A method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer

Abstract

This invention relates to the field of bio-organic fertilizer fermentation monitoring technology, specifically to a method for monitoring and optimizing Bacillus bio-organic fertilizer fermentation conditions. The method includes: acquiring data change characterization values ​​and related difference characterization values ​​for a first detection point; obtaining the target Bacillus acid production characterization value for the first detection point based on these values; further screening the first detection point based on the target Bacillus acid production characterization value; obtaining a predicted pH value for future monitoring times based on the screening results; and monitoring the Bacillus bio-organic fertilizer fermentation process based on the predicted pH value for future monitoring times. Furthermore, this invention improves the timeliness and effectiveness of monitoring the pH value of the fermentation broth in the fermenter by predicting the pH value.

Description

Technical Field

[0001] This invention relates to the field of bio-organic fertilizer fermentation monitoring technology, specifically to a method for monitoring and optimizing Bacillus bio-organic fertilizer fermentation conditions. Background Technology

[0002] During the fermentation of Bacillus bio-organic fertilizer, such as in the fermentation of Bacillus bereaves TA-1, the pH value of the fermentation liquid in the fermenter is usually monitored to ensure fermentation quality, production quality, or production efficiency of the bio-organic fertilizer. Since lactic acid and acetic acid are easily produced during the fermentation of Bacillus bio-organic fertilizer, excessive production of these substances can lead to a decrease in the pH value of the fermentation liquid in the fermenter or cause over-acidity, meaning the solution in the fermenter is highly acidic. Conversely, excessively low pH values ​​directly affect fermentation quality and speed, thus impacting the production quality or efficiency of the bio-organic fertilizer. Therefore, monitoring the pH value of the fermentation liquid in the fermenter is crucial during the fermentation process of Bacillus bio-organic fertilizer, specifically monitoring for over-acidity. The fermenter is the container used for the fermentation of Bacillus bio-organic fertilizer.

[0003] In existing technologies, pH values ​​are typically monitored in real time based on empirical values ​​or fixed thresholds. The pH value in the fermenter is the actual pH level. However, this method suffers from poor timeliness. For example, if the pH value collected at the current monitoring moment is lower than the preset threshold, the solution in the fermenter is already too acidic, affecting the fermentation process and ultimately the final result. Therefore, optimization of the monitoring process is necessary. Specifically, optimizing the pH monitoring process in the fermentation of Bacillus bio-organic fertilizer to ensure timely and effective monitoring is a pressing issue. Summary of the Invention

[0004] To address the above problems, this invention provides a method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer, the specific technical solution of which is as follows:

[0005] One embodiment of the present invention provides a method for monitoring and optimizing the fermentation conditions of Bacillus bio-organic fertilizer, comprising the following steps:

[0006] Obtain the pH monitoring time sequence corresponding to different detection points in the fermenter, wherein the fermenter is a container for fermenting Bacillus bio-organic fertilizer;

[0007] Based on the data in the pH monitoring time series, the detection points are screened to obtain a first detection point. Based on the pH monitoring time series corresponding to the first detection point, a target sequence corresponding to the first detection point is obtained. Based on the difference between adjacent pH values ​​in the target sequence, a data change characterization value for the first detection point is obtained. Based on the length of the target sequence, a set of sequences to be analyzed for the first detection point is obtained. Based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, a correlation difference characterization value for the first detection point is obtained. Based on the data change characterization value and the correlation difference characterization value for the first detection point, a target Bacillus acid production characterization value for the first detection point is obtained.

[0008] Based on the acid production characterization value of the target Bacillus, the first detection point is screened again, and the predicted pH value at the future monitoring time is obtained based on the screening results. The fermentation process of the Bacillus bio-organic fertilizer is monitored based on the predicted pH value at the future monitoring time.

[0009] Beneficial effects: This invention first obtains the pH monitoring time series corresponding to different detection points in the fermenter; then, based on the data in the pH monitoring time series, the detection points are screened to obtain the first detection point, and based on the pH monitoring time series corresponding to the first detection point, the target sequence corresponding to the first detection point is obtained; based on the difference between adjacent pH values ​​in the target sequence, the data change degree characterization value of the first detection point is obtained; based on the length of the target sequence, the set of sequences to be analyzed for the first detection point is obtained; and based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, the correlation difference characterization value of the first detection point is obtained; then, based on the data change degree characterization value and the correlation difference characterization value of the first detection point, the acid production characterization value of the target Bacillus at the first detection point is obtained; finally, based on the acid production characterization value of the target Bacillus, the first detection point is screened again, and based on the screening results, the predicted pH value at future monitoring times is obtained; and the fermentation process of Bacillus bio-organic fertilizer is monitored based on the predicted pH value at future monitoring times. Furthermore, by predicting the pH value, this invention can detect the tendency or risk of over-acidity in time before it occurs, and take corresponding measures to avoid the occurrence of over-acidity. In other words, by predicting the pH value, this invention can improve the timeliness or effectiveness of monitoring the pH value of the fermentation liquid in the fermenter. Attached Figure Description

[0010] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart of a method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer according to the present invention. Detailed Implementation

[0012] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the protection scope of the embodiments of the present invention.

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art.

[0014] This embodiment provides a method for monitoring and optimizing the fermentation conditions of Bacillus bio-organic fertilizer, which is described in detail below:

[0015] like Figure 1 As shown, the method for monitoring and optimizing the fermentation conditions of Bacillus bio-organic fertilizer includes the following steps:

[0016] Step S001: Obtain the pH monitoring time sequence corresponding to different detection points in the fermenter, wherein the fermenter is a container for Bacillus bio-organic fertilizer fermentation.

[0017] Currently, monitoring of the Bacillus bio-organic fertilizer fermentation process suffers from poor timeliness, which negatively impacts fermentation quality, bio-organic fertilizer production quality, and production efficiency. Therefore, this embodiment optimizes the existing monitoring process to improve timeliness and effectiveness. The existing monitoring process relies on empirical values ​​or fixed thresholds to monitor the real-time pH value, where the pH value is the solution pH in the fermentation tank. Since Bacillus bio-organic fertilizer is typically fermented in a fermentation tank, the timeliness of the monitoring process is crucial. The process is monitored, specifically the pH value of the solution in the fermenter. The fermenter is a container for Bacillus bio-organic fertilizer fermentation, and in this embodiment, it is a non-mechanically stirred and ventilated fermenter. Furthermore, since there are many types of Bacillus, this embodiment will subsequently describe the monitoring process of any one type of Bacillus bio-organic fertilizer fermentation as an example. For instance, in this embodiment, all Bacillus species mentioned are Bacillus belyssus TA-1, which is a bacterium belonging to the Bacillus genus and can be used as a plant disease control agent, especially showing significant application potential in the protection against plant pathogenic nematodes.

[0018] In addition, for ease of understanding and analysis, this embodiment will only monitor the pH value of the solution in any one fermenter. That is, the pH value collected in the future will be the pH value of the solution in the same fermenter. Furthermore, this embodiment mainly monitors the over-acidity caused by acidic substances produced during fermentation. The acidic substances produced during fermentation include, but are not limited to, lactic acid and acetic acid. That is, this embodiment mainly monitors the changes in pH value in the fermenter caused by acidic substances. Moreover, when over-acidity is detected or the risk of over-acidity is high, the pH of the solution in the fermenter will be adjusted immediately, such as by injecting an alkaline solution.

[0019] Firstly, to ensure monitoring accuracy, this embodiment arranges detection points at different locations within the fermenter. The number and distribution of these detection points need to be adaptively set based on factors such as the fermenter's volume and monitoring accuracy. For example, if the fermenter's volume is small, the number of detection points can be set to 4; if the volume is large, the number can be set to 8. Furthermore, the detection points can be configured to be evenly or randomly distributed within the fermenter. After determining the detection points within the fermenter, the corresponding data for each detection point is acquired using the detection instrument at various monitoring times within the current monitoring period. The pH value is then used to construct a time series of all pH values ​​corresponding to each detection point acquired during the current monitoring period. This time series is recorded as the pH monitoring time series corresponding to the detection point. The current monitoring period includes the current monitoring time. Therefore, the last pH value in the pH monitoring time series is the pH value collected at the current monitoring time. The detection instrument refers to a pH sensor. In addition, in specific applications, the implementer needs to set the time interval between adjacent monitoring times and the duration of the current monitoring period according to the actual situation. For example, in this embodiment, the time interval between adjacent monitoring times can be set to 1 second, and the duration of the current monitoring period can be set to 2 minutes.

[0020] Therefore, this embodiment can obtain the pH monitoring time sequence corresponding to each detection point in the fermenter through the above process during the current monitoring period.

[0021] Step S002: Based on the data in the pH monitoring time series, the detection points are screened to obtain a first detection point. Based on the pH monitoring time series corresponding to the first detection point, a target sequence corresponding to the first detection point is obtained. Based on the difference between adjacent pH values ​​in the target sequence, a data change degree characterization value for the first detection point is obtained. Based on the length of the target sequence, a set of sequences to be analyzed for the first detection point is obtained. Based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, a correlation difference characterization value for the first detection point is obtained. Based on the data change degree characterization value and the correlation difference characterization value for the first detection point, a target Bacillus acid production characterization value for the first detection point is obtained.

[0022] This embodiment will next analyze the acid production level of Bacillus at the detection points, i.e., the acid production characterization value of the target Bacillus, to obtain the predicted pH value in the fermenter at future monitoring times. Finally, the fermentation process of Bacillus bio-organic fertilizer will be monitored based on the predicted pH value. Using the predicted pH value can improve the timeliness or effectiveness of monitoring the fermentation process of Bacillus bio-organic fertilizer. However, before obtaining the acid production level of Bacillus, it is necessary to first screen all detection points based on the data in the pH monitoring time series corresponding to each detection point to obtain the first detection point. Then, based on the pH monitoring time series corresponding to the first detection point, the target sequence corresponding to the first detection point is obtained. Based on the difference between adjacent pH values ​​in the target sequence, the data change characterization value of the first detection point is obtained. The data change degree characterization value can reflect whether there is a decrease in pH caused by acidic substances at the corresponding detection point; then, based on the length of the target sequence, the set of sequences to be analyzed for the first detection point is obtained, and based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, the correlation difference characterization value of the first detection point is obtained, and the correlation difference characterization value can further reflect whether there is a decrease in pH caused by acidic substances at the corresponding detection point; then, based on the obtained data change degree characterization value and correlation difference characterization value of the first detection point, the acid production characterization value of the target Bacillus at the first detection point is obtained. In this embodiment, the acidic substances mentioned refer to the acidic substances produced as the fermentation process of Bacillus proceeds.

[0023] Therefore, based on the above analysis, this embodiment needs to first screen all detection points according to the data in the pH monitoring time series corresponding to each detection point to obtain the first detection point. The purpose of the screening is to obtain the detection points where the pH decrease is suspected to be caused by the presence of acidic substances at the current monitoring time. In this embodiment, the specific process of screening all detection points according to the data in the pH monitoring time series corresponding to each detection point to obtain the first detection point is as follows:

[0024] First, based on the data in the pH monitoring time series corresponding to each detection point, the suspected fermentation acid production characterization value corresponding to each detection point is obtained. This suspected fermentation acid production characterization value reflects the possibility of a decrease in pH value caused by the presence of acidic products at the corresponding detection point during the fermentation process of Bacillus bio-organic fertilizer. In this embodiment, the specific process for obtaining the suspected fermentation acid production characterization value corresponding to each detection point is as follows: For any detection point: First, the second-to-last pH value and the last pH value in the pH monitoring time series corresponding to that detection point are recorded as the first data value and the second data value, respectively. The result obtained by subtracting the second data value from the first data value is recorded as the first difference. Then, a predetermined number of consecutive data points preceding the last pH value in the pH monitoring time series corresponding to the detection point are obtained, and the new sequence formed by the predetermined number of consecutive data points is recorded as the feature subsequence. Then, the absolute value of the difference between the mean of the feature subsequence and the second data value is obtained and recorded as the second difference. The reciprocal of the result obtained by adding the second difference to a predetermined first constant is recorded as the first ratio. Finally, the product of the first difference and the first ratio is obtained and recorded as the suspected fermentation acid production characterization value corresponding to the detection point.

[0025] Therefore, this embodiment can obtain the suspected fermentation acid production characterization value corresponding to each detection point through the above process. The larger the suspected fermentation acid production characterization value corresponding to a detection point, the more likely the pH change at that detection point at the current monitoring time is caused by acidic substances leading to a pH decrease. Therefore, after obtaining the suspected fermentation acid production characterization value corresponding to each detection point, the detection points are initially screened. The specific process of the initial screening is as follows: For any detection point, firstly determine whether the suspected fermentation acid production characterization value corresponding to that detection point is positive. If so, then further determine whether the normalized value of the suspected fermentation acid production characterization value corresponding to that detection point is greater than or equal to the preset suspected judgment threshold. If so, it is determined that the pH change at that detection point at the current monitoring time is more likely to be caused by acidic substances leading to a pH decrease, and further analysis is required. Therefore, this detection point is recorded as the first detection point. Additionally, if the suspected fermentation acid production characterization value corresponding to a certain detection point is... A negative acid production characterization value indicates that the second-to-last pH value in the pH monitoring time series corresponding to that detection point is less than the last pH value in the same time series. This suggests that there is no acidic substance causing the pH decrease at that detection point at the current detection time. Furthermore, if all suspected fermentation acid production characterization values ​​corresponding to the detection point are negative, it can be directly determined that the fermentation process at the current time may not have produced any acidic substances. Therefore, over-acidity will not occur in the fermenter for a considerable period of time, and there is no need to adjust the pH of the solution in the fermenter at the current monitoring time. In other words, if all suspected fermentation acid production characterization values ​​corresponding to the detection point are negative, it is determined that no alkaline solution needs to be injected into the fermenter at the current monitoring time. The normalized value of the suspected fermentation acid production characterization value mentioned above refers to the value obtained by normalizing the corresponding suspected fermentation acid production characterization value using the normalization function Norm().

[0026] Furthermore, in specific applications, implementers need to set the values ​​of the preset suspected judgment threshold, the preset quantity, and the preset first constant according to the actual situation. For example, in this embodiment, the preset suspected judgment threshold can be set to 0.6, the preset quantity can be set to 100, and the preset first constant can be set to 0.1.

[0027] Therefore, this embodiment completes the initial screening of detection points through the above process and obtains all the first detection points among all detection points. After obtaining the first detection points, the pH monitoring time series corresponding to each first detection point is analyzed to obtain the target sequence corresponding to the first detection point. Furthermore, by analyzing the target sequence, the probability of pH decrease caused by acidic substances at each first detection point at the current monitoring time can be obtained more accurately. Therefore, the specific process of obtaining the target sequence corresponding to the first detection point in this embodiment is as follows:

[0028] For any first detection point: First, determine whether the sequence formed by the AB-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is a decreasing sequence, and whether the sequence formed by the A-(B-1)-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is not a decreasing sequence. If it is determined that the sequence formed by the AB-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is a decreasing sequence, and the sequence formed by the A-(B-1)-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is not a decreasing sequence, then the sequence formed by the AB-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is recorded as the target sequence corresponding to the first detection point. Furthermore, if the sequence formed by the AB-th pH value to the A-th pH value in the pH monitoring time series corresponding to the first detection point is a decreasing sequence and the first... If the sequence formed from the A-(B-1)th pH value to the Ath pH value in the pH monitoring time series corresponding to the detection point is not a decreasing sequence, it indicates that starting from the ABth pH value in the pH monitoring time series corresponding to the first detection point, acidic substances may have appeared at the location of the first detection point, causing a decrease in pH. As the fermentation time increases, the production of acidic substances also gradually increases, meaning that the pH value of the solution in the fermenter will gradually decrease with the increase in fermentation time. Based on this, it can be seen that only when the pH at the detection point shows a decreasing or decreasing characteristic can there be an over-acidity phenomenon, and only then is it worth monitoring. Therefore, this embodiment needs to extract the data collected at the current monitoring time and the decreasing sequence, which is the target sequence obtained above. Subsequently, the decreasing characteristic of the target sequence will be further analyzed to determine whether it is caused by acidic substances. In addition, A is the total number of data in the pH monitoring time series corresponding to the first detection point, and B is less than A and is a positive integer.

[0029] Therefore, this embodiment can obtain the target sequence corresponding to each first detection point through the above process. After obtaining the target sequence corresponding to each first detection point, the data change degree characterization value of each first detection point is obtained based on the difference between adjacent pH values ​​in the target sequence. The change degree characterization value can reflect the possibility that the target sequence is caused by acidic substances. Therefore, in this embodiment, the specific process of obtaining the data change degree characterization value of the first detection point is as follows:

[0030] For any first detection point: First, obtain the change feature value sequence of the target sequence corresponding to the first detection point, where the b-th change feature value in the change feature value sequence is the result of subtracting the (b+1)-th change feature value in the change feature value sequence from the b-th change feature value in the target sequence corresponding to the first detection point; then, obtain the reverse sequence of the change feature value sequence and denote it as the reverse sequence to be analyzed corresponding to the first detection point. For example, if the change feature value sequence is {u1,u2,u3}, then the reverse sequence to be analyzed is {u3,u2,u1}; then, obtain the feature difference sequence of the reverse sequence to be analyzed corresponding to the first detection point and the weight coefficient of each feature difference in the feature difference sequence, where the d-th feature difference in the feature difference sequence is the result of subtracting the (b+1)-th change feature value in the reverse sequence to be analyzed. The result of the (d+1)th change feature value in the sequence, and the weight coefficient of the dth feature difference in the feature difference sequence is the normalized value of the position representation value of the dth feature difference, and the position representation value of the dth feature difference is the reciprocal of d; in addition, the normalized value of the position representation value of the dth feature difference refers to the ratio of the position representation value of the dth feature difference to the comprehensive position representation value, and the comprehensive position representation value is the sum of the position representation values ​​of all feature differences in the feature difference sequence; then, the weighted feature difference sequence of the feature difference sequence is obtained, and the normalized value of the sum of all weighted feature differences in the weighted feature difference sequence is recorded as the data change degree representation value of the first detection point, and the product of the weight coefficient of the dth feature difference and the dth feature difference is the dth weighted feature difference in the weighted feature difference sequence.

[0031] Furthermore, the specific expression for calculating the data change characterization value of the first detection point is as follows:

[0032]

[0033] Where F is the data change characteristic value of the first detection point, D is the total number of feature differences in the feature difference sequence of the reverse sequence to be analyzed corresponding to the first detection point, and δ d G is the normalized value representing the position of the d-th feature difference in the feature difference sequence. d Let G be the d-th feature difference in the feature difference sequence. Norm() is the normalization function, and the above positional representation value can reflect the reference value of the feature difference. The closer to the current monitoring time, the higher the value of G. d The larger the value, the greater the reference value of the corresponding feature difference.

[0034] Furthermore, in actual Bacillus fermentation processes, due to the long fermentation time, the substances inside the pH sensor probe electrode can change over time or due to the effects of high-temperature steam sterilization. This can lead to data drift, a phenomenon more likely to occur during long-term Bacillus fermentation. However, sensor-induced drift can also cause continuously decreasing data segments in the monitored data sequence. In other words, the target sequence obtained above may also be due to drift. However, the continuously decreasing data segments caused by drift and the continuously decreasing data segments caused by acidic substances are further complicated by this drift. The data distribution within the segments differs, indicating that the drift phenomenon is influenced by long-term environmental conditions. Therefore, compared to the data segments experiencing a continuous decline due to acidic substances, the data changes in the data segments experiencing a continuous decline due to drift are slower and approximately linear. However, during the fermentation process, the number of Bacillus strains gradually increases, leading to a gradual increase in the amount of acidic substances produced. Therefore, compared to the data segments experiencing a continuous decline due to drift, the data segments experiencing a continuous decline due to acidic substances change more rapidly and the magnitude of these changes becomes increasingly drastic. Based on the aforementioned characterization value of the degree of data change calculated from the target sequence, it can be seen that when δ... d ×G d The larger the value of F, the larger the value of F. The larger the value of F, the more drastic the data change in the target sequence corresponding to the first detection point becomes. It also indicates that the data change in the target sequence corresponding to the first detection point is more likely to be caused by acidic substances produced by fermentation. In other words, the decreasing characteristic of the target sequence corresponding to the first detection point is more likely to be caused by acidic substances produced by fermentation.

[0035] Therefore, this embodiment can obtain the data change degree characterization value of each first detection point through the above process. After obtaining the data change degree characterization value of the first detection point, in order to further determine whether the data change in the target sequence corresponding to the first detection point is caused by acidic substances produced by fermentation, this embodiment will obtain the relevant difference characterization value of the first detection point. The relevant difference characterization value is also an important parameter reflecting whether the data change in the target sequence corresponding to the first detection point is caused by acidic substances produced by fermentation. However, before obtaining the relevant difference characterization value, it is necessary to obtain the set of sequences to be analyzed for each first detection point according to the length of the target sequence corresponding to each first detection point. The set of sequences to be analyzed is the basis for subsequent analysis to obtain the relevant difference characterization value. So, the specific process of obtaining the set of sequences to be analyzed for each first detection point in this embodiment is as follows:

[0036] For any first detection point: First, obtain all detection points in the fermenter except for the first detection point, and denote the set of all detection points in the fermenter except for the first detection point as the first set. Then, obtain the total number of data in the target sequence corresponding to the first detection point and denote it as the feature quantity representation value. Then, obtain the sequence to be analyzed corresponding to each detection point in the first set according to the feature quantity representation value, and denote the set of the sequence to be analyzed corresponding to all detection points in the first set as the sequence to be analyzed set of the first detection point. For the g-th detection point in the first set, denote the sequence constructed by the last feature quantity representation value and the pH value in the pH monitoring time series corresponding to the g-th detection point as the sequence to be analyzed corresponding to the g-th detection point. That is, the length of the sequence to be analyzed corresponding to the g-th detection point is the same as the length of the target sequence corresponding to the first detection point.

[0037] Therefore, this embodiment can obtain the set of sequences to be analyzed for each first detection point through the above process. The specific reason for obtaining the set of sequences to be analyzed for each first detection point is as follows: In the fermenter, the pH value directly monitored by the pH sensor is mainly affected by the bacterial strains around the sensor probe. That is, the degree of pH change caused by the acidic substances produced by Bacillus will vary with the strain density. Moreover, since the growth process of the strains during fermentation is an entropy-increasing process, the strain density is randomly distributed. That is, at different detection points, the correlation between pH changes caused by acidic substances is small. Therefore, for a certain first detection point, if the data change in the target sequence of that detection point is caused by the acidic substances produced by fermentation, then the target sequence of that detection point and each sequence in the set of sequences to be analyzed for that first detection point are considered to be related. The correlation between the analyzed sequences is relatively small. However, if the data change in the target sequence of the first detection point is not caused by acidic substances produced by fermentation, but by other reasons, such as high-temperature steam sterilization, then the correlation between the target sequence of the first detection point and each sequence to be analyzed in the set of sequences to be analyzed at the first detection point is relatively large. Based on the above analysis, after obtaining the set of sequences to be analyzed at the first detection point, this embodiment obtains the correlation difference characterization value of each first detection point according to the correlation between the target sequence corresponding to each first detection point and each sequence to be analyzed in the set of sequences to be analyzed at the corresponding first detection point. The specific process for obtaining the correlation difference characterization value of each first detection point is as follows:

[0038] For any first detection point: First, obtain the normalized Pearson correlation coefficient between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the first detection point, and record them as correlation characterization values. Then, the sequence constructed from all the obtained correlation characterization values ​​is recorded as the correlation characterization value sequence corresponding to the first detection point. The h-th correlation characterization value in the correlation characterization value sequence corresponding to the first detection point is the normalized Pearson correlation coefficient between the target sequence corresponding to the first detection point and the h-th sequence to be analyzed. The h-th sequence to be analyzed is the h-th sequence to be analyzed in the set of sequences to be analyzed for the first detection point, that is, the correlation characterization value sequence corresponding to the first detection point. The number of data points is consistent with the number of sequences to be analyzed in the set of sequences to be analyzed at the first detection point. Then, the mean value of the correlation characterization value sequence corresponding to the first detection point is obtained and recorded as the correlation mean. Then, the correlation mean is negatively correlated and mapped, and the mapping result is recorded as the correlation difference characterization value of the first detection point. The smaller the correlation mean, the larger the correlation difference characterization value of the first detection point. The larger the correlation difference characterization value of the first detection point, the greater the probability that the data change in the target sequence corresponding to the first detection point is caused by acidic substances produced by fermentation. That is, the probability that the decreasing feature of the target sequence corresponding to the first detection point is caused by acidic substances produced by fermentation is greater.

[0039] In addition, the normalized value of the Pearson correlation coefficient between the target sequence corresponding to the first detection point and the h-th sequence to be analyzed is... R h Let be the Pearson correlation coefficient between the target sequence corresponding to the first detection point and the h-th sequence to be analyzed. The normalization here is to control the correlation value between 0 and 1. The correlation difference value of the first detection point is exp(-N0), where exp() is an exponential function with a base of constant e, and N0 is the mean correlation value.

[0040] Therefore, this embodiment can obtain the relevant difference characterization value of each first detection point through the above process. Then, based on the data change degree characterization value and the relevant difference characterization value of each first detection point, the target Bacillus acid production characterization value of each first detection point is obtained. That is, the specific process of obtaining the target Bacillus acid production characterization value of the first detection point is as follows: for any first detection point, the average of the data change degree characterization value of the first detection point and the relevant difference characterization value of the first detection point is taken as the target Bacillus acid production characterization value of the first detection point. The larger the target Bacillus acid production characterization value of the first detection point, the greater the probability that the decreasing feature of the target sequence corresponding to the first detection point is caused by acidic substances produced by fermentation.

[0041] Therefore, this embodiment can obtain the acid production characterization value of the target Bacillus at each first detection point through the above process.

[0042] Step S003: Based on the acid production characterization value of the target Bacillus, the first detection point is screened again, and the predicted pH value at the future monitoring time is obtained based on the screening results. The fermentation process of the Bacillus bio-organic fertilizer is monitored based on the predicted pH value at the future monitoring time.

[0043] In this embodiment, after obtaining the acid production characterization values ​​of the target Bacillus at each first detection point, all first detection points are screened again based on these values. The predicted pH value for future monitoring times is then obtained based on the screening results. The specific acquisition process is as follows:

[0044] First, among all the first detection points, all first detection points whose target Bacillus acid production characterization value is not less than the preset Bacillus acid production characterization value threshold are obtained and recorded as second detection points. In this embodiment, the decreasing characteristic of the target sequence corresponding to the first detection point whose target Bacillus acid production characterization value is not less than the preset Bacillus acid production characterization value threshold is caused by acidic substances produced by fermentation. In addition, in specific applications, the implementer needs to set the preset Bacillus acid production characterization value threshold according to the actual situation. For example, in this embodiment, the preset Bacillus acid production characterization value threshold can be set to 0.4.

[0045] After obtaining the second detection point, the least squares method is used to perform linear fitting on the target sequence corresponding to each second detection point, and the straight line obtained by fitting is recorded as the fitting straight line of the corresponding second detection point. That is, a fitting straight line can be obtained by linearly fitting a second detection point. Then, the fitting straight lines of all second detection points are averaged to obtain the average fitting straight line, which is recorded as the target fitting straight line. The horizontal axis of the fitting straight line is time, and the vertical axis is pH value. The slope of the average fitting straight line is the mean of the slopes of the fitting straight lines of all second detection points.

[0046] Furthermore, since the straight line obtained through linear fitting can be used to predict future data, after obtaining the target fitted straight line, this embodiment obtains the predicted pH value for future monitoring times based on the obtained target fitted straight line. In this embodiment, the predicted pH value for future monitoring times refers to the predicted pH value corresponding to the next monitoring time after the current monitoring time. The predicted pH value corresponding to the next monitoring time after the current monitoring time refers to the ordinate value of the data point located on the target fitted straight line whose x-coordinate value is the same as the time corresponding to the next monitoring time after the current monitoring time. That is, if the time corresponding to the next monitoring time after the current monitoring time is T0, then the ordinate value of the data point on the target fitted straight line whose x-coordinate value is equal to T0 is the predicted pH value corresponding to the next monitoring time after the current monitoring time. And typically, the time corresponding to the next monitoring time after the current monitoring time is... The time interval between moments is consistent with the time interval between any adjacent monitoring moments in the current monitoring period. Alternatively, as another real-time method, the predicted pH value for future monitoring moments can also be the predicted pH value for each moment in the future monitoring period. This requires that the length of the future monitoring period is consistent with the length of the current monitoring period, the future monitoring period is adjacent to and continuous with the current monitoring period, and the future monitoring period is located after the current monitoring period in time. The predicted pH value for any moment t1 in the future monitoring period refers to the ordinate value of the data point on the target fitting line whose x-coordinate value is the same as the time corresponding to moment t1. That is, if the time corresponding to any moment t1 in the future monitoring period is T1, then the ordinate value of the data point on the target fitting line whose x-coordinate value is equal to T1 is the predicted pH value for moment t1.

[0047] Therefore, this embodiment obtains the predicted pH value at the future monitoring time through the above process. After obtaining the predicted pH value at the future monitoring time, this embodiment monitors the fermentation process of Bacillus bio-organic fertilizer based on the predicted pH value at the future monitoring time, specifically as follows:

[0048] First, determine if the predicted pH value is lower than the preset pH threshold. If so, it indicates a very high probability of over-acidity occurring at a future monitoring time or within a future monitoring period close to the current monitoring time. Therefore, to minimize the impact of over-acidity on fermentation, it is determined that an alkaline solution should be injected into the fermenter at the current monitoring time to reduce the probability of over-acidity. Otherwise, it indicates a very low probability of over-acidity occurring within a future monitoring period or within a future monitoring time close to the current monitoring time. In this case, there is no need to inject alkaline solution into the fermenter. Furthermore, this embodiment can detect the tendency towards over-acidity by predicting the pH value, even before the over-acidity phenomenon occurs, and then take corresponding measures to avoid the occurrence of over-acidity. Moreover, the amount of alkaline solution injected needs to be set by the implementer according to the actual situation. For example, the amount of alkaline solution injected can be determined based on the difference between the actual pH value collected at the current monitoring time and the preset pH threshold. It is also necessary to meet the requirement that the smaller the result of the actual pH value collected at the current monitoring time minus the preset pH threshold, the more alkaline solution needs to be injected.

[0049] In addition, in specific applications, implementers need to set a preset pH threshold according to the actual situation. In this embodiment, the preset pH threshold is the minimum pH threshold specified. For example, the preset pH threshold can be set to 5.5.

[0050] In summary, this embodiment first obtains the pH monitoring time series corresponding to different detection points in the fermenter; then, based on the data in the pH monitoring time series, the detection points are screened to obtain the first detection point, and based on the pH monitoring time series corresponding to the first detection point, the target sequence corresponding to the first detection point is obtained; based on the difference between adjacent pH values ​​in the target sequence, the data change degree characterization value of the first detection point is obtained; based on the length of the target sequence, the set of sequences to be analyzed for the first detection point is obtained; and based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, the correlation difference characterization value of the first detection point is obtained; then, based on the data change degree characterization value and the correlation difference characterization value of the first detection point, the acid production characterization value of the target Bacillus at the first detection point is obtained; finally, based on the acid production characterization value of the target Bacillus, the first detection point is screened again, and based on the screening results, the predicted pH value at future monitoring times is obtained; and the fermentation process of Bacillus bio-organic fertilizer is monitored based on the predicted pH value at future monitoring times. Furthermore, this embodiment can detect the tendency or risk of over-acidity in the fermentation liquid in the fermenter in a timely manner by predicting the pH value, even before the over-acidity phenomenon occurs, and take corresponding measures to avoid the occurrence of over-acidity. In other words, this embodiment can improve the timeliness or effectiveness of monitoring the pH value of the fermentation liquid in the fermenter by predicting the pH value.

[0051] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer, characterized in that, The method includes the following steps: Obtain the pH monitoring time sequence corresponding to different detection points in the fermenter, wherein the fermenter is a container for fermenting Bacillus bio-organic fertilizer; Based on the data in the pH monitoring time series, the detection points are screened to obtain the first detection point, and based on the pH monitoring time series corresponding to the first detection point, the target sequence corresponding to the first detection point is obtained. Based on the difference between adjacent pH values ​​in the target sequence, the data change degree characterization value of the first detection point is obtained. Based on the length of the target sequence, a set of sequences to be analyzed for the first detection point is obtained, and based on the correlation between the target sequence corresponding to the first detection point and each sequence to be analyzed in the set of sequences to be analyzed for the corresponding first detection point, the correlation difference characterization value of the first detection point is obtained. Based on the data change characterization value and related difference characterization value of the first detection point, the acid production characterization value of the target Bacillus at the first detection point is obtained; Based on the acid production characterization value of the target Bacillus, the first detection point is screened again, and the predicted pH value at the future monitoring time is obtained based on the screening results. The fermentation process of the Bacillus bio-organic fertilizer is monitored based on the predicted pH value at the future monitoring time. The method for selecting the first detection point based on data from the pH monitoring time series includes: Based on the data in the pH monitoring time series corresponding to each detection point, the suspected fermentation acid production characterization value corresponding to each detection point is obtained; For any detection point, if the suspected fermentation acid production characterization value corresponding to the detection point is positive, then it is further determined whether the normalized value of the suspected fermentation acid production characterization value corresponding to the detection point is not less than the preset suspected judgment threshold. If so, the detection point is recorded as the first detection point. The method for obtaining the suspected fermentation acid production characterization values ​​corresponding to each detection point includes: For any detection point, the penultimate and last pH values ​​in the pH monitoring time series corresponding to the detection point are respectively recorded as the first data value and the second data value. The result of subtracting the second data value from the first data value is recorded as the first difference. A new sequence consisting of a predetermined number of consecutive data points preceding the last pH value in the pH monitoring time series corresponding to the detection point is recorded as a feature subsequence. The absolute value of the difference between the mean of the feature subsequence and the second data value is recorded as the second difference. The reciprocal of the result obtained by adding the second difference to a predetermined first constant is recorded as the first ratio. The product of the first difference and the first ratio is recorded as the suspected fermentation acid production characterization value corresponding to the detection point. The method for obtaining the target sequence corresponding to the first detection point includes: For any first detection point, if the sequence formed from the AB-th pH value in the pH monitoring time series corresponding to the first detection point to the A-th pH value in the pH monitoring time series corresponding to the first detection point is a decreasing sequence, and the sequence formed from the A-(B-1)-th pH value in the pH monitoring time series corresponding to the first detection point to the A-th pH value is not a decreasing sequence, then the sequence formed from the AB-th pH value to the A-th pH value is recorded as the target sequence corresponding to the first detection point, where A is the total number of data in the pH monitoring time series corresponding to the first detection point, and B is less than A; The method for obtaining the predicted pH value at the future monitoring time includes: In all the first detection points, all first detection points whose target Bacillus acid production characterization value is not less than the preset Bacillus acid production characterization value threshold are recorded as second detection points; linear fitting is performed on the target sequence corresponding to each second detection point, and the straight line obtained by fitting is recorded as the fitting straight line corresponding to the second detection point, and the average fitting straight line of all fitting straight lines is recorded as the target fitting straight line, where the horizontal axis of the fitting straight line is time and the vertical axis is pH value; based on the target fitting straight line, the predicted pH value at future monitoring times is obtained.

2. The method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer as described in claim 1, characterized in that, The method for obtaining the data change characterization value of the first detection point includes: For any first detection point: Obtain the change feature value sequence of the target sequence corresponding to the first detection point. The b-th change feature value in the change feature value sequence is the result of subtracting the (b+1)-th change feature value in the change feature value sequence from the b-th change feature value in the target sequence corresponding to the first detection point. Record the reverse sequence of the change feature value sequence as the reverse sequence to be analyzed. Obtain the feature difference sequence of the reverse sequence to be analyzed and the weight coefficient of each feature difference in the feature difference sequence. The d-th feature difference in the feature difference sequence is the d-th change feature value in the reverse sequence to be analyzed. The result of subtracting the (d+1)th change feature value in the reverse sequence to be analyzed is given by the value. The weight coefficient of the dth feature difference in the feature difference sequence is the normalized value of the position representation value of the dth feature difference, and the position representation value of the dth feature difference is the reciprocal of d. The weighted feature difference sequence of the feature difference sequence is obtained, and the normalized value of the sum of all weighted feature differences in the weighted feature difference sequence is recorded as the data change degree representation value of the first detection point. The product of the dth feature difference and the weight coefficient of the dth feature difference is the dth weighted feature difference in the weighted feature difference sequence.

3. The method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer as described in claim 1, characterized in that, The method for obtaining the set of sequences to be analyzed at the first detection point includes: For any first detection point: the set of all detection points in the fermenter other than the first detection point is denoted as the first set, and the total number of data in the target sequence corresponding to the first detection point is denoted as the feature quantity characterization value; for the g-th detection point in the first set, the sequence constructed by the last feature quantity characterization value and pH value in the pH monitoring time series corresponding to the g-th detection point is denoted as the sequence to be analyzed corresponding to the g-th detection point; the set of the sequences to be analyzed corresponding to all detection points in the first set is denoted as the sequence to be analyzed set of the first detection point.

4. The method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer as described in claim 1, characterized in that, The method for obtaining the relevant difference characterization value of the first detection point includes: For any first detection point, a correlation characterization value sequence corresponding to the first detection point is obtained, and the mean of the correlation characterization value sequence is recorded as the correlation mean. The negative correlation mapping value of the correlation mean is recorded as the correlation difference characterization value of the first detection point. The h-th correlation characterization value in the correlation characterization value sequence is the normalized value of the Pearson correlation coefficient between the target sequence corresponding to the first detection point and the h-th sequence to be analyzed. The h-th sequence to be analyzed is the h-th sequence to be analyzed in the set of sequences to be analyzed for the first detection point.

5. The method for monitoring and optimizing fermentation conditions of Bacillus bio-organic fertilizer as described in claim 1, characterized in that, The method for obtaining the acid production characterization value of the target Bacillus at the first detection point includes: For any first detection point, the average of the data change characterization value of the first detection point and the relevant difference characterization value of the first detection point is used as the target Bacillus acid production characterization value of the first detection point.

Images