Root-fertilizer mutual-feeding synergistic fertilizer and preparation method thereof

Abstract

The present application discloses a method for preparing a root-fertilizer mutual-feedback synergistic fertilizer, which relates to the field of intelligent fertilizer production. The method uses a data analysis technology based on deep learning to perform a temporal correlation analysis of a fermentation temperature value and a real-time oxygen concentration value, thereby automatically judging whether to turn on the ventilation equipment according to the interactive response representation characteristics between the fermentation temperature temporal correlation characteristics and the oxygen concentration temporal correlation characteristics. In this way, the temperature and oxygen concentration changes during the fermentation process can be quickly responded to, and the accuracy and response speed of the ventilation equipment control can be improved based on the temporal interactive response and synergy between the actual fermentation temperature and oxygen concentration, thereby reducing the need for manual monitoring and realizing intelligent control of the fermentation process.

Description

Technical Field

[0001] The present application relates to the field of intelligent fertilizer production, and more specifically, to a root-fertilizer mutual-feedback synergistic fertilizer and a preparation method thereof. Background Art

[0002] Root-fertilizer mutual-feedback fertilizer is a new type of fertilizer that optimizes fertilizer formula and production process to improve the interaction between plant roots and fertilizer, thereby enhancing the plant's ability to absorb nutrients and overall growth effect. The design concept of this fertilizer is to improve the soil environment, promote the activity of beneficial microorganisms, and enhance the function of plant roots, so as to improve fertilizer utilization, enhance crop growth potential, and increase crop yield and quality.

[0003] In the preparation process of root fertilizer mutual feeding synergistic fertilizer, fermentation treatment is required to promote the activity and metabolic efficiency of microorganisms, thereby improving the decomposition of organic matter in fertilizers. During the fermentation process, multiple parameters will affect the activity and metabolic efficiency of microorganisms, among which temperature and oxygen concentration are two main factors. Temperature has a significant effect on the growth rate and metabolic activity of microorganisms. Different microorganisms have different optimal growth temperatures. Too high or too low temperatures may inhibit the growth of microorganisms and even cause the death of microorganisms. Within the appropriate temperature range, the metabolic activity of microorganisms will accelerate, thereby accelerating the decomposition of organic matter and the release of nutrients. Oxygen concentration is also crucial to the fermentation process. Aerobic microorganisms require sufficient oxygen to support their metabolic activities, while anaerobic microorganisms grow and metabolize under anaerobic conditions.

[0004] Ventilation control in traditional fermentation processes mainly relies on manual experience and feeling, and this method has some obvious defects. First, it is difficult to achieve precise control through manual adjustment, and the temperature is prone to fluctuate during the fermentation process, which will cause the metabolic activity of microorganisms to be unstable. If this small temperature change cannot be accurately controlled, it may disrupt the normal metabolic rhythm of microorganisms. Secondly, manual control often has a slow response speed and is difficult to respond to rapid changes in temperature in a timely manner, which may lead to the accumulation or loss of heat energy during the fermentation process, thereby affecting the growth and metabolism of microorganisms. In addition, traditional manual control and adjustment methods often ignore the interaction between oxygen concentration and temperature. In fact, both temperature and oxygen concentration affect the activity of microorganisms, and the synergistic effect between them is equally important for the growth and metabolism of microorganisms. Traditional ventilation control methods can often only consider these factors separately, without considering their mutual influence.

[0005] Therefore, a method for preparing a root-fertilizer mutual-feed synergistic fertilizer is desired. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present application provides a root-fertilizer mutual-feed synergistic fertilizer and a preparation method thereof.

[0007] According to one aspect of the present application, a method for preparing a root-fertilizer mutual-feed synergistic fertilizer is provided, comprising:

[0008] Mixing organic raw materials and adjusting the moisture content to 40%-60% to obtain adjusted mixed organic matter, wherein the organic raw materials include animal manure, plant residues and compost;

[0009] Adding beneficial microorganisms to the adjusted mixed organic matter and stirring the mixture to obtain a mixed organic matter containing microorganisms, wherein the beneficial microorganisms include nitrogen-fixing bacteria, phosphorus bacteria and silicate bacteria;

[0010] Placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation treatment to obtain fermented organic matter;

[0011] Screening the fermented organic matter to obtain screened fertilizer;

[0012] Drying and packaging the screened fertilizer to obtain a root-fertilizer mutual-feed synergistic fertilizer;

[0013] The mixed organic matter containing microorganisms is placed in a fermentation tank for fermentation treatment to obtain fermented organic matter, including: obtaining a time queue of fermentation temperature values ​​collected by a temperature sensor; obtaining a time queue of real-time oxygen concentration values ​​collected by an oxygen concentration sensor; inputting the time queue of fermentation temperature values ​​into a fermentation temperature sequence encoder based on a backward LSTM model to obtain a fermentation temperature time series associated feature vector; inputting the time queue of real-time oxygen concentration values ​​into an oxygen concentration sequence encoder based on a backward LSTM model to obtain an oxygen concentration time series associated feature vector; performing oxygen concentration-fermentation temperature time series interactive response processing on the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector to obtain an oxygen concentration-fermentation temperature interactive response representation feature; obtaining a control instruction based on the oxygen concentration-fermentation temperature interactive response representation feature, the control instruction being used to indicate whether to turn on the ventilation equipment; The characteristic vector and the oxygen concentration time series associated characteristic vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain oxygen concentration-fermentation temperature interactive response representation characteristics, including: characteristic decoupling of the oxygen concentration time series associated characteristic vector to obtain a sequence of oxygen concentration local time series associated characteristic vectors; calculating the characteristic interaction energy factor of each oxygen concentration local time series associated characteristic vector in the sequence of the fermentation temperature time series associated characteristic vector and the oxygen concentration local time series associated characteristic vector to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors; based on the sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, local-global interactive response processing is performed on each oxygen concentration local time series associated characteristic vector in the sequence of the fermentation temperature time series associated characteristic vector and the oxygen concentration local time series associated characteristic vector to obtain an oxygen concentration-fermentation temperature local-global interactive response representation matrix as the oxygen concentration-fermentation temperature interactive response representation characteristics.

[0014] In the preparation method of the above-mentioned root fertilizer mutual feed synergistic fertilizer, the characteristic interaction energy factor of each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector is calculated to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: performing a nonlinear transformation on the fermentation temperature time series association feature vector to obtain a fermentation temperature time series association feature vector after the nonlinear transformation; calculating the correlation matrix between the fermentation temperature time series association feature vector after the nonlinear transformation and each oxygen concentration local time series association feature vector in the sequence of the oxygen concentration local time series association feature vector to obtain a sequence of oxygen concentration-fermentation temperature interaction correlation matrices; calculating the characteristic interaction energy factor of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors.

[0015] In the preparation method of the above-mentioned root-fertilizer mutual feed synergistic fertilizer, the fermentation temperature time series correlation characteristic vector is nonlinearly transformed to obtain the fermentation temperature time series correlation characteristic vector after nonlinear transformation, including: multiplying the fermentation temperature time series correlation characteristic vector with the fermentation temperature transformation matrix, and then adding the obtained fermentation temperature time series correlation weight vector and fermentation temperature bias vector by position to obtain the fermentation temperature time series correlation characteristic vector after nonlinear transformation.

[0016] In the preparation method of the above-mentioned root-fertilizer mutual feed synergistic fertilizer, the correlation matrix between the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector is calculated to obtain a sequence of oxygen concentration-fermentation temperature interactive correlation matrices, including: vector multiplication of the transposed vector of the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector to obtain the sequence of the oxygen concentration-fermentation temperature interactive correlation matrix.

[0017] In the preparation method of the above-mentioned root fertilizer mutual feed synergistic fertilizer, the characteristic interaction energy factors of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix are calculated to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: respectively calculating the mean and variance of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain the oxygen concentration-fermentation temperature interaction correlation mean and the oxygen concentration-fermentation temperature interaction correlation variance; calculating the positional difference between the oxygen concentration-fermentation temperature interaction correlation matrix and the oxygen concentration-fermentation temperature interaction correlation mean to obtain the oxygen concentration-fermentation temperature interaction correlation difference matrix; calculating the fourth power of each eigenvalue in the oxygen concentration-fermentation temperature interaction correlation difference matrix to obtain the oxygen concentration-fermentation temperature interaction correlation modulation difference matrix; calculating the expected value of the oxygen concentration-fermentation temperature interaction correlation modulation difference matrix to obtain the oxygen concentration-fermentation temperature interaction correlation expected value; and dividing the oxygen concentration-fermentation temperature interaction correlation period by the square root of ... The expected value is divided by the square of the oxygen concentration-fermentation temperature interaction variance to obtain the oxygen concentration-fermentation temperature interaction kurtosis value; the sum of the oxygen concentration-fermentation temperature interaction variance and the hyperparameter is calculated to obtain the oxygen concentration-fermentation temperature interaction first energy factor; the square of the difference between the oxygen concentration-fermentation temperature interaction kurtosis value and the oxygen concentration-fermentation temperature interaction mean is calculated to obtain the oxygen concentration-fermentation temperature interaction difference value; the value obtained by multiplying the oxygen concentration-fermentation temperature interaction variance by constant two and the value obtained by multiplying the hyperparameter by constant two are added to the oxygen concentration-fermentation temperature interaction difference value to obtain the oxygen concentration-fermentation temperature interaction second energy factor; the oxygen concentration-fermentation temperature interaction first energy factor is divided by the oxygen concentration-fermentation temperature interaction second energy factor to obtain the oxygen concentration-fermentation temperature characteristic interaction energy factor corresponding to the oxygen concentration-fermentation temperature interaction matrix.

[0018] In the preparation method of the above-mentioned root fertilizer mutual feedback synergistic fertilizer, based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors, local-global interaction response processing is performed on each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix, including: inputting the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors into the Sigmoid function to obtain a sequence of normalized oxygen concentration-fermentation temperature characteristic interaction energy factors; using the sequence of the normalized oxygen concentration-fermentation temperature characteristic interaction energy factors as a sequence of weights, calculating the position-weighted sum of the sequence of the oxygen concentration-fermentation temperature interaction association matrix to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix.

[0019] In the above-mentioned method for preparing the root-fertilizer mutual feed synergistic fertilizer, a control instruction is obtained based on the oxygen concentration-fermentation temperature local-global interaction response representation matrix, including: inputting the oxygen concentration-fermentation temperature local-global interaction response representation matrix into a classifier-based ventilation equipment control module to obtain the control instruction.

[0020] In the preparation method of the above-mentioned root-fertilizer mutual feed synergistic fertilizer, the oxygen concentration-fermentation temperature local-global interaction response representation matrix is ​​input into a ventilation equipment control module based on a classifier to obtain the control instruction, including: expanding the oxygen concentration-fermentation temperature local-global interaction response representation matrix into an oxygen concentration-fermentation temperature local-global interaction response representation vector according to a row vector or a column vector; using the fully connected layer of the classifier to fully connect encode the oxygen concentration-fermentation temperature local-global interaction response representation vector to obtain a fully connected encoding feature vector; inputting the fully connected encoding feature vector into the Softmax classification function of the classifier to obtain the probability value of the oxygen concentration-fermentation temperature local-global interaction response representation matrix belonging to each classification label, and the classification label includes a label for indicating that the ventilation equipment is turned on and a label for indicating that the ventilation equipment is not turned on; and determining the classification label corresponding to the largest of the probability values ​​as the control instruction.

[0021] According to another aspect of the present application, a root-fertilizer mutual-feeding synergistic fertilizer is provided, and the root-fertilizer mutual-feeding synergistic fertilizer is prepared by any of the above-mentioned preparation methods of the root-fertilizer mutual-feeding synergistic fertilizer.

[0022] The root-fertilizer mutual-feed synergistic fertilizer and its preparation method provided by the present application use data analysis technology based on deep learning to perform time-series correlation analysis of fermentation temperature values ​​and real-time oxygen concentration values, so as to automatically judge whether to turn on the ventilation equipment according to the interactive response representation characteristics between the fermentation temperature time-series correlation characteristics and the oxygen concentration time-series correlation characteristics. In this way, it is possible to quickly respond to changes in temperature and oxygen concentration during the fermentation process, and improve the accuracy and response speed of the control of the ventilation equipment based on the time-series interactive response and synergy between the actual fermentation temperature and oxygen concentration, reduce the need for manual monitoring, and realize intelligent control of the fermentation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0024] Figure 1 The present invention is a flow chart of a method for preparing a root-fertilizer mutual-feed synergistic fertilizer according to an embodiment of the present application.

[0025] Figure 2 This is a flow chart of placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation treatment to obtain fermented organic matter in the preparation method of the root-fertilizer mutual feeding synergistic fertilizer according to an embodiment of the present application.

[0026] Figure 3 This is a data flow diagram of the method for preparing the root-fertilizer mutual-feed synergistic fertilizer according to an embodiment of the present application, in which the mixed organic matter containing microorganisms is placed in a fermentation tank for fermentation treatment to obtain the fermented organic matter.

[0027] Figure 4 In the method for preparing the root-fertilizer mutual feed synergistic fertilizer according to the embodiment of the present application, the fermentation temperature time series correlation feature vector and the oxygen concentration time series correlation feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain a flow chart representing the oxygen concentration-fermentation temperature interactive response characteristics.

[0028] Figure 5 In the method for preparing the root-fertilizer mutual feed synergistic fertilizer according to the embodiment of the present application, the characteristic interaction energy factor of each oxygen concentration local time series correlation feature vector in the sequence of the fermentation temperature time series correlation feature vector and the oxygen concentration local time series correlation feature vector is calculated to obtain a flow chart of the sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors.

[0029] Figure 6 In the preparation method of the root-fertilizer mutual feed synergistic fertilizer according to the embodiment of the present application, based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors, each oxygen concentration local time series correlation feature vector in the sequence of the fermentation temperature time series correlation feature vector and the oxygen concentration local time series correlation feature vector is subjected to local-global interactive response processing to obtain a flow chart of the oxygen concentration-fermentation temperature local-global interactive response representation matrix. DETAILED DESCRIPTION

[0030] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0031] Root-fertilizer mutual feedback fertilizer is a new type of fertilizer that aims to promote the interaction between plant roots and fertilizers by optimizing fertilizer formula and production process, thereby improving plant nutrient absorption efficiency and enhancing crop growth and yield.

[0032] In the preparation process of root-fertilizer mutual-feed synergistic fertilizer, the fermentation step is crucial, as it can activate microorganisms and improve the efficiency of organic matter decomposition. During the fermentation process, temperature and oxygen concentration are key parameters that directly affect the activity and metabolism of microorganisms. Suitable temperature can accelerate microbial metabolism and promote nutrient release; while oxygen concentration determines the growth conditions of aerobic or anaerobic microorganisms.

[0033] However, traditional fermentation ventilation control relies on manual experience, and has problems such as inaccurate control, slow response, and neglect of the interaction between temperature and oxygen. These problems may lead to unstable microbial metabolism and affect the fertilizer effect.

[0034] Based on this, the present application proposes a method for preparing a root-fertilizer mutual-feed synergistic fertilizer. Figure 1 FIG. 1 is a flow chart of a method for preparing a root-fertilizer mutual-feed synergistic fertilizer according to an embodiment of the present application. Figure 1 As shown, according to the preparation method of the root-fertilizer mutual-feed synergistic fertilizer of the embodiment of the present application, the method includes: S110, mixing the organic raw materials and adjusting the moisture content to 40%-60% to obtain an adjusted mixed organic matter, wherein the organic raw materials include animal manure, plant residues and compost; S120, adding beneficial microorganisms to the adjusted mixed organic matter, and stirring evenly to obtain a mixed organic matter containing microorganisms, wherein the beneficial microorganisms include nitrogen-fixing bacteria, phosphorus bacteria and silicate bacteria; S130, placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation to obtain fermented organic matter; S140, screening the fermented organic matter to obtain screened fertilizer; S150, drying and packaging the screened fertilizer to obtain the root-fertilizer mutual-feed synergistic fertilizer.

[0035] Accordingly, in the above-mentioned process of placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation treatment, the technical concept of the present application is to collect the time queue of the fermentation temperature value by the temperature sensor, and collect the time queue of the real-time oxygen concentration value by the oxygen concentration sensor, and use the data processing and analysis technology based on deep learning to perform the time series correlation analysis of the fermentation temperature value and the real-time oxygen concentration value, so as to automatically judge whether to turn on the ventilation equipment according to the interactive response representation characteristics between the fermentation temperature time series correlation characteristics and the oxygen concentration time series correlation characteristics. In this way, it is possible to quickly respond to the changes in temperature and oxygen concentration during the fermentation process, and improve the accuracy and response speed of the control of the ventilation equipment based on the time series interactive response and synergy between the actual fermentation temperature and oxygen concentration, reduce the need for manual monitoring, and realize intelligent control of the fermentation process.

[0036] Figure 2This is a flow chart of placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation treatment to obtain fermented organic matter in the preparation method of the root-fertilizer mutual feeding synergistic fertilizer according to an embodiment of the present application. Figure 3 This is a data flow diagram of the method for preparing the root-fertilizer mutual-feed synergistic fertilizer according to the embodiment of the present application, in which the mixed organic matter containing microorganisms is placed in a fermentation tank for fermentation treatment to obtain the fermented organic matter. Figure 2 and Figure 3 As shown, the mixed organic matter containing microorganisms is placed in a fermentation tank for fermentation treatment to obtain fermented organic matter, including: S131, obtaining a time queue of fermentation temperature values ​​collected by a temperature sensor; S132, obtaining a time queue of real-time oxygen concentration values ​​collected by an oxygen concentration sensor; S133, inputting the time queue of fermentation temperature values ​​into a fermentation temperature sequence encoder based on a backward LSTM model to obtain a fermentation temperature time series associated feature vector; S134, inputting the time queue of real-time oxygen concentration values ​​into an oxygen concentration sequence encoder based on a backward LSTM model to obtain an oxygen concentration time series associated feature vector; S135, performing oxygen concentration-fermentation temperature time series interactive response processing on the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector to obtain an oxygen concentration-fermentation temperature interactive response representation feature; S136, obtaining a control instruction based on the oxygen concentration-fermentation temperature interactive response representation feature, and the control instruction is used to indicate whether to turn on the ventilation equipment.

[0037] In step S131 and step S132, the time queue of the fermentation temperature value collected by the temperature sensor and the time queue of the real-time oxygen concentration value collected by the oxygen concentration sensor are respectively obtained. It should be understood that organic fermentation is a dynamic and complex biochemical reaction process. During the fermentation process, the activity of microorganisms is affected by temperature and oxygen concentration, and different microorganisms show different growth rates and metabolic efficiencies under different temperature and oxygen conditions. Considering that the temperature and oxygen concentration will change over time during the fermentation process, by collecting the time queue data of the fermentation temperature value and the time queue data of the oxygen concentration value, it can help to understand the state of the mixed organic fermentation environment in real time, and by analyzing the changing trend of the time queue data, it can be judged whether it is necessary to turn on the ventilation equipment to adjust the environmental conditions. That is, the time queue data of the fermentation temperature and oxygen concentration is the basis for realizing precise ventilation control, which provides a scientific basis for the control of the ventilation equipment, thereby optimizing the fermentation process of the root fertilizer mutual feedback synergistic fertilizer.

[0038] In step S133, the time queue of the fermentation temperature value is input into the fermentation temperature sequence encoder based on the backward LSTM model to obtain the fermentation temperature time series associated feature vector. Accordingly, considering that the change of the fermentation temperature during the fermentation process is a dynamic process that develops over time, and there are mutual correlations and dependencies between the various fermentation temperatures and the data at the current time point is affected by the future time, based on this, in the technical solution of the present application, the time queue of the fermentation temperature value is input into the fermentation temperature sequence encoder of the backward LSTM model to capture and mine the dependencies and complex time series patterns between different time periods, and obtain the fermentation temperature time series associated feature vector. It should be understood that LSTM (Long Short-Term Memory Network) can effectively process and memorize long time series data. It controls the flow of information by introducing a gating mechanism, thereby overcoming the gradient vanishing problem of traditional recurrent neural networks when processing long sequences. As part of the bidirectional LSTM, the backward LSTM can consider the impact of future information on the current time point, that is, when analyzing the fermentation temperature value at the current moment, not only the previous fermentation temperature value is referred to, but also the subsequent fermentation temperature changes are considered, which is conducive to improving the accuracy of the fermentation temperature time series feature extraction.

[0039] In step S134, the time queue of the real-time oxygen concentration value is input into the oxygen concentration sequence encoder based on the backward LSTM model to obtain the oxygen concentration time series correlation feature vector. It should be understood that since the time queue of the real-time oxygen concentration value is also constantly changing with the passage of time, and the recent changes may have a greater impact on the current fermentation state, therefore, in order to better capture and refine the interrelated influence relationship between each time segment, so as to more accurately reflect the current fermentation state, in the technical solution of the present application, the time queue of the real-time oxygen concentration value is input into the oxygen concentration sequence encoder based on the backward LSTM model to obtain the oxygen concentration time series correlation feature vector.

[0040] In step S135, the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain oxygen concentration-fermentation temperature interactive response representation characteristics. Accordingly, it is considered that the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector respectively reflect the trend and pattern of fermentation temperature and oxygen concentration changing over time during the fermentation process. And there is a mutual correlation relationship between the two in the local time series and the global range, that is, if the temperature rises, it may accelerate the metabolism of microorganisms, thereby causing the oxygen consumption rate to accelerate, and the oxygen concentration will decrease accordingly, and in the whole fermentation process, the change trend of temperature and oxygen concentration is usually consistent. In order to deeply understand and carefully analyze the associated response relationship between the two, in the technical solution of the present application, the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain the oxygen concentration-fermentation temperature local-global interactive response representation matrix as the oxygen concentration-fermentation temperature interactive response representation feature. In particular, the oxygen concentration-fermentation temperature time series interactive response processing can effectively extract and fuse the deep information in the feature vector by using the methods of feature decoupling, feature interaction and interactive metric learning. In this way, the complex dynamics of temperature and oxygen concentration changes during the fermentation process can be captured more accurately, thereby providing strong data support for the optimization of the fermentation process.

[0041] Specifically, first, the oxygen concentration time series correlation feature vector is feature decoupled to decompose the features of the entire time domain into a sequence of multiple oxygen concentration local time series correlation feature vectors containing local time segments, providing a basis for subsequent local interactions. Then, the fermentation temperature time series correlation feature vector is nonlinearly transformed to capture and mine more abundant fermentation temperature time series features to obtain the fermentation temperature time series correlation feature vector after nonlinear transformation. Furthermore, by calculating the time series correlation interaction between each oxygen concentration local time series correlation feature vector and the fermentation temperature time series correlation feature vector after nonlinear transformation, a set of fermentation temperature-oxygen concentration interaction correlation matrices is generated. In this way, the local dynamic correlation interaction relationship between temperature change and oxygen concentration can be quantified and captured, thereby providing a data basis for in-depth analysis and understanding of the complex interactions within the system. Subsequently, the characteristic interaction energy factor of each interaction correlation matrix is ​​calculated. In particular, the characteristic interaction energy factor is calculated by extracting the mean, variance and kurtosis from the correlation interaction matrix, which can more deeply reveal the mutual influence between features and thus identify the most influential interaction features. Then, the characteristic interaction energy factors of each interaction matrix are normalized based on the Sigmoid function to limit the values ​​to the range of [0,1] and converted into weights so that the model can focus more on important interaction features to balance the impact of each feature interaction on the model. Finally, each normalized factor is used as a weight to weight and calculate the set of fermentation temperature-oxygen concentration interaction correlation matrices to globally integrate various local time series correlation features and obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix.

[0042] Specifically, Figure 4 In the preparation method of the root-fertilizer mutual-feedback synergistic fertilizer according to the embodiment of the present application, the fermentation temperature time series correlation feature vector and the oxygen concentration time series correlation feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain a flow chart representing the oxygen concentration-fermentation temperature interactive response characteristics. Figure 4As shown, the fermentation temperature time series association feature vector and the oxygen concentration time series association feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain oxygen concentration-fermentation temperature interactive response representation characteristics, including: S1351, feature decoupling of the oxygen concentration time series association feature vector to obtain a sequence of oxygen concentration local time series association feature vectors; S1352, calculating the feature interaction energy factor of each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain a sequence of oxygen concentration-fermentation temperature feature interaction energy factors; S1353, based on the sequence of oxygen concentration-fermentation temperature feature interaction energy factors, local-global interactive response processing is performed on each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain an oxygen concentration-fermentation temperature local-global interactive response representation matrix as the oxygen concentration-fermentation temperature interactive response representation characteristics.

[0043] More specifically, Figure 5 The present invention is a flowchart of calculating the characteristic interaction energy factor of each oxygen concentration local time series correlation feature vector in the sequence of the fermentation temperature time series correlation feature vector and the oxygen concentration local time series correlation feature vector in the preparation method of the root fertilizer mutual feed synergistic fertilizer according to the embodiment of the present application to obtain the sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors. Figure 5 As shown, the characteristic interaction energy factor of each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector is calculated to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: S13521, performing a nonlinear transformation on the fermentation temperature time series association feature vector to obtain a fermentation temperature time series association feature vector after the nonlinear transformation; S13522, calculating the correlation matrix between the fermentation temperature time series association feature vector after the nonlinear transformation and each oxygen concentration local time series association feature vector in the sequence of the oxygen concentration local time series association feature vector to obtain a sequence of oxygen concentration-fermentation temperature interaction correlation matrices; S13523, calculating the characteristic interaction energy factor of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors.

[0044] More specifically, in an embodiment of the present application, the fermentation temperature time series association feature vector is nonlinearly transformed to obtain a fermentation temperature time series association feature vector after nonlinear transformation, including: multiplying the fermentation temperature time series association feature vector with the fermentation temperature transformation matrix, and then adding the obtained fermentation temperature time series association weight vector and the fermentation temperature bias vector by position to obtain the fermentation temperature time series association feature vector after nonlinear transformation.

[0045] More specifically, in an embodiment of the present application, a correlation matrix between the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector is calculated to obtain a sequence of oxygen concentration-fermentation temperature interactive correlation matrices, including: vector multiplying the transposed vector of the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector to obtain a sequence of oxygen concentration-fermentation temperature interactive correlation matrices.

[0046] More specifically, in an embodiment of the present application, the characteristic interaction energy factors of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix are calculated to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: respectively calculating the mean and variance of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain the oxygen concentration-fermentation temperature interaction correlation mean and the oxygen concentration-fermentation temperature interaction correlation variance; calculating the positional difference between the oxygen concentration-fermentation temperature interaction correlation matrix and the oxygen concentration-fermentation temperature interaction correlation mean to obtain an oxygen concentration-fermentation temperature interaction correlation difference matrix; calculating the fourth power of each eigenvalue in the oxygen concentration-fermentation temperature interaction correlation difference matrix to obtain an oxygen concentration-fermentation temperature interaction correlation modulation difference matrix; calculating the expected value of the oxygen concentration-fermentation temperature interaction correlation modulation difference matrix to obtain an oxygen concentration-fermentation temperature interaction correlation expected value; and adding the oxygen concentration-fermentation temperature interaction correlation expected value to the matrix. Divide it by the square of the oxygen concentration-fermentation temperature interaction variance to obtain the oxygen concentration-fermentation temperature interaction kurtosis value; calculate the sum of the oxygen concentration-fermentation temperature interaction variance and the hyperparameter to obtain the oxygen concentration-fermentation temperature interaction first energy factor; calculate the square of the difference between the oxygen concentration-fermentation temperature interaction kurtosis value and the oxygen concentration-fermentation temperature interaction mean to obtain the oxygen concentration-fermentation temperature interaction difference value; add the value obtained by multiplying the oxygen concentration-fermentation temperature interaction variance by constant two and the value obtained by multiplying the hyperparameter by constant two to the oxygen concentration-fermentation temperature interaction difference value to obtain the oxygen concentration-fermentation temperature interaction second energy factor; divide the oxygen concentration-fermentation temperature interaction first energy factor by the oxygen concentration-fermentation temperature interaction second energy factor to obtain the oxygen concentration-fermentation temperature characteristic interaction energy factor corresponding to the oxygen concentration-fermentation temperature interaction matrix.

[0047] More specifically, Figure 6 In the preparation method of the root fertilizer mutual feeding synergistic fertilizer according to the embodiment of the present application, based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factor, each oxygen concentration local time series correlation feature vector in the sequence of the fermentation temperature time series correlation feature vector and the oxygen concentration local time series correlation feature vector is subjected to local-global interaction response processing to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix. Figure 6As shown, based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors, local-global interaction response processing is performed on each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix, including: S13531, inputting the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors into the Sigmoid function to obtain a sequence of normalized oxygen concentration-fermentation temperature characteristic interaction energy factors; S13532, using the sequence of the normalized oxygen concentration-fermentation temperature characteristic interaction energy factors as a sequence of weights, calculating the position-weighted sum of the sequence of the oxygen concentration-fermentation temperature interaction association matrix to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix.

[0048] In the embodiment of the present application, specifically, the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain the oxygen concentration-fermentation temperature interactive response representation feature, which can be expressed by the formula:

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] ;

[0054] ;

[0055] ;

[0056] in, is the oxygen concentration time series correlation feature vector, For feature decoupling operations, are the first, second, ..., first, second, third, fourth, fifth, fifth, sixth, sixth, sixth, seventh, fifth, fifth, sixth, seventh ... ,..., The local time series correlation feature vector of oxygen concentration, is the number of feature vectors in the sequence of the local time series associated feature vectors of the oxygen concentration, is the fermentation temperature time series associated feature vector, and are the fermentation temperature transformation matrix and the fermentation temperature bias vector respectively, is the time series correlation feature vector of fermentation temperature after the nonlinear transformation, yes The transposed vector of represents vector multiplication, For the The oxygen concentration-fermentation temperature interactive correlation matrix between the oxygen concentration local time series correlation feature vector and the fermentation temperature time series correlation feature vector after the nonlinear transformation, yes The eigenvalues ​​at each position in , yes The mean of To calculate the expected value of a matrix, yes The variance of for The kurtosis value of is a hyperparameter, yes The corresponding oxygen concentration-fermentation temperature characteristic interaction energy factor, yes function, yes The corresponding normalized oxygen concentration-fermentation temperature characteristic interaction energy factor, It is the point multiplication by position. is the oxygen concentration-fermentation temperature local-global interaction response representation matrix.

[0057] In step S136, based on the oxygen concentration-fermentation temperature interactive response representation feature, a control instruction is obtained, and the control instruction is used to indicate whether to turn on the ventilation equipment. Specifically, in an embodiment of the present application, based on the oxygen concentration-fermentation temperature local-global interactive response representation matrix, a control instruction is obtained, including: inputting the oxygen concentration-fermentation temperature local-global interactive response representation matrix into a classifier-based ventilation equipment control module to obtain the control instruction. More specifically, in an embodiment of the present application, the oxygen concentration-fermentation temperature local-global interactive response representation matrix is ​​input into a ventilation equipment control module based on a classifier to obtain the control instruction, including: expanding the oxygen concentration-fermentation temperature local-global interactive response representation matrix into an oxygen concentration-fermentation temperature local-global interactive response representation vector according to a row vector or a column vector; using the fully connected layer of the classifier to fully connect the oxygen concentration-fermentation temperature local-global interactive response representation vector to obtain a fully connected encoding feature vector; inputting the fully connected encoding feature vector into the Softmax classification function of the classifier to obtain the probability value of the oxygen concentration-fermentation temperature local-global interactive response representation matrix belonging to each classification label, the classification label includes a label for indicating that the ventilation equipment is turned on and a label for indicating that the ventilation equipment is not turned on; and determining the classification label corresponding to the largest probability value as the control instruction. That is, the oxygen concentration-fermentation temperature local-global interactive response representation matrix obtained by interactively responding the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector is used for classification processing, so as to automatically determine whether to turn on the ventilation equipment. In this way, it is possible to quickly respond to changes in temperature and oxygen concentration during the fermentation process, and based on the temporal interactive response and synergy between the actual fermentation temperature and oxygen concentration, improve the accuracy and response speed of ventilation equipment control, reduce the need for manual monitoring, and realize intelligent control of the fermentation process.

[0058] In particular, the present application takes into account that the fermentation temperature time series correlation feature vector and the oxygen concentration time series correlation feature vector respectively represent the bidirectional short-range-long-range time series correlation characteristics of the fermentation temperature value and the real-time oxygen concentration value. When they are passed through a feature local-global interactive response module based on feature decoupling, the time series correlation feature distribution pattern based on the source time series distribution will also have local-global time series interactive response differences. Therefore, it is expected to improve the convergence mapping iterative consistency of the oxygen concentration-fermentation temperature local-global interactive response representation matrix based on the time series correlation representation differences, from the detailed time series interactive response semantics to the predetermined class probability category representation in the probability space, so as to improve the accuracy of the classification results.

[0059] In a preferred example, inputting the oxygen concentration-fermentation temperature local-global interaction response representation matrix into the classifier-based ventilation equipment control module to obtain a control instruction includes: expanding the oxygen concentration-fermentation temperature local-global interaction response representation matrix into an oxygen concentration-fermentation temperature local-global interaction response representation vector according to row vectors or column vectors; optimizing the oxygen concentration-fermentation temperature local-global interaction response representation vector to obtain an optimized oxygen concentration-fermentation temperature local-global interaction response representation vector; using the fully connected layer of the classifier to fully connect encode the optimized oxygen concentration-fermentation temperature local-global interaction response representation vector to obtain a fully connected encoded feature vector; inputting the fully connected encoded feature vector into the Softmax classification function of the classifier to obtain the probability value of the oxygen concentration-fermentation temperature local-global interaction response representation matrix belonging to each classification label, the classification label including a label for indicating turning on the ventilation equipment and a label for indicating not turning on the ventilation equipment; and determining the classification label corresponding to the largest of the probability values ​​as the control instruction.

[0060] Based on this, the present application optimizes the oxygen concentration-fermentation temperature local-global interaction response representation vector, including: firstly, expanding the oxygen concentration-fermentation temperature local-global interaction response representation matrix into an oxygen concentration-fermentation temperature local-global interaction response representation vector, and optimizing the oxygen concentration-fermentation temperature local-global interaction response representation vector according to the first eigenvalue of the oxygen concentration-fermentation temperature local-global interaction response representation vector. and the second eigenvalue , calculate the absolute value of the difference divided by the sum of the absolute values ​​of the two and then correct the relative probability value, wherein the correction makes the relative probability value greater than zero and less than one; calculate the first product of the first eigenvalue multiplied by the relative probability value, the second product of the second eigenvalue multiplied by the absolute value of the difference between the relative probability value and one, the first quotient of the first eigenvalue divided by the absolute value of the difference between the relative probability value and one, and the second quotient of the second eigenvalue divided by the relative probability value; after adding the first product and the second product, further subtract the difference between the first quotient and the second quotient to obtain the same value as the first eigenvalue and the second eigenvalue The corresponding eigenvalues ​​of the corresponding optimization matrix ; Multiplying the optimization matrix with the oxygen concentration-fermentation temperature local-global interaction response representation vector to obtain an optimized oxygen concentration-fermentation temperature local-global interaction response representation vector.

[0061] Specifically, it is expressed as:

[0062] ;

[0063] ;

[0064] ;

[0065] in, represents the oxygen concentration-fermentation temperature local-global interaction response representation vector, and represent the first eigenvalue and the second eigenvalue of the oxygen concentration-fermentation temperature local-global interaction response representation vector respectively, represents the relative probability value, Represents the first eigenvalue and the second eigenvalue The corresponding eigenvalues ​​of the corresponding optimization matrix, Represents the first eigenvalue and the second eigenvalue The corresponding optimization matrix is, represents matrix multiplication, A vector representing the optimized oxygen concentration-fermentation temperature local-global interaction response;

[0066] Among them, the first eigenvalue and the second eigenvalue The corresponding relative probability value It is expressed as:

[0067] ;

[0068] in, and represent the first eigenvalue and the second eigenvalue of the oxygen concentration-fermentation temperature local-global interaction response representation vector respectively, represents the predetermined hyperparameters, Represents the relative probability value.

[0069] Here, for the region-boundary integral relationship of the high-dimensional feature manifold of the oxygen concentration-fermentation temperature local-global interactive response representation vector, by defining the probability distribution of the feature set of the oxygen concentration-fermentation temperature local-global interactive response representation vector as a manifold constraint boundary, the high-dimensional manifold of the feature set of the oxygen concentration-fermentation temperature local-global interactive response representation vector, which is composed of eigenvalue pairs, is approximated to avoid the mapping ambiguity of the diversified feature representation of the oxygen concentration-fermentation temperature local-global interactive response representation vector to the local manifold representation in the high-dimensional feature classification regression convergence space, and the execution iteration consistency of each local feature distribution in the mapping task is improved to improve the feature classification regression convergence effect, and improve the accuracy of the control instructions obtained by the oxygen concentration-fermentation temperature local-global interactive response representation matrix input based on the classifier ventilation equipment control module. In this way, the temperature and oxygen concentration changes in the fermentation process can be quickly responded to, and the accuracy and response speed of the ventilation equipment control can be improved based on the time-series interactive response and synergy between the actual fermentation temperature and oxygen concentration, reducing the need for manual monitoring and realizing intelligent control of the fermentation process.

[0070] In summary, the preparation method of the root-fertilizer mutual-feed synergistic fertilizer based on the embodiment of the present application is explained, which uses a data analysis technology based on deep learning to perform a time series correlation analysis of the fermentation temperature value and the real-time oxygen concentration value, so as to automatically judge whether to turn on the ventilation equipment according to the interactive response representation characteristics between the fermentation temperature time series correlation characteristics and the oxygen concentration time series correlation characteristics. In this way, it is possible to quickly respond to changes in temperature and oxygen concentration during the fermentation process, and improve the accuracy and response speed of the control of the ventilation equipment based on the time series interactive response and synergy between the actual fermentation temperature and oxygen concentration, thereby reducing the need for manual monitoring and realizing intelligent control of the fermentation process.

[0071] The present application also provides a root-fertilizer mutual-feedback synergistic fertilizer, which can be used according to Figures 1 to 6 The preparation method of the root-fertilizer mutual-feed synergistic fertilizer is prepared as described above, therefore, its repeated description will be omitted.

[0072] The above is only a preferred embodiment of the present application and does not constitute any form of limitation to the present application. Although the present application has been disclosed as a preferred embodiment as above, it is not intended to limit the present application. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of the technical contents disclosed above without departing from the scope of the technical solution of the present application. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application still falls within the scope of the technical solution of the present application.

Claims

1. A method for preparing a root-fertilizer mutual-feed synergistic fertilizer, characterized in that: include: Mixing organic raw materials and adjusting the moisture content to 40%-60% to obtain adjusted mixed organic matter, wherein the organic raw materials include animal manure, plant residues and compost; Adding beneficial microorganisms to the adjusted mixed organic matter and stirring the mixture to obtain a mixed organic matter containing microorganisms, wherein the beneficial microorganisms include nitrogen-fixing bacteria, phosphorus bacteria and silicate bacteria; Placing the mixed organic matter containing microorganisms in a fermentation tank for fermentation treatment to obtain fermented organic matter; Screening the fermented organic matter to obtain screened fertilizer; Drying and packaging the screened fertilizer to obtain a root-fertilizer mutual-feed synergistic fertilizer; The mixed organic matter containing microorganisms is placed in a fermentation tank for fermentation treatment to obtain fermented organic matter, including: Obtaining a time queue of fermentation temperature values ​​collected by a temperature sensor; Obtain a time queue of real-time oxygen concentration values ​​collected by an oxygen concentration sensor; Inputting the time queue of the fermentation temperature values ​​into a fermentation temperature sequence encoder based on a backward LSTM model to obtain a fermentation temperature time series correlation feature vector; Inputting the time queue of the real-time oxygen concentration value into an oxygen concentration sequence encoder based on a backward LSTM model to obtain an oxygen concentration time series correlation feature vector; Performing oxygen concentration-fermentation temperature time series interactive response processing on the fermentation temperature time series associated feature vector and the oxygen concentration time series associated feature vector to obtain oxygen concentration-fermentation temperature interactive response representation features; Based on the oxygen concentration-fermentation temperature interactive response representation characteristics, a control instruction is obtained, wherein the control instruction is used to indicate whether to turn on the ventilation equipment; The fermentation temperature time series correlation feature vector and the oxygen concentration time series correlation feature vector are subjected to oxygen concentration-fermentation temperature time series interactive response processing to obtain oxygen concentration-fermentation temperature interactive response representation features, including: Performing feature decoupling on the oxygen concentration time series correlation feature vector to obtain a sequence of oxygen concentration local time series correlation feature vectors; Calculating the characteristic interaction energy factor of each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors; Based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors, performing local-global interaction response processing on each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain an oxygen concentration-fermentation temperature local-global interaction response representation matrix as the oxygen concentration-fermentation temperature interaction response representation feature; Calculating the characteristic interaction energy factor of each oxygen concentration local time series correlation feature vector in the sequence of the fermentation temperature time series correlation feature vector and the oxygen concentration local time series correlation feature vector to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: Performing a nonlinear transformation on the fermentation temperature time series associated characteristic vector to obtain a nonlinearly transformed fermentation temperature time series associated characteristic vector; Calculating the correlation matrix between the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector to obtain a sequence of oxygen concentration-fermentation temperature interactive correlation matrices; Calculating the characteristic interaction energy factor of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors; Performing a nonlinear transformation on the fermentation temperature time series associated feature vector to obtain a nonlinearly transformed fermentation temperature time series associated feature vector, comprising: multiplying the fermentation temperature time series associated feature vector by a fermentation temperature transformation matrix, and then adding the obtained fermentation temperature time series associated weight vector and fermentation temperature bias vector by position to obtain the nonlinearly transformed fermentation temperature time series associated feature vector; Calculating the correlation matrix between the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector to obtain a sequence of oxygen concentration-fermentation temperature interactive correlation matrices, including: vector multiplying the transposed vector of the fermentation temperature time series correlation feature vector after the nonlinear transformation and each oxygen concentration local time series correlation feature vector in the sequence of the oxygen concentration local time series correlation feature vector to obtain a sequence of the oxygen concentration-fermentation temperature interactive correlation matrix; Calculating the characteristic interaction energy factor of each oxygen concentration-fermentation temperature interaction correlation matrix in the sequence of oxygen concentration-fermentation temperature interaction correlation matrices to obtain a sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors, including: Respectively calculating the mean and variance of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain the oxygen concentration-fermentation temperature interaction correlation mean and the oxygen concentration-fermentation temperature interaction correlation variance; Calculating the positional difference between the oxygen concentration-fermentation temperature interaction correlation matrix and the oxygen concentration-fermentation temperature interaction correlation mean to obtain an oxygen concentration-fermentation temperature interaction correlation difference matrix; Calculating the fourth power of each eigenvalue in the oxygen concentration-fermentation temperature interactive correlation difference matrix to obtain an oxygen concentration-fermentation temperature interactive correlation modulation difference matrix; Calculating the expected value of the oxygen concentration-fermentation temperature interactive correlation modulation difference matrix to obtain the oxygen concentration-fermentation temperature interactive correlation expected value; Dividing the oxygen concentration-fermentation temperature cross-correlation expected value by the square of the oxygen concentration-fermentation temperature cross-correlation variance to obtain the oxygen concentration-fermentation temperature cross-correlation kurtosis value; Calculating the sum of the oxygen concentration-fermentation temperature interaction variance and the hyperparameter to obtain the oxygen concentration-fermentation temperature interaction first energy factor; Calculating the square of the difference between the oxygen concentration-fermentation temperature cross-correlation kurtosis value and the oxygen concentration-fermentation temperature cross-correlation mean to obtain the oxygen concentration-fermentation temperature cross-correlation difference value; The oxygen concentration-fermentation temperature cross-correlation second energy factor is obtained by adding the value obtained by multiplying the oxygen concentration-fermentation temperature cross-correlation variance by constant 2 and the value obtained by multiplying the hyperparameter by constant 2 to the oxygen concentration-fermentation temperature cross-correlation difference value; Dividing the oxygen concentration-fermentation temperature interactive correlation first energy factor by the oxygen concentration-fermentation temperature interactive correlation second energy factor to obtain the oxygen concentration-fermentation temperature characteristic interactive energy factor corresponding to the oxygen concentration-fermentation temperature interactive correlation matrix; Based on the sequence of the oxygen concentration-fermentation temperature characteristic interaction energy factors, local-global interaction response processing is performed on each oxygen concentration local time series association feature vector in the sequence of the fermentation temperature time series association feature vector and the oxygen concentration local time series association feature vector to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix, including: Inputting the sequence of oxygen concentration-fermentation temperature characteristic interaction energy factors into a Sigmoid function to obtain a sequence of normalized oxygen concentration-fermentation temperature characteristic interaction energy factors; Using the sequence of the normalized oxygen concentration-fermentation temperature characteristic interaction energy factors as a sequence of weights, calculating the position-weighted sum of the sequence of the oxygen concentration-fermentation temperature interaction correlation matrix to obtain the oxygen concentration-fermentation temperature local-global interaction response representation matrix; Based on the oxygen concentration-fermentation temperature local-global interaction response representation matrix, obtaining a control instruction includes: inputting the oxygen concentration-fermentation temperature local-global interaction response representation matrix into a classifier-based ventilation equipment control module to obtain the control instruction; Inputting the oxygen concentration-fermentation temperature local-global interaction response representation matrix into a classifier-based ventilation equipment control module to obtain the control instruction, including: Expanding the oxygen concentration-fermentation temperature local-global interaction response representation matrix into an oxygen concentration-fermentation temperature local-global interaction response representation vector according to row vectors or column vectors; Using the fully connected layer of the classifier to perform fully connected encoding on the oxygen concentration-fermentation temperature local-global interaction response representation vector to obtain a fully connected encoding feature vector; Inputting the fully connected encoded feature vector into the Softmax classification function of the classifier to obtain the probability value of the oxygen concentration-fermentation temperature local-global interaction response representation matrix belonging to each classification label, wherein the classification label includes a label for indicating that the ventilation device is turned on and a label for indicating that the ventilation device is not turned on; The classification label corresponding to the largest probability value among the probability values ​​is determined as the control instruction.

2. A root-fertilizer mutual-feed synergistic fertilizer, characterized in that: The root-fertilizer mutual-feed synergistic fertilizer is prepared by the preparation method of the root-fertilizer mutual-feed synergistic fertilizer according to claim 1.

Images