing and weight coefficients, highlight the influence of key indicators (such as reflecting the importance of each parameter through weight coefficients), and reasonably integrate information with the help of the maximum value of the transfer function, so that the reference value more accurately reflects the comprehensive level of disinfection efficiency.

[0057] The effective veterinary disinfectant raw material historical ratio of each batch of veterinary disinfectant is obtained, the effective veterinary disinfectant raw material historical ratio and the corresponding disinfection efficiency reference value are taken as the input of the random forest model, and the reference value-ratio formula is output. The reference value-ratio formula is:

[0058] S=ω1*a1+ω2*a2+...+ω n *an ;

[0059] wherein ω1, ω2, …, ω n are weight coefficients, a1, a2, …, a n the closed is the ratio of each effective veterinary disinfectant raw material, that is, the ratio of the mass of each effective veterinary disinfectant raw material to the total mass; the energy of each raw material ratio on the overall performance is quantified, which provides a clear mathematical basis for backstepping the ratio according to the expected benchmark value, and improves the accuracy of the ratio calculation.

[0060] The expected disinfection performance benchmark value is obtained by fusing the expected standard data of the veterinary disinfectant.

[0061] Based on the benchmark value-ratio formula, the effective veterinary disinfectant raw material ratio solution set corresponding to the expected disinfection performance benchmark value is determined, if there are multiple solutions in the effective veterinary disinfectant raw material ratio solution set, the optimal solution is determined as the effective veterinary disinfectant raw material ratio, and if there is only one solution, it is directly output as the effective veterinary disinfectant raw material ratio.

[0062] Based on the formula, the ratio solution set is determined, and the optimal solution is selected or directly output according to the number of solutions, which can cope with multiple solutions, ensure that the final ratio is more optimal (such as cost, effect matching degree, etc.) while meeting the performance, and directly output when there is only one solution to improve efficiency and ensure that the ratio meets the expected requirements.

[0063] Determining the optimal solution includes the following steps: determining the effective veterinary disinfectant raw material to-be-determined ratio corresponding to each solution, performing similarity analysis on the effective veterinary disinfectant raw material to-be-determined ratio and the raw material historical ratio, determining the raw material historical ratio corresponding to each solution, and then determining the corresponding veterinary disinfectant historical actual standard data; verifying the feasibility of the to-be-determined ratio through historical data, providing actual effect reference for subsequent evaluation, and ensuring the rationality of the solution.

[0064] The sum C of the cost of each solution is calculated; the cost factor is taken into account in the evaluation to avoid focusing only on the effect and ignoring the economy, and the balance between cost and effect is achieved. The cosine similarity value B of the veterinary disinfectant historical actual standard data and the veterinary disinfectant expected standard data corresponding to each solution is calculated; the matching degree of the actual effect and the expected effect corresponding to the solution is quantified, and the selected solution can meet the quality requirements. The evaluation value of the solution is calculated: wherein sC is the sum of the raw material cost reference value, e is the natural constant, λ1 and λ2 are weight factors; the solution corresponding to the maximum evaluation value is taken as the optimal solution.

[0065] The influence weight of cost and effect is flexibly adjusted by using natural constant and weight factor, both are quantified as unified evaluation value, and the comprehensive and objective evaluation of solution is realized. The solution corresponding to the maximum evaluation value is selected as the optimal solution, which has the advantages of screening the best comprehensive effect (both close to the expected quality and reasonable cost) among multiple feasible solutions, and improving the economy and product quality reliability of production.

[0066] Based on the determined effective veterinary disinfectant raw material ratio, each effective veterinary disinfectant raw material is mixed, and disinfectant monitoring data of each mixing stage is obtained in the mixing process. The mixing stage includes the initial mixing stage, the reaction stabilization stage and the homogenization curing stage. Based on the disinfectant monitoring data, it is determined whether the disinfectant mixing effect of each mixing stage meets the standard. The disinfectant monitoring data of the mixing process is obtained in three stages of initial mixing, reaction stabilization and homogenization curing. The dynamic changes of the mixing process can be accurately captured in stages, and the targeted data for effect evaluation of each stage is provided, avoiding the ambiguity of overall monitoring.

[0067] K-means clustering analysis is performed on the historical actual standard data of veterinary disinfectants to determine multiple standard data clustering centers. The dispersed historical data is classified, and representative standard modes are extracted to provide a clear reference benchmark for effect evaluation. Based on the blockchain, the initial mixing stage disinfectant historical monitoring data, the reaction stabilization stage disinfectant historical monitoring data and the homogenization curing stage disinfectant historical monitoring data corresponding to the veterinary disinfectant historical actual standard data in each standard data clustering center are obtained. The veterinary disinfectant historical actual standard data corresponding to each standard data clustering center is processed by mean value and normalized to obtain veterinary disinfectant historical actual standard characteristic data. The dimensional difference of data is eliminated, and the data of different batches and different indicators can be directly compared to improve the universality of the characteristic data.

[0068] The initial mixing stage disinfectant historical monitoring data, the reaction stabilization stage disinfectant historical monitoring data and the homogenization curing stage disinfectant historical monitoring data corresponding to each standard data clustering center are processed by mean value and normalized to obtain historical monitoring characteristic data, including initial mixing stage disinfectant historical monitoring characteristic data, reaction stabilization stage disinfectant historical monitoring characteristic data and homogenization curing stage disinfectant historical monitoring characteristic data. Forming a standardized monitoring reference template of each stage facilitates comparison with current monitoring data and simplifies the evaluation process.

[0069] After normalizing the expected standard data of veterinary disinfectants, the similarity of the veterinary disinfectant historical actual standard characteristic data corresponding to each standard data clustering center is compared, and the standard data clustering center corresponding to the most similar veterinary disinfectant historical actual standard characteristic data is determined. By normalizing the data scale, the closest historical case to the expected effect is accurately matched by using similarity, providing an adaptive reference standard for current monitoring.

[0070]

[0071] wherein SIC represents the similarity between the normalized expected data of the veterinary disinfectant and the historical actual data of the veterinary disinfectant, the greater the SIC, the more similar, A i is the i-th normalized data in the normalized expected data of the veterinary disinfectant, B i is the i-th data of the historical actual data of the veterinary disinfectant; the similarity calculation formula quantifies the matching degree of each corresponding item in the expected data and the historical characteristic data, objectively reflects the similarity of the two, provides clear numerical basis for the selection of clustering center, and improves the accuracy of matching.

[0072] The disinfectant monitoring data of the effective veterinary disinfectant raw material in the mixing stage is obtained and normalized, and cosine similarity analysis is performed on the historical monitoring characteristic data of the corresponding mixing stage to obtain a monitoring data similarity value; if the monitoring data similarity value is greater than the set monitoring similarity threshold value, the disinfectant mixing effect of the corresponding mixing stage meets the standard; the consistency of the current monitoring data and the historical data meeting the standard is quantified by cosine similarity, and the mixing effect is scientifically judged. According to the comparison result of the monitoring data similarity value and the threshold value, it is judged whether the mixing effect meets the standard, the standardization of effect evaluation is realized by the clear threshold value, and it is ensured that the mixing effect of each stage meets the expectation, and the quality of the final product is guaranteed.

[0073] If the monitoring data similarity value is not greater than the set monitoring similarity threshold value, the disinfectant mixing effect of the corresponding mixing stage does not meet the standard.

[0074] If it meets the standard, continue to mix until the mixing is completed;

[0075] If it does not meet the standard, the initial equipment parameter adjustment is performed on the mixing equipment based on the Q value strategy, and the initial equipment parameters include the initial mixing stage equipment initial parameter, the reaction stabilization stage equipment initial parameter and the homogenization and curing stage equipment initial parameter.

[0076] The initial equipment parameter acquisition process is as follows: based on the most similar veterinary disinfectant historical actual data meeting the standard characteristic data corresponding to the data clustering center, the corresponding initial mixing stage equipment historical parameter, reaction stabilization stage equipment historical parameter and homogenization and curing stage equipment historical parameter are obtained; ensure that the obtained historical parameters are highly matched with the current production expected effect, and provide a reference basis for the actual demand of the initial parameter setting.

[0077] The initial mixing stage equipment historical parameters, the reaction stable stage equipment historical parameters and the homogenization and aging stage equipment historical parameters are counted to determine the initial mixing stage equipment historical parameter interval, the reaction stable stage equipment historical parameter interval and the homogenization and aging stage equipment historical parameter interval; the equipment historical parameters of each stage are counted to determine the parameter interval, which has the advantages that the reasonable range of the parameters is determined, the boundary for subsequent parameter adjustment is drawn, and the parameters are prevented from exceeding the effective working range. The initial mixing stage equipment historical parameters are subjected to K-means clustering analysis to determine the initial mixing stage equipment parameter clustering center corresponding to the initial mixing stage equipment historical parameters, and the initial mixing stage equipment parameter clustering center corresponding to the most samples; the equipment historical parameters of each stage are subjected to K-means clustering to determine the clustering center corresponding to the most samples, and the most widely used and most stable parameter mode in the historical production is extracted, so that the initial parameters are more representative and reliable.

[0078] The reaction stable stage equipment historical parameters are subjected to K-means clustering analysis to determine the reaction stable stage equipment parameter clustering center corresponding to the reaction stable stage equipment historical parameters, and the reaction stable stage equipment parameter clustering center corresponding to the most samples; the homogenization and aging stage equipment historical parameters are subjected to K-means clustering analysis to determine the homogenization and aging stage equipment parameter clustering center corresponding to the homogenization and aging stage equipment historical parameters, and the homogenization and aging stage equipment parameter clustering center corresponding to the most samples; the samples corresponding to the initial mixing stage equipment parameter clustering center corresponding to the most samples, the reaction stable stage equipment parameter clustering center corresponding to the most samples and the homogenization and aging stage equipment parameter clustering center corresponding to the most samples are subjected to mean value processing to obtain the initial mixing stage equipment initial parameters, the reaction stable stage equipment initial parameters and the homogenization and aging stage equipment initial parameters. The individual data fluctuation influence is eliminated through mean value calculation to obtain the optimal initial value of the equipment parameters of each stage, and stable starting settings are provided for production.

[0079] The equipment parameter adjustment of the mixing equipment based on the Q value strategy includes the following steps: obtaining the disinfectant monitoring data of the current mixing stage, which is defined as a state space; the real-time state of the current equipment operation and mixing effect can be accurately captured, which provides an initial basis for subsequent parameter adjustment. A set of action spaces stored in the database is obtained, each action in the set of action spaces corresponds to a set of equipment parameter adjustment amounts of the mixing equipment, and each action also corresponds to a stored Q value; a diversified predefined scheme is provided for parameter adjustment, and the Q value can reflect the effectiveness of the action, which facilitates the rapid selection of a suitable adjustment strategy.

[0080] The action corresponding to the maximum Q value is determined, and the initial device parameter of the current mixing stage is adjusted based on the device parameter adjustment amount corresponding to the action to obtain an adjusted device parameter; the adjustment scheme with better historical performance is preferentially selected to improve the effectiveness of parameter adjustment and quickly approach the target state. The device parameter limit interval obtained from the database is obtained, and the adjusted device parameter is verified based on the device parameter limit interval, the adjusted device parameter, and the device parameter adjustment amount, to ensure that the adjusted parameter is within the safe and effective operation range of the device, and to avoid device failure or product quality problems caused by abnormal parameters.

[0081] The specific verification process is to determine whether the adjusted device parameter exceeds the device parameter limit interval, and if so, the verification fails, and then it is determined whether the device parameter adjustment amount is greater than the set adjustment threshold, and if so, the verification fails.

[0082] If the verification passes, the mixing device is adjusted based on the adjusted device parameter; if the verification fails, the Q value is reduced based on the reduction value set in the database, and the action corresponding to the maximum Q value is determined again, and the adjusted device parameter verification is performed again; by reducing the Q value of the invalid action, the probability of its being selected again is reduced, the action selection is continuously optimized, and the adjustment efficiency is improved.

[0083] After adjusting the mixing device based on the adjusted device parameter, the disinfectant monitoring data after a set time is obtained, and the state space is updated; the actual effect after parameter adjustment is reflected in a timely manner to provide the latest state reference for subsequent evaluation and adjustment again. The disinfectant mixing effect is evaluated again based on the updated disinfectant monitoring data, and it is determined whether to adjust the device parameter again, and the Q value is updated. Through the feedback mechanism, the Q value is dynamically optimized, the effectiveness evaluation of the action is more in line with the actual situation, and the accuracy and adaptability of subsequent parameter adjustment are improved.

[0084] Based on the updated disinfectant monitoring data, the disinfectant mixing effect is evaluated again, it is determined whether to adjust the device parameter again, and the Q value is updated, including the following steps: obtaining disinfectant monitoring data after a set time after device adjustment, and obtaining short-term evaluation effect based on disinfectant monitoring data after a set time; the immediate effect after parameter adjustment is fed back in a timely manner to provide a short-term basis for subsequent evaluation and timely capture the effect change in the early stage of adjustment.

[0085] The disinfectant monitoring data after a set time and the current adjusted device parameter are input into the trained LSTM model to obtain disinfectant monitoring data after the current mixing stage, and the long-term evaluation effect is obtained based on the disinfectant monitoring data after the current mixing stage; the long-term evaluation effect is obtained by using the LSTM model to predict the monitoring data after the current mixing stage, and the influence of adjustment on the final effect of the current stage is determined in advance, and the forward-looking of the evaluation is enhanced.

[0086] The disinfectant monitoring data after the current mixing stage ends and the initial equipment parameters of the next mixing stage are input into the trained LSTM model to obtain the disinfectant monitoring data after the next mixing stage ends, and the auxiliary long-time evaluation effect is obtained based on the disinfectant monitoring data after the next mixing stage ends; considering the potential influence of adjustment on the subsequent stage, the whole process effect prediction is realized, and the adverse effect of local adjustment on the overall production is avoided. Based on the short-time evaluation effect Long-time evaluation effect And auxiliary long-time evaluation effect Obtain disinfectant mixing prediction effect Ψ ：

[0087] Wherein, k is the number of auxiliary long-time evaluation effect , e is a natural constant; the disinfectant mixing prediction effect formula combines short-time, long-time and auxiliary long-time evaluation effect, introduces natural constant and auxiliary effect number k, the advantage is to integrate multi-dimensional evaluation results by weighting, and the immediate and long-term, local and overall effects are considered, so that the prediction is more comprehensive and objective.

[0088] The disinfectant monitoring data includes uniformity, effective component conversion rate, system temperature, system viscosity and target sterilization efficiency. The equipment parameters include stirring speed, vacuum degree, stirring paddle angle and pressure in the stirring tank.

[0089] When the current mixing stage is the initial mixing stage, the short-time evaluation effect The calculation formula is as follows:

[0090]

[0091] JY for uniformity, ZH for effective component conversion rate, WD for system temperature, ND for system viscosity, BX for target sterilization efficiency, Jy for required uniformity of initial mixing stage, Zh for required effective component conversion rate of initial mixing stage, Wd for required system temperature of initial mixing stage, Nd for required system viscosity of initial mixing stage, Bx for required target sterilization efficiency of initial mixing stage;

[0092] And the long-time evaluation effect is the disinfectant monitoring data at the end of the initial mixing stage and the calculation formula is as follows: Auxiliary long-time evaluation effect is the calculation result between the predicted disinfectant monitoring data of the reaction stabilization stage and the required disinfectant monitoring data of the reaction stabilization stage, and the calculation result between the predicted disinfectant monitoring data of the homogeneous maturation stage and the required disinfectant monitoring data of the homogeneous maturation stage, k is 2;

[0093] When the current mixing phase is the reaction stabilization phase, the auxiliary long-time evaluation effect When the current mixing phase is the reaction stabilization phase, the auxiliary long-time evaluation effect When the current mixing phase is the reaction stabilization phase, the auxiliary long-time evaluation effect When the current mixing phase is the reaction stabilization phase, the auxiliary long-time evaluation effect

[0094] If the disinfectant mixing prediction effect is less than the set prediction effect evaluation value (different prediction effect evaluation values exist in different cases), no device parameter adjustment is performed, and the Q value is increased based on the increase value stored in the database, and the Q value is updated; if the disinfectant mixing prediction effect is not less than the set prediction effect evaluation value, device parameter adjustment is performed, and the Q value is reduced based on the reduction value stored in the database, and the Q value is updated.

[0095] The animal disinfectant production line monitoring method and device based on artificial intelligence, as shown in Figure 2 The animal disinfectant production line monitoring method and device based on artificial intelligence, as shown in The animal disinfectant production line monitoring method and device based on artificial intelligence, as shown in

[0096] An electronic device, comprising: a processor; a computer readable storage medium; and a memory having stored therein computer program instructions, which when executed by the processor, cause the processor to perform the artificial intelligence based veterinary disinfectant production line monitoring method as described above; a computer readable storage medium for storing a program, which when executed by a processor implements the artificial intelligence based veterinary disinfectant production line monitoring method as described above.

[0097] Those skilled in the art will appreciate that embodiments of the application can be supplied as a method, a system, or a computer program product. Accordingly, the application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code thereon.

[0098] The application is described in reference to the flowchart and / or block diagrams of the system, apparatus (system), and computer program product according to embodiments of the application. It will be understood that each block of the flowchart and / or block diagrams, and combinations of blocks in the flowchart and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processing device, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for performing the function specified by the flowchart and / or block diagram block or blocks.

[0099] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for performing the function specified by the flowchart and / or block diagram block or blocks.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for performing the function specified by the flowchart and / or block diagram block or blocks.

[0101] While the preferred embodiments of the application have been described, additional variations and modifications can be made to these embodiments by those skilled in the art once they have the benefit of the present disclosure without departing from the spirit and scope of the application. Accordingly, it is intended that the appended claims include all such modifications and variations as fall within the scope of the present application.

[0102] It is apparent that those skilled in the art can make various changes and modifications to the application without departing from the spirit and scope of the application. It is therefore intended that the present application cover all such changes and modifications that are within its scope.

Claims

1. A veterinary disinfectant production line monitoring method based on artificial intelligence, characterized in that: The following steps are involved: Obtain the expected compliance data of veterinary disinfectants, screen the effective raw materials of veterinary disinfectants based on the expected compliance data, and determine the effective veterinary disinfectant raw materials. The expected compliance data of veterinary disinfectants include the livestock and poultry skin and mucous membrane irritation index, the risk value of drug-resistant bacteria gene induction, the risk value of raw material mutagenicity, the compatibility index with veterinary vaccines, the degradation rate of organic impurities, the targeted bactericidal efficacy and the thermal stability coefficient; After determining the effective veterinary disinfectant raw materials, random forest decision-making is used to process the expected compliance data of veterinary disinfectants to determine the effective veterinary disinfectant raw material ratio; Based on the determined effective veterinary disinfectant raw material ratio, the effective veterinary disinfectant raw materials are mixed, and disinfectant monitoring data of each mixing stage is obtained during the mixing process. The mixing stage includes an initial mixing stage, a reaction stabilization stage, and a homogenization and maturation stage; Determine whether the disinfectant mixing effect at each mixing stage meets the standards based on disinfectant monitoring data: If the standard is met, continue mixing until the mixing is completed; If the standards are not met, the initial equipment parameters of the mixing equipment will be adjusted based on the Q value strategy. The initial equipment parameters include the initial equipment parameters of the initial mixing stage, the initial equipment parameters of the reaction stabilization stage, and the initial equipment parameters of the homogenization and maturation stage.

2. The artificial intelligence-based veterinary disinfectant production line monitoring method according to claim 1, characterized in that: The effective raw material screening of veterinary disinfectant raw materials is performed based on the expected standard-reaching data of veterinary disinfectants to determine the effective veterinary disinfectant raw materials, including the following steps: Obtain the historical characteristic spectrum matrix of various effective veterinary disinfectant raw materials for each historical batch, and at the same time obtain the historical actual compliance data of veterinary disinfectants corresponding to the effective veterinary disinfectant raw materials of the same batch, the historical ratio of raw materials, the historical parameters of equipment in the initial mixing stage, the historical parameters of equipment in the reaction stabilization stage, the historical parameters of equipment in the homogenization and maturation stage, the historical monitoring data of disinfectants in the initial mixing stage, the historical monitoring data of disinfectants in the reaction stabilization stage, and the historical monitoring data of disinfectants in the homogenization and maturation stage; Based on the blockchain, the historical characteristic spectrum matrix and the corresponding historical actual compliance data of veterinary disinfectants, historical raw material ratios, historical equipment parameters of the initial mixing stage, historical equipment parameters of the reaction stabilization stage, historical equipment parameters of the homogenization and maturation stage, historical monitoring data of disinfectants in the initial mixing stage, historical monitoring data of disinfectants in the reaction stabilization stage, and historical monitoring data of disinfectants in the homogenization and maturation stage are stored; Perform similarity analysis on the expected compliance data of veterinary disinfectants and the historical actual compliance data of various veterinary disinfectants stored in the blockchain, determine the actual compliance data of veterinary disinfectants that is most similar to the expected compliance data of veterinary disinfectants, and then determine the corresponding historical characteristic spectrum matrix of various effective veterinary disinfectant raw materials; Obtain characteristic spectrum matrices of various veterinary disinfectant raw materials, perform cosine similarity analysis on the characteristic spectrum matrices of various veterinary disinfectant raw materials and the corresponding historical characteristic spectrum matrices of effective veterinary disinfectant raw materials, and obtain cosine similarity values ​​corresponding to various veterinary disinfectant raw materials; If the cosine similarity value is greater than the set similarity threshold, the corresponding veterinary disinfectant raw material is an effective veterinary disinfectant raw material; If the cosine similarity value is not greater than the set similarity threshold, the corresponding veterinary disinfectant raw material is not a valid veterinary disinfectant raw material and is replaced.

3. The artificial intelligence-based veterinary disinfectant production line monitoring method according to claim 2, characterized in that: Determining the effective raw material ratio of veterinary disinfectant includes the following steps: The actual compliance data of each batch of veterinary disinfectants in history are fused and processed to obtain the historical disinfection efficacy benchmark value, the historical disinfection efficacy benchmark value S ： Among them, F1 and F2 are both transition functions, max(F1+F2) is the maximum function value of transition functions F1 and F2, LPSI is the normalized livestock and poultry skin and mucosal irritation index, DRG is the normalized risk value of drug-resistant bacteria gene induction, RMM is the normalized raw material mutagenicity risk value, CIV is the normalized compatibility index with veterinary vaccines, OID is the normalized organic impurity degradation rate, oid is the normalized organic impurity degradation rate reference value, TBE is the normalized targeted bactericidal efficacy, TSC is the normalized thermal stability coefficient, α1, α2, α3, α4, β1, β2 and β3 are all weight coefficients; Obtain the historical ratio of effective veterinary disinfectant raw materials for each batch of veterinary disinfectants, use the historical ratio of effective veterinary disinfectant raw materials and the corresponding disinfection efficacy benchmark value as the input of the random forest model, and output the benchmark value-ratio formula: S=ω1*a1+ω2*a2+...+ω n *a n ; Among them, ω1, ω2, …, ω n are all weight coefficients, a1, a2, ..., a n The closed ratio is the ratio of each effective veterinary disinfectant raw material, that is, the ratio of the mass of each effective veterinary disinfectant raw material to the total mass; The expected compliance data of veterinary disinfectants are integrated and processed to obtain the expected disinfection efficacy benchmark value; Based on the benchmark value-ratio formula, the effective veterinary disinfectant raw material ratio solution set corresponding to the expected disinfection efficacy benchmark value is determined. If there are multiple solutions in the effective veterinary disinfectant raw material ratio solution set, they are screened to determine the optimal solution as the effective veterinary disinfectant raw material ratio. If there is only one solution, it is directly output as the effective veterinary disinfectant raw material ratio.

4. The method for monitoring a veterinary disinfectant production line based on artificial intelligence according to claim 3, characterized in that: Determining the optimal solution includes the following steps: Determine the pending ratio of effective veterinary disinfectant raw materials corresponding to each solution, perform similarity analysis on the pending ratio of effective veterinary disinfectant raw materials and the historical ratio of raw materials, determine the historical ratio of raw materials corresponding to each solution, and then determine the corresponding historical actual compliance data of veterinary disinfectants; Calculate the sum C of the raw material costs corresponding to each solution; Calculate the cosine similarity value B between the historical actual compliance data of veterinary disinfectants and the expected compliance data of veterinary disinfectants corresponding to each solution; Compute the evaluated value of the solution: Among them, sC is the reference value of the sum of raw material costs, e is a natural constant, and λ1 and λ2 are weight factors; The solution corresponding to the largest evaluation value is regarded as the optimal solution.

5. The method for monitoring a veterinary disinfectant production line based on artificial intelligence according to claim 2, characterized in that: Determining whether the disinfectant mixing effect at each mixing stage meets the standards based on the disinfectant monitoring data includes the following steps: Perform K-means cluster analysis on the historical actual compliance data of veterinary disinfectants to determine multiple compliance data cluster centers. Based on the blockchain, obtain the historical monitoring data of disinfectants in the initial mixing stage, the historical monitoring data of disinfectants in the reaction stabilization stage, and the historical monitoring data of disinfectants in the homogenization and maturation stage corresponding to the historical actual compliance data of veterinary disinfectants in each compliance data cluster center; The historical actual compliance data of veterinary disinfectants corresponding to each compliance data cluster center are averaged and normalized to obtain the historical actual compliance feature data of veterinary disinfectants; The historical monitoring data of the disinfectant in the initial mixing stage, the historical monitoring data of the disinfectant in the reaction stabilization stage, and the historical monitoring data of the disinfectant in the homogenization and maturation stage corresponding to each cluster center of the standard-compliant data are averaged and normalized to obtain historical monitoring feature data, including the historical monitoring feature data of the disinfectant in the initial mixing stage, the historical monitoring feature data of the disinfectant in the reaction stabilization stage, and the historical monitoring feature data of the disinfectant in the homogenization and maturation stage; After normalizing the expected compliance data of veterinary disinfectants, perform similarity comparison with the historical actual compliance feature data of veterinary disinfectants corresponding to each compliance data cluster center, and determine the compliance data cluster center corresponding to the most similar historical actual compliance feature data of veterinary disinfectants; Obtain and normalize the disinfectant monitoring data of the mixing stage of the effective veterinary disinfectant raw materials, and perform cosine similarity analysis with the historical monitoring feature data of the corresponding mixing stage to obtain the monitoring data similarity value; If the similarity value of the monitoring data is greater than the set monitoring similarity threshold, the disinfectant mixing effect in the corresponding mixing stage meets the standard; If the similarity value of the monitoring data is not greater than the set monitoring similarity threshold, the disinfectant mixing effect in the corresponding mixing stage does not meet the standard.

6. The artificial intelligence-based veterinary disinfectant production line monitoring method according to claim 5, characterized in that: The process of obtaining initial device parameters is as follows: Based on the standard data cluster center corresponding to the most similar historical actual standard-compliant feature data of veterinary disinfectants, the corresponding historical parameters of the equipment in the initial mixing stage, the historical parameters of the equipment in the reaction stabilization stage, and the historical parameters of the equipment in the homogenization and maturation stage are obtained; Collect statistics on the historical parameters of the equipment in the initial mixing stage, the historical parameters of the equipment in the reaction stabilization stage, and the historical parameters of the equipment in the homogenization and maturation stage, and determine the historical parameter intervals of the equipment in the initial mixing stage, the historical parameter intervals of the equipment in the reaction stabilization stage, and the historical parameter intervals of the equipment in the homogenization and maturation stage; Perform K-means cluster analysis on the historical parameters of the equipment in the initial mixing stage to determine the cluster center of the equipment parameters in the initial mixing stage corresponding to the historical parameters of the equipment in the initial mixing stage, and determine the cluster center of the equipment parameters in the initial mixing stage corresponding to the most samples; Perform K-means cluster analysis on the historical parameters of the equipment in the reaction stable stage to determine the cluster center of the equipment parameters in the reaction stable stage corresponding to the historical parameters of the equipment in the reaction stable stage, and determine the cluster center of the equipment parameters in the reaction stable stage corresponding to the most samples; Perform K-means cluster analysis on the historical parameters of the equipment in the homogenization and maturation stage to determine the cluster center of the equipment parameters in the homogenization and maturation stage corresponding to the historical parameters of the equipment in the homogenization and maturation stage, and determine the cluster center of the equipment parameters in the homogenization and maturation stage corresponding to the most samples; The samples corresponding to the cluster center of the equipment parameters in the initial mixing stage corresponding to the most samples, the cluster center of the equipment parameters in the reaction stabilization stage corresponding to the most samples, and the cluster center of the equipment parameters in the homogenization and maturation stage corresponding to the most samples are respectively processed by means of mean value to obtain the initial parameters of the equipment in the initial mixing stage, the initial parameters of the equipment in the reaction stabilization stage, and the initial parameters of the equipment in the homogenization and maturation stage.

7. The artificial intelligence-based veterinary disinfectant production line monitoring method according to claim 6, characterized in that: Adjusting the device parameters of a hybrid device based on the Q value strategy includes the following steps: Obtain the disinfectant monitoring data of the current mixing stage, which is defined as the state space; Obtain an action space set stored in a database, where each action in the action space set corresponds to a device parameter adjustment amount for a set of hybrid devices, and each action also corresponds to a stored Q value; Determine the action corresponding to the maximum Q value, and adjust the initial device parameters of the current mixing stage based on the device parameter adjustment amount corresponding to the action to obtain the adjusted device parameters; Obtain the device parameter limit interval obtained from the database, and verify the adjusted device parameters based on the device parameter limit interval, the adjusted device parameters, and the device parameter adjustment amount: If the verification passes, the hybrid device is adjusted based on the adjusted device parameters; If the verification fails, the Q value is reduced based on the reduction value set in the database, and the action corresponding to the maximum Q value is re-determined, and the adjusted device parameters are verified again; After adjusting the mixing device based on the adjusted device parameters, the disinfectant monitoring data after the set time is obtained and the state space is updated; Based on the updated disinfectant monitoring data, the disinfectant mixing effect is evaluated again to determine whether the equipment parameters need to be adjusted again and the Q value is updated.

8. The method for monitoring a veterinary disinfectant production line based on artificial intelligence according to claim 7, characterized in that: Based on the updated disinfectant monitoring data, the disinfectant mixing effect is evaluated again to determine whether the equipment parameters need to be adjusted again and the Q value is updated, including the following steps: Obtain disinfectant monitoring data after a set time after equipment adjustment, and obtain short-term evaluation effects based on the disinfectant monitoring data after the set time; The disinfectant monitoring data after the set time and the current adjusted equipment parameters are input into the trained LSTM model to obtain the disinfectant monitoring data after the current mixing stage. The long-term evaluation effect is obtained based on the disinfectant monitoring data after the current mixing stage. The disinfectant monitoring data after the current mixing stage and the initial equipment parameters of the next mixing stage are input into the trained LSTM model to obtain the disinfectant monitoring data after the next mixing stage. Based on the disinfectant monitoring data after the next mixing stage, the auxiliary long-term evaluation effect is obtained; Based on short-term evaluation effect Long-term evaluation effect and assist in long-term evaluation of the effect Get the predicted effect of disinfectant mixing Ψ ： Among them, k is the auxiliary long-term evaluation effect The number of , e is a natural constant; If the predicted effect of the disinfectant mixture is less than the set predicted effect evaluation value, the equipment parameters are not adjusted, and the Q value is increased based on the increase value stored in the database to update the Q value; If the predicted effect of the disinfectant mixture is not less than the set predicted effect evaluation value, the equipment parameters are adjusted, and the Q value is reduced based on the reduction value stored in the database to update the Q value.

9. A method and device for monitoring a veterinary disinfectant production line based on artificial intelligence, characterized in that: It includes an effective raw material determination module, a ratio determination module and a mixing adjustment module, wherein: An effective raw material determination module is used to obtain expected compliance data for veterinary disinfectants, screen the effective raw materials for veterinary disinfectants based on the expected compliance data, and determine the effective veterinary disinfectant raw materials. The expected compliance data for veterinary disinfectants include the livestock and poultry skin and mucous membrane irritation index, the risk value of drug-resistant bacteria gene induction, the raw material mutagenicity risk value, the compatibility index with veterinary vaccines, the degradation rate of organic impurities, the targeted bactericidal efficacy, and the thermal stability coefficient; A ratio determination module is used to determine the effective veterinary disinfectant raw materials, and then use random forest decision-making to process the expected compliance data of veterinary disinfectants to determine the effective veterinary disinfectant raw material ratio; A mixing adjustment module is used to mix the effective veterinary disinfectant raw materials based on the determined effective veterinary disinfectant raw material ratio, and obtain disinfectant monitoring data at each mixing stage during the mixing process. The mixing stages include the initial mixing stage, the reaction stabilization stage, and the homogenization and maturation stage; Determine whether the disinfectant mixing effect at each mixing stage meets the standards based on disinfectant monitoring data: If the standard is met, continue mixing until the mixing is completed; If the standards are not met, the initial equipment parameters of the mixing equipment will be adjusted based on the Q value strategy. The initial equipment parameters include the initial equipment parameters of the initial mixing stage, the initial equipment parameters of the reaction stabilization stage, and the initial equipment parameters of the homogenization and maturation stage.

10. Electronic equipment, including: processor; computer-readable storage medium; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the veterinary disinfectant production line monitoring method based on artificial intelligence according to any one of claims 1 to 8; A computer-readable storage medium for storing a program, which, when executed by a processor, implements the artificial intelligence-based veterinary disinfectant production line monitoring method as described in any one of claims 1 to 8.

Images