CN112765900A - Alcohol content online detection method and system for segmented liquor picking - Google Patents
Alcohol content online detection method and system for segmented liquor picking Download PDFInfo
- Publication number
- CN112765900A CN112765900A CN202110148060.8A CN202110148060A CN112765900A CN 112765900 A CN112765900 A CN 112765900A CN 202110148060 A CN202110148060 A CN 202110148060A CN 112765900 A CN112765900 A CN 112765900A
- Authority
- CN
- China
- Prior art keywords
- pipeline
- layer
- alcoholic strength
- value
- real
- 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.)
- Granted
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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K7/00—Measuring temperature based on the use of electric or magnetic elements directly sensitive to heat ; Power supply therefor, e.g. using thermoelectric elements
- G01K7/02—Measuring temperature based on the use of electric or magnetic elements directly sensitive to heat ; Power supply therefor, e.g. using thermoelectric elements using thermoelectric elements, e.g. thermocouples
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N25/00—Investigating or analyzing materials by the use of thermal means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N9/00—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity
- G01N9/26—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity by measuring pressure differences
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N9/00—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity
- G01N9/36—Analysing materials by measuring the density or specific gravity, e.g. determining quantity of moisture
-
- 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
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Pathology (AREA)
- Analytical Chemistry (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Investigating Or Analyzing Materials Using Thermal Means (AREA)
Abstract
The invention discloses an alcohol content online detection method and system for sectional liquor picking. The method comprises the following steps: building an alcoholic strength online detection system for segmented liquor picking; collecting a real-time density value and a real-time detection temperature value; and converting the real-time density value and the real-time temperature value into a real-time alcohol degree value of the base wine to be detected through an alcohol degree calculation model in the controller. The system comprises a density sensor, a damping device, a wine receiving device, a power device, a flow stabilizing device, a temperature sensor, a controller, an exhaust device, a wine flowing pipeline, a flow stabilizing pipeline and an alcohol degree measuring pipeline. The method can realize real-time online detection of the alcohol content, effectively improve the speed, timeliness and detection precision of the alcohol content detection, greatly save the production time, improve the production efficiency of a liquor production line, reduce the waste of human resources, avoid various adverse effects caused by human factors, and provide timely and accurate index data for production.
Description
Technical Field
The invention belongs to the technical field of wine brewing, and particularly relates to an alcohol content online detection method and system for segmented wine storage.
Background
In the current wine making industry, methods of wine extraction by sections, wine extraction by looking at flowers and the like are common in the production process of distilled liquor, and wine extraction is carried out according to the gradient change of the alcoholic strength, so that the real-time detection of the alcoholic strength is very important work, the alcoholic strength of distillate is accurately detected, and the production efficiency is improved to a great extent.
In the prior art, the commonly used alcohol degree detection method for the white spirit production line is a measuring method of 'looking at flowers' or adopting a combination of an alcohol meter and a thermometer and the like. The method needs to cultivate professional staff, and the working skill is difficult to grasp, the alcohol degree judgment standards are different, and a large difference exists. The measuring method of the measuring tool is simple to operate, but real-time detection cannot be carried out, the reading of the measuring tool is easy to generate large errors due to different reading personnel, the reading result is greatly influenced by human factors, and the environment temperature of the liquor production workshop is unstable, so that the detection result is influenced. The application number 201110089637.9 discloses a method for extracting strong aromatic Chinese spirits, which realizes the extraction of the spirits by sections by observing the sizes of hops through controlling the water content and the gas phase temperature. However, the traditional liquor picking method for picking liquor by watching flowers is still adopted in essence, and the alcohol picking degree is still influenced by human errors and environmental factors. The document with the application number of 201910418319.9 discloses a liquor picking method based on near infrared spectrum, which is characterized in that a near infrared model is established by collecting liquor near infrared spectrum data, liquor alcoholic strength is judged, and accordingly liquor is picked in sections.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to solve the technical problem of providing an online detection method and system for alcohol content of sectional liquor picking.
The technical scheme for solving the technical problem of the method is to provide the online detection method for the alcoholic strength of the wine picked in sections, which is characterized by comprising the following steps of:
firstly, building an alcohol content online detection system for segmented liquor picking: the system comprises a density sensor, a damping device, a power device, a flow stabilizing device, a temperature sensor, a controller, a flow stabilizing pipeline and an alcohol degree measuring pipeline;
one end of the steady flow pipeline is a flowing wine inlet, and the other end of the steady flow pipeline is connected with the inlet end of the alcoholic strength measuring pipeline; a flow stabilizing device is arranged on the flow stabilizing pipeline; the alcoholic strength measuring pipeline is vertically arranged, and a damping device and a density sensor are sequentially arranged on the alcoholic strength measuring pipeline according to the flowing direction of liquid; the temperature sensor is arranged on the alcoholic strength measuring pipeline and is tightly attached to the density sensor; the outlet end of the alcoholic strength measuring pipeline is externally connected with an outflow pipeline; the controller is respectively in communication connection with the power device, the density sensor and the temperature sensor and realizes control;
the flow stabilizing device is used for changing the flowing wine into a steady flow state; the damping device is used for reducing the influence of working vibration of the power device on the density sensor;
step two, data acquisition: the system starts to operate, and the raw wine conveyed by the power device flows to the alcohol degree measuring pipeline through the flow stabilizing pipeline and the damping device; the density sensor collects a real-time density value rho, the temperature sensors collect corresponding number of real-time detection temperature values, and electric signals of the real-time density value rho and the real-time detection temperature values are transmitted to the controller;
thirdly, calculating the alcoholic strength value: the controller calculates the average value of a plurality of real-time detection temperature values as a real-time temperature value T, and converts the real-time density value rho and the real-time temperature value T into a real-time alcohol degree value q of the base wine to be detected through an alcohol degree calculation model in the controller;
the alcohol degree calculation model is as follows:
wherein the content of the first and second substances,weight constant V1=-8.78009788,V2=-120.34663676,V3=-11.92250949,V4=5.36664326,V57.37168811, W11 ═ 0.35875759, W12 ═ 0.13787137, W13 ═ 0.33458189, W14 ═ 0.50523374, W15 ═ 1.69823758, W21 ═ 10.29658039, W22 ═ 0.80742288, W23 ═ 4.12574959, W24 ═ 11.19394148, W25 ═ 8.05011810; bias constant b1=-11.35753556,b2=0.04014640,b3=-2.66309245,b4=9.68923738,b5=12.42979910,b=105.67968188。
The invention provides a system for executing the alcohol content online detection method for sectional liquor picking, which is characterized by comprising a density sensor, a damping device, a liquor receiving device, a power device, a flow stabilizing device, a temperature sensor, a controller, an exhaust device, a liquor flowing pipeline, a flow stabilizing pipeline and an alcohol content measuring pipeline;
one end of the wine flowing pipeline is connected with the tail end of the wine receiving device, and the other end of the wine flowing pipeline is connected with one end of the flow stabilizing pipeline; the wine flowing pipeline is provided with a power device and an exhaust device; the other end of the flow stabilizing pipeline is connected with the inlet end of the alcoholic strength measuring pipeline; a flow stabilizing device is arranged on the flow stabilizing pipeline; the alcoholic strength measuring pipeline is vertically arranged, and a damping device and a density sensor are sequentially arranged on the alcoholic strength measuring pipeline according to the flowing direction of liquid; the temperature sensor is arranged on the alcoholic strength measuring pipeline and is tightly attached to the density sensor; the outlet end of the alcoholic strength measuring pipeline is externally connected with an outflow pipeline; and the controller is respectively in communication connection with the power device, the density sensor and the temperature sensor and realizes control.
Compared with the prior art, the invention has the beneficial effects that:
(1) the method can realize real-time online detection of the alcohol content, effectively improve the speed, timeliness and detection precision of the alcohol content detection, greatly save the production time, overcome the key difficult problem of the production line, improve the production efficiency of the liquor production line, reduce the waste of human resources, avoid various adverse effects caused by human factors, provide timely and accurate index data for production, provide a theoretical basis for the production of high-quality liquor, and can be applied to the segmented liquor picking link in the liquor production industry.
(2) According to the method, the alcohol accuracy calculation model is fitted through the neural network, and the density and the temperature value which are measured in real time are input into the alcohol degree calculation model to obtain the real-time alcohol degree value, so that the real-time detection of the alcohol degree is indirectly realized. The higher the detection precision of the temperature value and the density value is, the higher the detection precision of the alcoholic strength is, the detection precision of the alcoholic strength is improved to a certain extent, and +/-0.5% vol can be realized.
Drawings
FIG. 1 is a schematic diagram of a system according to an embodiment of the present invention;
in the figure: 1. a density sensor; 2. a damping device; 3. a wine receiving device; 4. a power plant; 5. a pulsation damper; 6. a temperature sensor; 7. a back pressure valve; 8. an exhaust device; 9. a liquor flowing pipeline; 10. a flow stabilizing pipeline; 11. alcohol degree measuring pipeline.
Detailed Description
Specific examples of the present invention are given below. The specific examples are only intended to illustrate the invention in further detail and do not limit the scope of protection of the claims of the present application.
The invention provides an alcohol content online detection method (short method) for segmented liquor picking, which is characterized by comprising the following steps of:
firstly, building an alcohol content online detection system (system for short) for segmented liquor picking:
the system comprises a density sensor 1, a damping device 2, a wine receiving device 3, a power device 4, a flow stabilizer, a temperature sensor 6, a controller, an exhaust device 8, a wine flowing pipeline 9, a flow stabilizing pipeline 10 and an alcoholic strength measuring pipeline 11;
one end of the wine flowing pipeline 9 is connected with the tail end of the wine receiving device 3, and the other end of the wine flowing pipeline is connected with one end of the steady flow pipeline 10; the wine flowing pipeline 9 is provided with a power device 4 and an exhaust device 8; the other end of the flow stabilizing pipeline 10 is connected with the inlet end of the alcoholic strength measuring pipeline 11; a flow stabilizing device is arranged on the flow stabilizing pipeline 10; the alcoholic strength measuring pipeline 11 is vertically arranged, and a damping device 2 and a density sensor 1 are sequentially arranged on the alcoholic strength measuring pipeline according to the flowing direction of liquid; the temperature sensor 6 is arranged on the alcoholic strength measuring pipeline 11 and is tightly attached to the density sensor 1; the outlet end of the alcoholic strength measuring pipeline 11 is externally connected with an outflow pipeline; the controller is respectively in communication connection with the power device 4, the density sensor 1 and the temperature sensor 6 and realizes control.
Preferably, the wine receiving device 3 is of a funnel-shaped structure and is used for receiving the wine base flowing out from a condenser of the wine production equipment.
Preferably, the exhaust device 8 is used for exhausting air in the pipeline in the wine discharging process, the accuracy of density measurement is improved, and an exhaust pipe or an exhaust hole can be adopted; when the exhaust pipe is adopted, the position of the gas outlet of the exhaust pipe is higher than the position of the liquid inlet of the power device 4, so that gas is discharged conveniently, liquid is prevented from flowing out, and the gas outlet is provided with a net structure; when the exhaust hole is adopted, the position of the exhaust hole is higher than the position of a liquid inlet of the power device 4, so that the exhaust of gas is facilitated, meanwhile, liquid is prevented from flowing out, and a net-shaped structure is arranged at the exhaust hole.
Preferably, the power device 4 adopts a diaphragm metering pump and is used for accurately controlling the raw wine flow speed and the raw wine flow rate of sectional wine-picking equipment of the white wine production line.
Preferably, the flow stabilizer is used for changing the raw wine flowing out of the power device 4 into a steady flow state and improving the accuracy of density measurement, and comprises a pulsation damper 5 with a pressure gauge and a back pressure valve 7; the pulsation damper 5 and the back pressure valve 7 are sequentially disposed on the flow stabilizing pipe 10 in a liquid flowing direction.
Preferably, the steady flow pipeline 10 is horizontally arranged, the backpressure valve 7 is horizontally arranged on the steady flow pipeline 10, and the pulsation damper 5 is vertically arranged on the steady flow pipeline 10, so that part of air in the pipeline enters a cavity of the pulsation damper 5 to be squeezed and buffered.
Preferably, the damping device 2 is arranged at the inlet end of the alcoholic strength measuring pipeline 11, and is used for reducing the influence of the power device 4 on the density sensor 1 due to working vibration and improving the accuracy of density measurement of the density sensor 1. The damping device 2 adopts a food-grade flexible connection transparent pipeline; the other end of the steady flow pipeline 10 is connected with one end of a food-grade flexible connection transparent pipeline; the other end of the food-grade flexible connection transparent pipeline is connected with the inlet end of the alcoholic strength measuring pipeline 11.
Preferably, the density sensor 1 is a differential pressure density sensor, and is divided into two measuring heads, and the two measuring heads are connected to the alcoholic strength measuring pipeline 11 at intervals through flanges, and the vertical distance between the two measuring heads is H. According to the principle of fluid statics, in a liquid with constant flow rate, the pressure difference between two vertical points is a certain value. The raw wine to be measured forms a pressure difference delta P by two measuring heads with a vertical distance of H, and the pressure values measured at the two measuring heads are P respectively1And P2,Δp=P1-P2The density of the wine base to be measured in the pipeline with the vertical distance H can be obtained according to the hydrostatic principle
The temperature sensors 6 are at least one thermocouple temperature sensor, the number of the thermocouple temperature sensors is at least one, the thermocouple temperature sensors are uniformly distributed on the alcoholic strength measuring pipeline 11, one temperature sensor 6 is tightly attached to the rear part of a measuring head of the density sensor 1, the measuring head is positioned on the rear side according to the liquid flowing direction, and the white spirit temperature value is the average value measured by the temperature sensors 6; when one temperature sensor 6 is used, the temperature sensor 6 is attached to one of the two measuring heads of the density sensor 1 according to the liquid flow direction (preferably, the probe of the temperature sensor 6 is attached to the rear of the measuring head of the density sensor 1 which is located on the rear side according to the liquid flow direction, considering that the liquid pressure may be affected by the probe, and thus the flow stabilizing effect which has been achieved is not affected); when two temperature sensors 6 are adopted, the two temperature sensors 6 are respectively clung to two measuring heads of the density sensor 1, wherein one temperature sensor 6 is clung to the rear part of the measuring head of the density sensor 1 which is positioned at the rear side according to the flowing direction of the liquid; when three temperature sensors 6 are used, two of the temperature sensors 6 are respectively attached to two measuring heads of the density sensor 1, one of the temperature sensors 6 is attached to the density sensor 1 behind the measuring head located behind the density sensor 1 in the direction of liquid flow, and the other temperature sensor 6 is located at the midpoint of the two temperature sensors 6 or the midpoint of the two measuring heads of the density sensor 1.
Preferably, the controller is a PLC or a single chip microcomputer.
Step two, data acquisition: the system starts to operate, the raw wine flows into the power device 4 through the wine receiving device 3 and the wine flowing pipeline 9, and the raw wine conveyed by the power device 4 flows to the alcoholic strength measuring pipeline 11 through the flow stabilizing pipeline 10 and the damping device 2; the density sensor 1 collects a real-time density value rho, the temperature sensors 6 collect corresponding number of real-time detection temperature values, and electric signals of the real-time density value rho and the real-time detection temperature values are transmitted to the controller;
thirdly, calculating the alcoholic strength value: the controller calculates the average value of a plurality of real-time detection temperature values as a real-time temperature value T, and converts the real-time density value rho and the real-time temperature value T into a real-time alcohol degree value q of the base wine to be detected through an alcohol degree calculation model in the controller;
the alcohol degree calculation model is as follows:
wherein the content of the first and second substances,weight constant V1=-8.78009788,V2=-120.34663676,V3=-11.92250949,V4=5.36664326,V57.37168811, W11 ═ 0.35875759, W12 ═ 0.13787137, W13 ═ 0.33458189, W14 ═ 0.50523374, W15 ═ 1.69823758, W21 ═ 10.29658039, W22 ═ 0.80742288, W23 ═ 4.12574959, W24 ═ 11.19394148, W25 ═ 8.05011810; bias constant b1=-11.35753556,b2=0.04014640,b3=-2.66309245,b4=9.68923738,b5=12.42979910,b=105.67968188。
Preferably, in the third step, the obtaining process of the alcohol degree calculation model is as follows:
step 1, data acquisition: collecting a plurality of pieces of data (4141 pieces in the embodiment) from a field or a database as training samples, wherein each piece of data comprises a density value, a temperature value and a corresponding alcohol degree value, the temperature value collection precision is 0.1, the density value collection precision is 0.001, and the alcohol degree value collection precision is 0.5; the alcohol degree difference value between the collected groups of data is a fixed value, and the alcohol degree value is in an arithmetic sequence;
step 2, data preprocessing: carrying out standardization processing on the density value and the temperature value in the training sample to obtain an input variable;
input variable X of BP neural networkk=[xkd,xkt](k-1, 2, … N), wherein XkInput variable, x, representing the k-th training samplekdDenotes a density value variable, x, among input variables of the k-th training samplektRepresenting a temperature value variable in input variables of a kth training sample, wherein N represents the number of input training samples; y iskPredicting the alcohol accuracy value of the actual output value of the kth training sample; dkRepresenting the expected output value of the kth training sample, namely the actual alcoholic strength value, namely the label;
the topological structure of the BP neural network comprises an input layer, a hidden layer and an output layer; the input variables are two variables of temperature value and density value, so the number of neurons of the input layer is 2; by selecting the number of different hidden layer units for comparison, the number of hidden layer units is 5, the error is small, and the model is good, so that the number of neurons of the hidden layer is 5; the output is a variable of the alcoholic strength value, so the number of neurons of an output layer is 1; i.e. a BP neural network of 2-5-1 topology is established.
The basic parameters comprise a learning rate mu, an iteration number m, the weight from an input layer to an implicit layer, the weight from the implicit layer to an output layer, the bias from the input layer to the implicit layer, the bias from the implicit layer to the output layer and an allowable error epsilon;
the transfer functions include a hidden layer transfer function f (z) and an output layer transfer function; the above-mentionedz is a hidden layer input; the output layer transfer function is a linear transfer function;
step 4, initializing basic parameters of the BP neural network: the learning rate mu is 0.01, the iteration number m is 1000, the initial values of the weight from the input layer to the hidden layer, the weight from the hidden layer to the output layer, the bias from the input layer to the hidden layer and the bias from the hidden layer to the output layer are all random numbers within (-1,1), and the allowable error epsilon is 0.005;
step 5, training a BP neural network: the input variable set X of the training sample is equal to [ X ═ X1,X2,…XN]And a desired output set D ═ D1,D2,…DN]Sequentially inputting the data into a BP neural network, carrying out nonlinear fitting on the BP neural network by using MATLAB software to obtain an alcoholic strength calculation model with independent variable of temperature and density and dependent variable of alcoholic strength, and inputting the alcoholic strength calculation model into a controller;
preferably, in step 5, the performing nonlinear fitting on the BP neural network specifically includes: in the training sample, the alcoholic strength range is 0-100%, the temperature range is 0-40 degrees, and the density value is the corresponding density value; performing no more than m iterations from the input layer to the output layer for each training sampleCalculating the input and output of each layer of neuron of BP neural network layer by layer and calculating the output error E of the kth training samplekAnd the total output error E of all training samples; wherein
When the E is smaller than the allowable error epsilon or reaches the specified iteration number m, finishing the fitting training to obtain an alcoholic strength calculation model; if E is larger than the allowable error epsilon and does not reach the specified iteration number m, reversely calculating the output error E of each layer of neuron layer by layer from the output layer to the input layerkAnd then, adjusting each weight value and bias value of the BP neural network according to an error gradient descent method until the condition that E is smaller than an allowable error epsilon or reaches a specified iteration number m is met, enabling the final actual output of the adjusted BP neural network to be close to the expected output, and finishing the fitting training.
Nothing in this specification is said to apply to the prior art.
Claims (10)
1. An alcohol content online detection method for sectional liquor picking is characterized by comprising the following steps:
firstly, building an alcohol content online detection system for segmented liquor picking: the system comprises a density sensor, a damping device, a power device, a flow stabilizing device, a temperature sensor, a controller, a flow stabilizing pipeline and an alcohol degree measuring pipeline;
one end of the steady flow pipeline is a flowing wine inlet, and the other end of the steady flow pipeline is connected with the inlet end of the alcoholic strength measuring pipeline; a flow stabilizing device is arranged on the flow stabilizing pipeline; the alcoholic strength measuring pipeline is vertically arranged, and a damping device and a density sensor are sequentially arranged on the alcoholic strength measuring pipeline according to the flowing direction of liquid; the temperature sensor is arranged on the alcoholic strength measuring pipeline and is tightly attached to the density sensor; the outlet end of the alcoholic strength measuring pipeline is externally connected with an outflow pipeline; the controller is respectively in communication connection with the power device, the density sensor and the temperature sensor and realizes control;
the flow stabilizing device is used for changing the flowing wine into a steady flow state; the damping device is used for reducing the influence of working vibration of the power device on the density sensor;
step two, data acquisition: the system starts to operate, and the raw wine conveyed by the power device flows to the alcohol degree measuring pipeline through the flow stabilizing pipeline and the damping device; the density sensor collects a real-time density value rho, the temperature sensors collect corresponding number of real-time detection temperature values, and electric signals of the real-time density value rho and the real-time detection temperature values are transmitted to the controller;
thirdly, calculating the alcoholic strength value: the controller calculates the average value of a plurality of real-time detection temperature values as a real-time temperature value T, and converts the real-time density value rho and the real-time temperature value T into a real-time alcohol degree value q of the base wine to be detected through an alcohol degree calculation model in the controller;
the alcohol degree calculation model is as follows:
wherein the content of the first and second substances,weight constant V1=-8.78009788,V2=-120.34663676,V3=-11.92250949,V4=5.36664326,V57.37168811, W11 ═ 0.35875759, W12 ═ 0.13787137, W13 ═ 0.33458189, W14 ═ 0.50523374, W15 ═ 1.69823758, W21 ═ 10.29658039, W22 ═ 0.80742288, W23 ═ 4.12574959, W24 ═ 11.19394148, W25 ═ 8.05011810; bias constant b1=-11.35753556,b2=0.04014640,b3=-2.66309245,b4=9.68923738,b5=12.42979910,b=105.67968188。
2. The online alcoholic strength detection method for segmented liquor taking as claimed in claim 1, wherein in the third step, the alcoholic strength calculation model is obtained through a process of:
step 1, data acquisition: collecting a plurality of pieces of data as training samples, wherein each piece of data comprises a density value, a temperature value and a corresponding alcohol accuracy value;
step 2, data preprocessing: carrying out standardization processing on the density value and the temperature value in the training sample to obtain an input variable;
step 3, establishing a BP neural network: determining variables, topological structures, basic parameters and transfer functions of the BP neural network;
step 4, initializing basic parameters of the BP neural network;
step 5, training a BP neural network: the input variable set X of the training sample is equal to [ X ═ X1,X2,…XN]And a desired output set D ═ D1,D2,…DN]And sequentially inputting the data into a BP neural network, carrying out nonlinear fitting on the BP neural network to obtain an alcoholic strength calculation model with independent variable of temperature and density and dependent variable of alcoholic strength, and inputting the alcoholic strength calculation model into a controller.
3. The online alcoholic strength detection method for segmented liquor taking according to claim 2, wherein in the step 1, the temperature value collection precision is 0.1, the density value collection precision is 0.001, and the alcoholic strength value collection precision is 0.5; the alcohol degree difference value between the collected data is a fixed value.
4. The online alcoholic strength detection method for segmented liquor taking as claimed in claim 2, wherein in the step 3, an input variable X of a BP neural networkk=[xkd,xkt](k-1, 2, … N), wherein XkInput variable, x, representing the k-th training samplekdDenotes a density value variable, x, among input variables of the k-th training samplektRepresenting a temperature value variable in input variables of a kth training sample, wherein N represents the number of input training samples; y iskThe actual output value of the kth training sample; dkRepresenting an expected output value of the kth training sample;
the topological structure of the BP neural network comprises an input layer, a hidden layer and an output layer; the number of the neurons of the input layer is 2, the number of the neurons of the hidden layer is 5, and the number of the neurons of the output layer is 1;
the basic parameters comprise a learning rate mu, an iteration number m, the weight from an input layer to an implicit layer, the weight from the implicit layer to an output layer, the bias from the input layer to the implicit layer, the bias from the implicit layer to the output layer and an allowable error epsilon;
5. The online alcoholic strength detection method for segmented liquor taking according to claim 4, wherein the step 4 specifically comprises: the learning rate μ is 0.01, the number of iterations m is 1000, the initialized values of the weights from the input layer to the hidden layer, the weights from the hidden layer to the output layer, the bias from the input layer to the hidden layer and the bias from the hidden layer to the output layer are all random numbers within (-1,1), and the allowable error ∈ is 0.005.
6. The online alcoholic strength detection method for segmented liquor taking according to claim 4, wherein in the step 5, the nonlinear fitting of the BP neural network specifically comprises: performing no more than m iterations on each training sample, wherein the iteration is to forward calculate the input and output of each layer of neuron of the BP neural network layer by layer from an input layer to an output layer, and calculate the output error E of the kth training samplekAnd the total output error E of all training samples; wherein
When the E is smaller than the allowable error epsilon or reaches the specified iteration number m, finishing the fitting training to obtain an alcoholic strength calculation model; if E is larger than the allowable error epsilon and does not reach the specified iteration number m, the output layer outputs to the inputIn-layer and layer-by-layer reverse calculation of output error E of neuron in each layerkAnd then, adjusting each weight value and bias value of the BP neural network according to an error gradient descent method until the condition that E is smaller than an allowable error epsilon or reaches a specified iteration number m is met, enabling the final actual output of the adjusted BP neural network to be close to the expected output, and finishing the fitting training.
7. A system for executing the online alcoholic strength detection method for the segmented liquor extraction of any one of claims 1 to 6, wherein the system comprises a density sensor, a damping device, a liquor receiving device, a power device, a flow stabilizing device, a temperature sensor, a controller, an exhaust device, a liquor flowing pipeline, a flow stabilizing pipeline and an alcoholic strength measuring pipeline;
one end of the wine flowing pipeline is connected with the tail end of the wine receiving device, and the other end of the wine flowing pipeline is connected with one end of the flow stabilizing pipeline; the wine flowing pipeline is provided with a power device and an exhaust device; the other end of the flow stabilizing pipeline is connected with the inlet end of the alcoholic strength measuring pipeline; a flow stabilizing device is arranged on the flow stabilizing pipeline; the alcoholic strength measuring pipeline is vertically arranged, and a damping device and a density sensor are sequentially arranged on the alcoholic strength measuring pipeline according to the flowing direction of liquid; the temperature sensor is arranged on the alcoholic strength measuring pipeline and is tightly attached to the density sensor; the outlet end of the alcoholic strength measuring pipeline is externally connected with an outflow pipeline; and the controller is respectively in communication connection with the power device, the density sensor and the temperature sensor and realizes control.
8. The system for implementing the alcohol degree on-line detection method for the segmented liquor picking according to the claim 7, wherein the exhaust device is used for exhausting air in the pipeline in the liquor picking process, the accuracy of density measurement is improved, and an exhaust pipe or an exhaust hole is adopted;
the flow stabilizer comprises a pulsation damper and a back pressure valve; the pulsation damper and the back pressure valve are sequentially arranged on the flow stabilizing pipeline according to the flowing direction of the liquid; the steady flow pipeline is horizontally arranged, the back pressure valve is horizontally arranged on the steady flow pipeline, and the pulsation damper is vertically arranged on the steady flow pipeline;
the damping device adopts a food-grade flexible connection transparent pipeline; the other end of the steady flow pipeline is connected with one end of the food-grade flexible connection transparent pipeline; the other end of the food-grade flexible connection transparent pipeline is connected with the inlet end of the alcohol degree measuring pipeline.
9. The system for performing the alcohol content on-line measuring method for the sectional liquor picking of the claim 7, wherein the density sensor is a differential pressure density sensor which is divided into two measuring heads and is connected to the alcohol content measuring pipeline at intervals through a flange, and the vertical distance between the two measuring heads is H; the raw wine to be measured forms a pressure difference delta p by two measuring heads with a vertical distance H, and the density of the raw wine to be measured in the pipeline with the vertical distance H can be obtained according to the hydrostatic principle
10. The system for performing the alcohol content on-line measuring method for the segmented liquor taking according to the claim 9, wherein the number of the temperature sensors is at least one, and the temperature sensors are uniformly distributed on the alcohol content measuring pipeline, wherein one temperature sensor is tightly attached to the rear of a measuring head of the density sensor, which is positioned at the rear side according to the liquid flowing direction.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110148060.8A CN112765900B (en) | 2021-02-03 | 2021-02-03 | Alcohol content online detection method and system for segmented liquor picking |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110148060.8A CN112765900B (en) | 2021-02-03 | 2021-02-03 | Alcohol content online detection method and system for segmented liquor picking |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112765900A true CN112765900A (en) | 2021-05-07 |
CN112765900B CN112765900B (en) | 2022-05-03 |
Family
ID=75704790
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110148060.8A Active CN112765900B (en) | 2021-02-03 | 2021-02-03 | Alcohol content online detection method and system for segmented liquor picking |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112765900B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113484189A (en) * | 2021-07-02 | 2021-10-08 | 中国电信股份有限公司 | Online alcoholic strength acquisition system and method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS63253235A (en) * | 1987-04-09 | 1988-10-20 | Shibaura Denshi Kogyo Kk | Instrument for measuring alcohol concentration for spirits |
CN201561909U (en) * | 2009-09-07 | 2010-08-25 | 青岛澳科仪器有限责任公司 | Pressure-differential type density sensor and density measurement device comprising the same |
CN108318379A (en) * | 2018-02-05 | 2018-07-24 | 衡阳鑫山机械设备制造有限公司 | A kind of alcohol concentration on-line measuring device |
CN110009053A (en) * | 2019-04-12 | 2019-07-12 | 浙江树人学院(浙江树人大学) | A kind of comprehensive classification judgment method of the yellow rice wine based on BP deep neural network |
CN111537685A (en) * | 2020-05-19 | 2020-08-14 | 河北工业大学 | Alcohol content online detection wine picking and blending system |
CN111912945A (en) * | 2020-08-05 | 2020-11-10 | 湖南酃渌酒业有限公司 | White spirit alcohol degree monitoring device and monitoring method |
CN112180832A (en) * | 2020-11-04 | 2021-01-05 | 尧米(重庆)智能科技有限公司 | Artificial intelligent wine picking device and wine picking method thereof |
-
2021
- 2021-02-03 CN CN202110148060.8A patent/CN112765900B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS63253235A (en) * | 1987-04-09 | 1988-10-20 | Shibaura Denshi Kogyo Kk | Instrument for measuring alcohol concentration for spirits |
CN201561909U (en) * | 2009-09-07 | 2010-08-25 | 青岛澳科仪器有限责任公司 | Pressure-differential type density sensor and density measurement device comprising the same |
CN108318379A (en) * | 2018-02-05 | 2018-07-24 | 衡阳鑫山机械设备制造有限公司 | A kind of alcohol concentration on-line measuring device |
CN110009053A (en) * | 2019-04-12 | 2019-07-12 | 浙江树人学院(浙江树人大学) | A kind of comprehensive classification judgment method of the yellow rice wine based on BP deep neural network |
CN111537685A (en) * | 2020-05-19 | 2020-08-14 | 河北工业大学 | Alcohol content online detection wine picking and blending system |
CN111912945A (en) * | 2020-08-05 | 2020-11-10 | 湖南酃渌酒业有限公司 | White spirit alcohol degree monitoring device and monitoring method |
CN112180832A (en) * | 2020-11-04 | 2021-01-05 | 尧米(重庆)智能科技有限公司 | Artificial intelligent wine picking device and wine picking method thereof |
Non-Patent Citations (4)
Title |
---|
BOUTILLON 等: ""Indirect measurement of the alcohol concentration during the industrial distillation of cognac"", <SCI. ALIMENTS> * |
LUO, WZ 等: ""Terahertz Reflection Time-domain Spectroscopy for Measuring Alcohol Concentration"", 《PROCEEDINGS OF SPIE》 * |
李冲伟 等: ""BP神经网络在酒精发酵过程建模的应用"", 《中国酿造》 * |
范明明: ""基于近红外技术的白酒摘酒在线检测装置开发"", 《中国优秀硕士学位论文全文数据库(工程科技Ⅰ辑)》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113484189A (en) * | 2021-07-02 | 2021-10-08 | 中国电信股份有限公司 | Online alcoholic strength acquisition system and method |
Also Published As
Publication number | Publication date |
---|---|
CN112765900B (en) | 2022-05-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110084367A (en) | A kind of Forecast of Soil Moisture Content method based on LSTM deep learning model | |
CN101790955B (en) | Method for controlling irrigation according to crop water shortage degree and device thereof | |
CN112765900B (en) | Alcohol content online detection method and system for segmented liquor picking | |
CN111783987A (en) | Farmland reference crop evapotranspiration prediction method based on improved BP neural network | |
CN101630376A (en) | Soft-sensing modeling method and soft meter of multi-model neural network in biological fermentation process | |
CN104964719A (en) | Household electronic water meter flow metering method based on BP neural network | |
CN112926265A (en) | Atmospheric porous probe measurement calibration method based on genetic algorithm optimization neural network | |
CN107341359A (en) | Stalk fermentation produces the flexible measurement method of ethanol process key parameters | |
CN112446091A (en) | Artificial neural network-based pulsating pressure prediction method | |
CN101929993A (en) | Dynamic fuzzy neural network based penicillin fermentation process soft measuring modeling method | |
CN110346005A (en) | Coriolis mass flowmeter digital signal processing method based on deep learning | |
CN110633790A (en) | Method and system for measuring residual oil quantity of airplane oil tank based on convolutional neural network | |
CN101673096B (en) | Soft-measuring method for density in concentration process of salvia miltiorrhiza injection production | |
CN114169242A (en) | Intelligent control algorithm for analyzing ecological oxygenation of river channel based on parameter uncertainty | |
CN116519555A (en) | Dilution method flue gas monitoring system | |
CN103267826A (en) | Soft measurement method for online detection of gelatin concentration | |
CN105035311B (en) | A kind of aircraft gust alleviation adaptive feedforward control system | |
CN114117919A (en) | Instant learning soft measurement modeling method based on sample collaborative representation | |
CN111914488B (en) | Data area hydrologic parameter calibration method based on antagonistic neural network | |
CN205103165U (en) | Device that combines vortex correlation systematic observation ozone flux | |
CN115907204A (en) | Forest transpiration water consumption prediction method for optimizing BP neural network by sparrow search algorithm | |
CN103530521A (en) | Sunlight temperature time interval simulation method based on Fourier series and ARMA model | |
CN109933031A (en) | A kind of system and method automatically correcting soft measuring instrument according to analysis data | |
CN106599391A (en) | Relevance vector machine soft measurement modeling method based on triangular angle value dynamic weighting | |
CN113008440B (en) | Flexible liquid injection sensor detection method based on genetic algorithm optimization neural network |
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 |