CN113139342B - Aerobic compost monitoring system and result prediction method - Google Patents
Aerobic compost monitoring system and result prediction method Download PDFInfo
- Publication number
- CN113139342B CN113139342B CN202110444505.7A CN202110444505A CN113139342B CN 113139342 B CN113139342 B CN 113139342B CN 202110444505 A CN202110444505 A CN 202110444505A CN 113139342 B CN113139342 B CN 113139342B
- Authority
- CN
- China
- Prior art keywords
- compost
- result
- data
- prediction
- composting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- C—CHEMISTRY; METALLURGY
- C05—FERTILISERS; MANUFACTURE THEREOF
- C05F—ORGANIC FERTILISERS NOT COVERED BY SUBCLASSES C05B, C05C, e.g. FERTILISERS FROM WASTE OR REFUSE
- C05F17/00—Preparation of fertilisers characterised by biological or biochemical treatment steps, e.g. composting or fermentation
- C05F17/70—Controlling the treatment in response to process parameters
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02W—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO WASTEWATER TREATMENT OR WASTE MANAGEMENT
- Y02W30/00—Technologies for solid waste management
- Y02W30/40—Bio-organic fraction processing; Production of fertilisers from the organic fraction of waste or refuse
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Chemical & Material Sciences (AREA)
- Biotechnology (AREA)
- Biochemistry (AREA)
- Chemical Kinetics & Catalysis (AREA)
- General Chemical & Material Sciences (AREA)
- Microbiology (AREA)
- Organic Chemistry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Processing Of Solid Wastes (AREA)
Abstract
The invention discloses a method for predicting an aerobic compost monitoring result, which comprises the steps of acquiring process parameter data; measuring result index data; preprocessing the data; establishing an LSTM model and carrying out model training and prediction; verifying the prediction accuracy of the prediction sequence and the original result quality sequence; testing the fitting degree of the predicted sequence and the actual sequence of the LSTM model, analyzing the composting decomposition trend and judging the optimal regulation time point according to the predicted value of the result quality data; the temperature, the humidity and the oxygen concentration in the compost bin and the pH and the conductivity in the leachate tank are monitored by adopting the sensors, monitoring media are converted, effective monitoring on compost indexes is realized, the compost indexes are displayed by a touch screen, a user can observe the situation of a compost body in the reactor in real time, historical data can be downloaded to train a prediction model, and therefore regulation and control time and regulation and control means are convenient to determine.
Description
Technical Field
The invention relates to the technical field of solid waste resource utilization, in particular to an aerobic composting monitoring system and a result prediction method.
Background
The industrial and agricultural production and daily life often face the problem that waste cannot be effectively treated or a large amount of manpower and financial resources are consumed when the waste is treated, and the organic matters of a large amount of straws and municipal wastes, leftover materials in food and beverage processing and the like produced in the agricultural planting industry are common organic waste. At present, the modes for treating the organic wastes in China mainly comprise incineration, landfill, aerobic composting and anaerobic digestion, and secondary pollution to the environment can still be caused due to improper treatment.
The composting technology is an economic and effective means for stabilizing and detoxifying organic wastes; in the existing production process, only the temperature in the material pile body can realize automatic measurement, a large number of process parameters cannot realize on-line monitoring due to the lack of corresponding medium sensors, a mechanical manual detection device is mostly adopted for measurement, the change process of various index parameters of compost cannot be effectively monitored in real time, the quality of the compost cannot be ensured, the compost cannot be completely decomposed, and agricultural and environmental problems such as seed germination, crop seedling burning, malodor and the like can be inhibited after the compost is applied to a farmland; in addition, at present, the prediction of the compost maturity degree is based on empirical judgment, and the result quality judgment has larger deviation and inconsistent product quality.
Disclosure of Invention
This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and title of the application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.
The invention is provided in view of the problems that the change process of various index parameters of compost cannot be effectively monitored in real time, the quality of the compost cannot be ensured and the compost cannot be completely decomposed in the existing coolable high ultraviolet aging test box.
Therefore, the invention aims to provide an aerobic compost monitoring system and an outcome prediction method.
In order to solve the technical problems, the invention provides the following technical scheme: a method for predicting an aerobic compost monitoring result comprises the steps of obtaining process parameter data; measuring result index data; preprocessing the data; establishing an LSTM model and carrying out model training and prediction; verifying the prediction accuracy of the prediction sequence and the original result quality sequence; and (4) checking the fitting degree of the prediction sequence and the actual sequence of the LSTM model, analyzing the compost maturity trend according to the result quality data predicted value, and judging the optimal regulation and control time point.
As a preferable aspect of the method for predicting the result of monitoring aerobic compost of the present invention, there is provided: acquiring process parameter data comprises acquiring process parameter data by using an aerobic composting reactor monitoring system, wherein the process parameter data specifically comprises material temperature, oxygen concentration content in a reactor, leachate pH, leachate conductivity and material water content (weight conversion).
As a preferable aspect of the method for predicting the result of monitoring aerobic compost of the present invention, there is provided: the measurement result index data comprises measurement result index data which is obtained by taking materials from a reactor material taking port, and specifically comprises a seed Germination Index (GI).
As a preferable aspect of the aerobic compost monitoring system and the result prediction method of the present invention, there is provided: the preprocessing of the data includes preprocessing of the process parameter data and the Germination Index (GI) of the seeds, such as format adjustment and standardization.
As a preferable aspect of the method for predicting the result of monitoring aerobic compost of the present invention, there is provided: establishing an LSTM model and carrying out model training and prediction;
s1, through a forgetting gate, an output value ht-1 at the previous moment and an input value xt at the current moment are calculated through a sigmoid activation function represented by sigma (). Where Wf is the weight matrix corresponding to ht-1 and xt, bf is the offset, and the result is ft. Outputting a number between 0 and 1 to determine the amount of information to be reserved, wherein "1" represents that the information is completely reserved, and "0" represents that the information is completely forgotten ";
f t ＝σ(w f ·[h t-1 ，x τ ]+b f )
s2: determining the information to be preserved in the neuron cell by an input gate; at the moment, the sigmoid activation function updates the neuron information to it, the tanh activation function generates a new candidate state to Ct, and W and b are corresponding weight matrix and offset respectively;
i t ＝σ(w i ·[h t-1 ，x t ]+b i )
s3: updating an old neuron state Ct-1 to a new neuron state Ct, multiplying the old state by ft, forgetting information which is determined to be forgotten before forgetting, and then increasing it x-Ct;
s4: finally, the cell state is filtered through a filter of an output gate. Firstly, a sigmoid activation function is used to obtain ot of [0,1] interval value, and then the ot is multiplied by the value of the cell state Ct processed by the tanh activation function, namely the output ht of the layer.
o t ＝σ(w o ·[h t-1 ·x t ]+b o )
h t ＝o t ×tanh(C t )
As a preferable scheme of the method for predicting the aerobic compost monitoring result, the method comprises the following steps: verifying the prediction accuracy of the prediction sequence and the original result quality sequence comprises verifying the prediction accuracy of the composting result quality prediction sequence and the original result quality sequence by using a root mean square error function and an average absolute percentage error function.
As a preferable scheme of the method for predicting the aerobic compost monitoring result, the method comprises the following steps: and (3) checking the fitting degree of the prediction sequence and the actual sequence of the LSTM model, analyzing the composting decomposition trend and judging the optimal regulation and control time point according to the result quality data prediction value, wherein when the composting result prediction shows that the compost is in an undegraded state, namely GI (GI) is lower than 90%, if the temperature of the compost is reduced or the oxygen concentration is reduced at the moment, the compost fails, and selecting to keep high temperature at the moment to promote the material decomposition, so that the material decomposition can be effectively ensured.
An aerobic compost monitoring system comprises a bearing unit and a real-time detection system, wherein the bearing unit comprises a compost bin body, and an aeration port and a feeding box which are arranged at the top of the compost bin body; and the real-time detection system comprises an oxygen content probe and a gas pressure probe which are arranged on the top of the inner wall of the composting bin body.
As a preferable mode of the aerobic compost monitoring system of the invention, wherein: the top of the inner wall of the composting bin body is provided with a spray pipe in a penetrating manner, the bottom of the composting bin body is provided with a percolate collecting box and a discharge hole, one side of the composting bin body is provided with a liquid discharge pipe, and the percolate collecting box is communicated with the liquid discharge pipe.
As a preferable mode of the aerobic compost monitoring system of the invention, wherein: the real-time detection system further comprises a PH probe and a conductivity probe which are arranged in the percolate collecting box, and a weighing device which is positioned at the bottom of the composting bin body, wherein a PLC control cabinet is arranged on one side of the weighing device, the oxygen content probe, the air pressure probe, the PH probe and the conductivity probe are all electrically connected with the PLC control cabinet, and the PLC control cabinet is electrically connected with a display screen.
The invention has the beneficial effects that: the temperature, the humidity and the oxygen concentration in the compost bin and the pH and the conductivity in the leachate tank are monitored by adopting each sensor, monitoring media are converted, effective monitoring on compost indexes is realized, the compost indexes are displayed through a touch screen, a user can observe the situation of a compost body in the reactor in real time, historical data can be downloaded to train a prediction model, and therefore regulation and control time and regulation and control means are convenient to determine; the composting process data has the characteristic of time sequence, the circular neural network is a neural network for processing sequence data, the feedback neuron is utilized to enable output data to be related to current time input data and to be influenced by last time output data, and neural network parameters are updated through data training to achieve higher prediction capability; according to the method, the characteristics are utilized, the long-short term memory neural network prediction model is trained according to the historical data of the obtained composting process data, so that the gas concentration can be predicted based on the long-short term memory neural network, the underground gas concentration can be well predicted, and a basis is provided for later-stage targeted processing and personnel work.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
FIG. 1 is a schematic diagram of the overall structure of the coolable high ultraviolet aging test chamber of the present invention.
FIG. 2 is a schematic flow chart of a method for predicting the result of the coolable high ultraviolet aging test chamber according to the present invention.
FIG. 3 is a schematic diagram of a process for establishing and training a prediction model of the coolable UV weathering test chamber according to the present invention.
FIG. 4 is a schematic diagram of the calculation steps of the coolable UV weathering chamber of the present invention, which is LSTM.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention more comprehensible, embodiments accompanying figures of the present invention are described in detail below.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Furthermore, the references herein to "one embodiment" or "an embodiment" refer to a particular feature, structure, or characteristic that may be included in at least one implementation of the present invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
Furthermore, the present invention is described in detail with reference to the drawings, and for convenience of illustration, the cross-sectional views illustrating the device structures are not enlarged partially according to the general scale when describing the embodiments of the present invention, and the drawings are only exemplary, which should not limit the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.
Example 1
Referring to fig. 2 to 4, a schematic flow chart of a method for predicting an aerobic compost monitoring result is provided, and as shown in fig. 2 to 4, the method for predicting the aerobic compost monitoring result comprises the following steps:
step 2, taking measurement result index data from a reactor material taking port, wherein the measurement result index data specifically comprises a seed germination index
Step 3, preprocessing the data in the step 1 and the step 2 such as format adjustment and standardization, so that the data conform to the input format of the neural network, and in order to eliminate dimension difference, the input and output data are subjected to unified standardization processing;
step 4, establishing an LSTM model for the data set obtained in the step 3, and carrying out model training and prediction
Step 5, verifying the prediction accuracy of the composting result quality prediction sequence and the original result quality sequence in the step 4 by using a root mean square error function and an average absolute percentage error function;
and 6, checking the fitting degree of the predicted sequence and the actual sequence of the LSTM model, analyzing the compost maturity trend and judging the optimal regulation and control time point according to the result quality data predicted value.
Step 4, referring to fig. 3, a process diagram of an aerobic composting result prediction method based on a long-short term memory neural network provided by an embodiment of the present application is shown in the figure, where the method may include the following steps:
s1: determining model training parameters:
and acquiring process parameter data by using an aerobic composting reactor monitoring system. Specifically, the temperature of the materials in the reactor, the oxygen content in the reactor, the percolation pH, the conductivity and the water content (weighing) of the materials can be obtained through a monitoring system, and historical process data can be downloaded from a server to a hard disk.
And taking measurement result index data from a reactor material taking port. Specifically, the Germination Index (GI) of the seeds can be measured by adopting a Chinese cabbage seed germination experiment.
S2: establishing a database:
preprocessing work such as format adjustment and standardization is carried out on the data in the S1;
adjusting data to a form conforming to an input format for model training, selecting n data points in batch compost data in a sliding manner, taking a process parameter as an input variable (Xm) according to a parameter (m) multiplied by a sample group number (n) form vector, taking a result index parameter as an output variable (Ym) according to a parameter (m) multiplied by a sample group number (n) form vector, and forming a batch of Xm and corresponding Ym into a batch for training; in order to facilitate calculation and model training, the data can be subjected to standardization processing after being acquired. The average value of the samples and the standard deviation of the samples are obtained in the formula, a normalized data set which is subject to normal distribution and has the average value of 0 and the standard deviation of 1 is obtained after normalization, dimensional data are converted into dimensionless data, and data processing in the later period is facilitated.
The data set is divided into 7:3, randomly dividing the training set into a training set and a testing set, inputting the training set into the long-short term memory neural network, and training the long-short term memory neural network prediction model. The LSTM network is written in the pytorech library in python, and the data format is adjusted to float _ pointer and then trained.
S3: establishing an LSTM model for the database obtained in the S2, and carrying out model training and prediction:
the Long Short-Term Memory network (LSTM) is a time-cycle neural network, gives consideration to the strong nonlinear approximation capability of the cycle neural network, and simultaneously utilizes an input gate and a forgetting gate to process Memory neurons (important information is reserved and unimportant information is forgotten), thereby overcoming the problems of gradient explosion and disappearance when the cycle neural network is used. The specific calculation process can be divided into the following four stages as shown in fig. 4:
the first stage is as follows: with the forget gate, the output value ht-1 at the previous time and the input value xt at the current time are calculated by a sigmoid activation function represented by σ (). Where Wf is the weight matrix corresponding to ht-1 and xt, bf is the offset, and the result is ft. Outputting a number between 0 and 1 to determine the amount of information to be reserved, wherein "1" represents that the information is completely reserved, and "0" represents that the information is completely forgotten ";
and a second stage: the input gate determines the information that needs to be preserved in the neuronal cells. At the moment, the sigmoid activation function updates the neuron information to it, the tanh activation function generates a new candidate state to Ct, and W and b are corresponding weight matrix and offset respectively;
and a third stage: updating an old neuron state Ct-1 to a new neuron state Ct, multiplying the old state by ft, forgetting information which is determined to be forgotten before forgetting, and then increasing it x-Ct;
a fourth stage: finally, the cell state is filtered through a filter of an output gate. Firstly, obtaining ot of a [0,1] interval value by using a sigmoid activation function, and then multiplying the ot by a value of a cell state Ct processed by a tanh activation function, namely the output ht of the layer.
Firstly, establishing a neural network, initializing the number, wherein the parameters can be divided into weight type parameters and bias type parameters; all ownership heavy parameters in the three stages are initialized by values sampled from normal distribution with a flat value of 0 and a standard deviation of 0.01, and all deviation type parameters are initialized to be 0; such initialization may prevent the neural network from falling into local minima to some extent; and setting the number of hidden layers of the neural network, the number of nodes, the model training learning rate and the iteration times.
S4: verifying the prediction accuracy of the composting result quality prediction sequence and the original result quality sequence in the S3 by using a root mean square error function and an average absolute percentage error function;
and (3) utilizing a loss function Mean Square Error (Mean Square Error MSE) to adjust the weight of the network, and utilizing a gradient descent-based method-Adam to optimize until the prediction result of the long-short term memory neural network prediction model reaches a set Error. And adjusting the neural network structure, and selecting the structure network with the highest training efficiency as a prediction model. And detecting the effect of the prediction model on the test set, and judging whether the result can be accurately predicted or not and whether an overfitting phenomenon exists or not.
S5: and (4) checking the fitting degree of the predicted sequence and the actual sequence of the LSTM model, analyzing the composting decomposition trend and judging the optimal regulation and control time point according to the result quality data predicted value.
When the compost result is predicted to show that the compost is in an undegraded state, i.e., GI is less than 90%, the compost is composted if the temperature of the compost is lowered or the oxygen concentration is lowered.
Example 2
Referring to fig. 1, a schematic flow diagram of an aerobic compost monitoring system is provided, and as shown in fig. 1, an aerobic compost monitoring system comprises: the device comprises a bearing unit 100 and a composting bin body 101 for bearing aerobic compost, an aeration port 102 for allowing air to enter the composting bin body 101, a feeding box 103 for allowing the fertilizer to be reacted to enter the composting bin body 101, a real-time detection system 200 for monitoring the aerobic compost in real time, an oxygen content probe 201 for measuring the oxygen content, an air pressure probe 202 for measuring the air pressure, a spraying pipe 101a for spraying when the water content of materials is reduced too fast, a percolate collecting box 104 for collecting percolate, a liquid discharge pipe 101b for discharging liquid, a PH probe 203 for detecting compost PH, a conductivity probe 204 for detecting the conductivity of the compost, a weigher 205 for detecting the weight of the compost, a PLC control cabinet 206 and a display screen 207.
Further, the conductivity probe 204 and the weighing device 205 are connected and communicated through a data line, and the data transmission format is 4-20ma; multiplying the pH and the conductivity of the leachate by a certain proportion to estimate the pH and the conductivity of the bulk material; the weighing device 205 records the initial mass M0 of the stack, updates the mass M every day, and calculates the water content of the stack as (M0-M)/M0 × 100%; the data obtained by the oxygen content probe 201, the air pressure probe 202, the PH probe 203 and the conductivity probe 204 are all transmitted to the PLC control cabinet 206 and displayed in real time through the display screen 207; all data can be stored in a database, the display screen 207 supports future historical data viewing, and a user can download and view historical data of all compost batches through the communication interface
Specifically, the bearing unit 100 for bearing aerobic compost comprises a composting bin body 101 for bearing the aerobic compost, and an aeration port 102 arranged at the top of the composting bin body 101 for allowing air to enter the composting bin body 101 and a feeding box 103 for allowing fertilizer to be reacted to enter the composting bin body 101; and the real-time detection system 200 for monitoring the aerobic compost in real time comprises an oxygen content probe 201 and an air pressure probe 202, wherein the oxygen content probe 201 is arranged at the top of the inner wall of the composting bin body 101 and is used for measuring the oxygen content, and the air pressure probe 202 is used for measuring the air pressure.
A spraying pipe 101a for spraying when the water content of the materials is reduced too fast penetrates through the top of the inner wall of the composting bin body 101, a percolate collecting box 104 and a discharge port 105 for collecting percolate are arranged at the bottom of the composting bin body 101, a liquid discharge pipe 101b for discharging liquid is arranged on one side of the composting bin body 101, and the percolate collecting box 104 is communicated with the liquid discharge pipe 101 b; the real-time detection system 200 further comprises a PH probe 203 for detecting the PH of the compost, a conductivity probe 204 for detecting the conductivity of the compost, and a weighing device 205 for detecting the weight of the compost, which is positioned at the bottom of the compost cabin body 101, wherein the weighing device 205 is arranged on one side of the weighing device 205, the oxygen content probe 201, the air pressure probe 202, the PH probe 203 and the conductivity probe 204 are all electrically connected with the PLC control cabinet 206, and the PLC control cabinet 206 is electrically connected with a display screen 207.
It is important to note that the construction and arrangement of the present application as shown in the various exemplary embodiments is illustrative only. Although only a few embodiments have been described in detail in this disclosure, those skilled in the art who review this disclosure will readily appreciate that many modifications are possible (e.g., variations in sizes, dimensions, structures, shapes and proportions of the various elements, values of parameters (e.g., temperatures, pressures, etc.), mounting arrangements, use of materials, colors, orientations, etc.) without materially departing from the novel teachings and advantages of the subject matter recited in this application. For example, elements shown as integrally formed may be constructed of multiple parts or elements, the position of elements may be reversed or otherwise varied, and the nature or number of discrete elements or positions may be altered or varied. Accordingly, all such modifications are intended to be included within the scope of this invention. The order or sequence of any process or method steps may be varied or re-sequenced according to alternative embodiments. In the claims, any means-plus-function clause is intended to cover the structures described herein as performing the recited function and not only structural equivalents but also equivalent structures. Other substitutions, modifications, changes and omissions may be made in the design, operating conditions and arrangement of the exemplary embodiments without departing from the scope of the present inventions. Therefore, the present invention is not limited to a particular embodiment, but extends to various modifications that nevertheless fall within the scope of the appended claims.
Moreover, in an effort to provide a concise description of the exemplary embodiments, all features of an actual implementation may not have been described (i.e., those unrelated to the presently contemplated best mode of carrying out the invention, or those unrelated to enabling the invention).
It should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort might be complex and time consuming, but would nevertheless be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill having the benefit of this disclosure, without undue experimentation.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (7)
1. A method for predicting an aerobic compost monitoring result is characterized by comprising the following steps: comprises the steps of (a) preparing a substrate,
acquiring process parameter data;
measuring result index data;
preprocessing the data;
establishing an LSTM model and performing model training and prediction including,
s1: through a forgetting gate, an output value ht-1 at the previous moment and an input value xt at the current moment are calculated through a sigmoid activation function represented by sigma (); wherein Wf is a weight matrix corresponding to ht-1 and xt, bf is an offset, and the calculation result is ft; outputting a number between 0 and 1 to determine the amount of information to be reserved, wherein "1" represents that the information is completely reserved, and "0" represents that the information is completely forgotten ";
f t ＝σ(w f ·[h t-1 ，x τ ]+b f )
s2: determining the information to be preserved in the neuronal cells by the input gate; at the moment, the sigmoid activation function updates neuron information to be it, the tanh activation function generates a new candidate state of Ct, and W and b are corresponding weight matrix and offset respectively;
i t ＝σ(w i ·[h t-1 ，x t ]+b i )
s3: updating an old neuron state Ct-1 to a new neuron state Ct, multiplying the old state by ft, forgetting information which is determined to be forgotten before forgetting, and then increasing it x-Ct;
s4: finally, determining how many cell states need to be filtered through a filter of an output gate; firstly, obtaining ot of a [0,1] interval value by using a sigmoid activation function, and then multiplying the ot by a value of a cell state Ct processed by a tanh activation function, namely the output ht of the layer;
o t ＝σ(w o ·[h t-1 ·x t ]+b o )
h t ＝o t ×tanh(C t )
verifying the prediction accuracy of the prediction sequence and the original result quality sequence comprises verifying the prediction accuracy of the compost result quality prediction sequence and the original result quality sequence by using a root mean square error function and an average absolute percentage error function;
and (3) checking the fitting degree of the prediction sequence and the actual sequence of the LSTM model, analyzing the composting decomposition trend and judging the optimal regulation and control time point according to the result quality data prediction value, wherein when the composting result prediction shows that the compost is in an undegraded state, namely GI (GI) is lower than 90%, if the temperature of the compost is reduced or the oxygen concentration is reduced at the moment, the compost fails, and selecting to keep high temperature at the moment to promote the material decomposition, so that the material decomposition can be effectively ensured.
2. The method for predicting an aerobic compost monitoring result as set forth in claim 1, wherein: acquiring process parameter data comprises acquiring process parameter data by using an aerobic composting reactor monitoring system, wherein the process parameter data specifically comprises material temperature, oxygen concentration content in a reactor, leachate pH, leachate conductivity and material water content.
3. The method for predicting the result of monitoring aerobic compost according to claim 2, wherein: the measurement result index data comprises measurement result index data, specifically comprising a seed Germination Index (GI), which is obtained by taking materials from a reactor material taking port.
4. The method for predicting the result of monitoring aerobic compost according to claim 3, wherein: the preprocessing of the data includes preprocessing of the process parameter data and the Germination Index (GI) of the seeds, such as format adjustment and standardization.
5. An aerobic composting monitoring system for use in a method according to claims 1-4, characterised in that: comprises the steps of (a) preparing a substrate,
the loading unit (100) comprises a composting bin body (101), and an aeration port (102) and a feeding box (103) which are arranged at the top of the composting bin body (101); and (c) a second step of,
the real-time detection system (200) comprises an oxygen content probe (201) and an air pressure probe (202) which are arranged on the top of the inner wall of the composting bin body (101).
6. The aerobic compost monitoring system of claim 5, wherein: the top of the inner wall of the composting bin body (101) is provided with a spray pipe (101 a) in a penetrating manner, the bottom of the composting bin body (101) is provided with a percolate collecting box (104) and a discharge hole (105), one side of the composting bin body (101) is provided with a liquid discharge pipe (101 b), the other side of the composting bin body (101) is provided with a thermometer (101 c), and the percolate collecting box (104) is communicated with the liquid discharge pipe (101 b).
7. The aerobic compost monitoring system of claim 6, wherein: the real-time detection system (200) further comprises a PH probe (203) and a conductivity probe (204) which are arranged in the percolate collecting box (104), and a weighing device (205) which is positioned at the bottom of the composting cabin body (101), wherein a PLC (programmable logic controller) control cabinet (206) is arranged on one side of the weighing device (205), the oxygen content probe (201), the air pressure probe (202), the PH probe (203) and the conductivity probe (204) are all electrically connected with the PLC control cabinet (206), and the PLC control cabinet (206) is electrically connected with a display screen (207).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110444505.7A CN113139342B (en) | 2021-04-23 | 2021-04-23 | Aerobic compost monitoring system and result prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110444505.7A CN113139342B (en) | 2021-04-23 | 2021-04-23 | Aerobic compost monitoring system and result prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113139342A CN113139342A (en) | 2021-07-20 |
| CN113139342B true CN113139342B (en) | 2022-12-27 |
Family
ID=76811857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110444505.7A Active CN113139342B (en) | 2021-04-23 | 2021-04-23 | Aerobic compost monitoring system and result prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113139342B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113793645A (en) * | 2021-09-16 | 2021-12-14 | 浙江大学 | Compost maturity prediction method based on machine learning model |
| CN113805627A (en) * | 2021-09-24 | 2021-12-17 | 领先生物农业股份有限公司 | Nano-film aerobic composting Internet of things control method and device |
| CN114048897A (en) * | 2021-10-29 | 2022-02-15 | 西藏电建成勘院工程有限公司 | Method for constructing straw compost organic matter content prediction model based on temperature and humidity conditions |
| CN116199532A (en) * | 2023-04-10 | 2023-06-02 | 常熟理工学院 | Method for promoting kitchen waste composting |
| CN116659589B (en) * | 2023-07-25 | 2023-10-27 | 澳润(山东)药业有限公司 | Donkey-hide gelatin cake preservation environment monitoring method based on data analysis |
| CN118151614A (en) * | 2024-05-08 | 2024-06-07 | 镇江贝思特有机活性肥料有限公司 | Control method and control system for adjusting organic material fermentation process by septic tank water |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102603382A (en) * | 2012-03-16 | 2012-07-25 | 清华大学 | Organic waste aerobic composting test device |
| CN202400981U (en) * | 2011-12-23 | 2012-08-29 | 北京科技大学 | Human feces aerobic composting device |
| CN108682006A (en) * | 2018-04-25 | 2018-10-19 | 南京农业大学 | Contactless canned compost maturity judgment method |
| CN111056865A (en) * | 2019-12-24 | 2020-04-24 | 石家庄佳盛生物科技有限公司 | System and method for degrading household garbage |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210355045A1 (en) * | 2020-05-18 | 2021-11-18 | Brad Morehead | Smart composting apparatus |
-
2021
- 2021-04-23 CN CN202110444505.7A patent/CN113139342B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202400981U (en) * | 2011-12-23 | 2012-08-29 | 北京科技大学 | Human feces aerobic composting device |
| CN102603382A (en) * | 2012-03-16 | 2012-07-25 | 清华大学 | Organic waste aerobic composting test device |
| CN108682006A (en) * | 2018-04-25 | 2018-10-19 | 南京农业大学 | Contactless canned compost maturity judgment method |
| CN111056865A (en) * | 2019-12-24 | 2020-04-24 | 石家庄佳盛生物科技有限公司 | System and method for degrading household garbage |
Non-Patent Citations (2)
| Title |
|---|
| 基于BP神经网络的堆肥物料抗剪强度预测模型;李成杰等;《中国给水排水》;20200401(第07期);全文 * |
| 沼渣好氧堆肥种子发芽指数快速预测可行性分析;黄光群等;《农业机械学报》;20160531(第05期);正文第177-181页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113139342A (en) | 2021-07-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113139342B (en) | Aerobic compost monitoring system and result prediction method | |
| Bhagat et al. | Heavy metal contamination prediction using ensemble model: Case study of Bay sedimentation, Australia | |
| Stenberg | Monitoring soil quality of arable land: microbiological indicators | |
| Lin et al. | Modeling of substrate degradation and oxygen consumption in waste composting processes | |
| Smith et al. | Soil biota and global change at the ecosystem level: describing soil biota in mathematical models | |
| Korsaeth et al. | Modelling the competition for nitrogen between plants and microflora as a function of soil heterogeneity | |
| Ebrahimi et al. | Comparison of artificial neural network and multivariate regression models for prediction of Azotobacteria population in soil under different land uses | |
| CN101370926A (en) | Method for on-line future performance estimation of fermentation apparatus | |
| Gabrielle et al. | Simulation of carbon and nitrogen dynamics in arable soils: a comparison of approaches | |
| Esbensen et al. | Representative process sampling—in practice: Variographic analysis and estimation of total sampling errors (TSE) | |
| Nair et al. | Real-time monitoring of enhanced biological phosphorus removal in a multistage EBPR-MBBR using a soft-sensor for phosphates | |
| Wan et al. | Machine learning framework for intelligent prediction of compost maturity towards automation of food waste composting system | |
| Han et al. | Prediction of sludge bulking using the knowledge-leverage-based fuzzy neural network | |
| Cai et al. | [Retracted] Simulation and Prediction of Fungal Community Evolution Based on RBF Neural Network | |
| Soto-Paz et al. | A new approach for the optimization of biowaste composting using artificial neural networks and particle swarm optimization | |
| Tan et al. | Resolving stability issue of thermophilic high-rate anaerobic palm oil mill effluent treatment via adaptive neuro-fuzzy inference system predictive model | |
| Yu et al. | Comparison of different machine learning algorithms to estimate liquid level for bioreactor management | |
| Li et al. | Prediction of composting maturity and identification of critical parameters for green waste compost using machine learning | |
| Dong et al. | Virtual sample generation empowers machine learning-based effluent prediction in constructed wetlands | |
| CN113793645A (en) | Compost maturity prediction method based on machine learning model | |
| Xi et al. | Characterizing effects of uncertainties in MSW composting process through a coupled fuzzy vertex and factorial-analysis approach | |
| Gasser et al. | Transfer of carbon incubation parameters to model the SOC and SON dynamics of a field trial with energy crops applying digestates as organic fertilizers | |
| Bakht et al. | Ingredient analysis of biological wastewater using hybrid multi-stream deep learning framework | |
| CN115959933A (en) | Intelligent control method and system for aerobic composting | |
| Irizar et al. | Model-based design of a software sensor for real-time diagnosis of the stability conditions in high-rate anaerobic reactors–Full-scale application to Internal Circulation technology |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |