CN107506862B - Online real-time grinding particle size prediction system and method based on Internet of things - Google Patents
Online real-time grinding particle size prediction system and method based on Internet of things Download PDFInfo
- Publication number
- CN107506862B CN107506862B CN201710747634.7A CN201710747634A CN107506862B CN 107506862 B CN107506862 B CN 107506862B CN 201710747634 A CN201710747634 A CN 201710747634A CN 107506862 B CN107506862 B CN 107506862B
- Authority
- CN
- China
- Prior art keywords
- grinding
- real
- neural network
- time
- data
- 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
- 238000000227 grinding Methods 0.000 title claims abstract description 191
- 239000002245 particle Substances 0.000 title claims abstract description 43
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000003062 neural network model Methods 0.000 claims abstract description 40
- 238000013528 artificial neural network Methods 0.000 claims abstract description 33
- 238000012549 training Methods 0.000 claims abstract description 25
- 238000013523 data management Methods 0.000 claims abstract description 22
- 230000005540 biological transmission Effects 0.000 claims abstract description 8
- 238000011156 evaluation Methods 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims description 19
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 14
- 238000012360 testing method Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 7
- 238000003801 milling Methods 0.000 claims description 6
- 238000004088 simulation Methods 0.000 claims description 5
- 230000001502 supplementing effect Effects 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000013500 data storage Methods 0.000 claims description 4
- 230000005284 excitation Effects 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 abstract description 7
- 238000004519 manufacturing process Methods 0.000 abstract description 6
- 238000001514 detection method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 229910052500 inorganic mineral Inorganic materials 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 239000011707 mineral Substances 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010494 dissociation reaction Methods 0.000 description 1
- 230000005593 dissociations Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000178 monomer Substances 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000002945 steepest descent method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Disintegrating Or Milling (AREA)
Abstract
The invention relates to an online real-time grinding grain size prediction system and method based on the Internet of things, wherein the system comprises a data acquisition unit, a wireless data transmission unit, a data management unit, an MES data reading unit, a grinding grain size prediction unit and a BP neural network model evaluation unit; the grinding particle size prediction unit is used for training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding particle size. According to the prediction method, the grinding parameters are collected in real time through the Internet of things, so that the production flow is optimized; and the BP neural network model is used for carrying out online real-time prediction on the grinding granularity, so that the instantaneity and the accuracy of the grinding granularity prediction are improved on the premise that a mechanism model is not allowed to be established.
Description
Technical Field
The invention relates to the technical field of Internet of things and grinding, in particular to an online real-time grinding particle size prediction method and system based on the Internet of things.
Background
The grinding classification operation is an important link in the ore dressing production process, and the purpose of grinding is to crush large-particle ores to a certain degree, separate useful minerals from gangue minerals and present a monomer dissociation state so as to be beneficial to the separation of the useful minerals. The grinding granularity is not only the most important production index of the grinding operation, but also the key factor influencing the concentrate grade and the metal recovery rate of the subsequent sorting operation. The accurate real-time acquisition of the granularity information is the key for controlling the ore grinding process and improving the ore grinding efficiency and the product quality. Therefore, the detection of the granularity is of great practical significance.
In the actual industrial production process, the ore dressing plant adopts an artificial off-line test on the ore grinding granularity to obtain the ore grinding granularity data, but the real-time requirement of control cannot be met. Therefore, the realization of the online detection of granularity is of great importance. At present, an online detection patent method for realizing granularity mainly comprises the step of carrying out online real-time prediction on the granularity of the grinding ore by solving a mechanism model for predicting the granularity of the grinding ore (201310349946.4 (an online prediction system and method for the granularity of the grinding ore). The mechanism model for predicting the grinding granularity is complex, the influence factors are more, and the influence of each factor has real-time change, so that the mechanism model for predicting the grinding granularity is difficult to accurately establish.
Disclosure of Invention
The embodiment of the invention provides an online real-time grinding particle size prediction system and method based on the Internet of things, which can improve the real-time performance and accuracy of grinding particle size prediction on the premise of not establishing a mechanism model
The invention provides an online real-time grinding particle size prediction system based on the Internet of things, which comprises the following steps:
the data acquisition unit is used for simulating and generating real-time values of grinding parameters and real-time values of grinding granularity, and carrying out data processing on the real-time values of the grinding parameters to obtain standard voltage data;
the wireless data transmission unit is used for transmitting the standard voltage data to the data management unit;
the data management unit is used for receiving the standard voltage data, restoring the standard voltage data into real-time values of grinding parameters and storing the real-time values;
the MES data reading unit is used for reading the real-time value of the grinding parameter stored by the data management unit and storing the historical value of the grinding parameter;
the grinding particle size prediction unit is used for training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding particle size;
and the BP neural network model evaluation unit is used for comparing the predicted real-time value of the grinding granularity with the real-time value of the grinding granularity generated by simulation so as to evaluate the BP neural network model.
In the online real-time grinding particle size prediction system based on the Internet of things, the data acquisition unit comprises:
the real-time data generation module is used for establishing a grinding virtual object model to simulate the data acquisition condition of an industrial field, and generating real-time values of a group of grinding parameters every 5S;
and the data processing module is used for carrying out standardization processing on the generated real-time value of the grinding parameter and converting the real-time value into standard voltage data of [0,5V ].
In the online real-time grinding particle size prediction system based on the Internet of things, the data management unit comprises:
the data display module is used for converting the received standard voltage data into standard voltage signal labels for display;
the data conversion module is used for reducing the received standard voltage quantity into a real-time value of the grinding parameter;
and the data storage module is used for storing real-time values of the grinding parameters.
In the online real-time grinding particle size prediction system based on the Internet of things, the grinding particle size prediction unit comprises:
the initialization module is used for establishing an initial model based on the BP neural network, wherein the initial model comprises an input layer, an implicit layer and an output layer;
the parameter setting module is used for setting parameters of the initial model;
the learning module is used for training the initial model of the BP neural network by taking the historical value of the ore grinding parameter as a training set of the initial model of the BP neural network to obtain a trained BP neural network model;
and the grinding granularity prediction module takes the real-time value of the grinding parameter as a test set of the trained BP neural network model, and outputs the real-time value as the grinding granularity prediction value.
The invention also provides an online real-time grinding particle size prediction method based on the Internet of things, which comprises the following steps of:
step 1: simulating to generate real-time values of grinding parameters and real-time values of grinding granularity, and performing data processing on the real-time values of the grinding parameters to obtain standard voltage data;
step 2: transmitting the standard voltage data to a data management unit;
step 3: receiving standard voltage data, and restoring the standard voltage data into real-time values of grinding parameters for storage;
step 4: reading a real-time value of the grinding parameter stored by the data management unit, and storing a historical value of the grinding parameter;
step 5: training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding granularity;
step 6: the predicted real-time value of the grind size is compared with the simulated real-time value of the grind size to assess the BP neural network model.
In the online real-time grinding particle size prediction method based on the Internet of things, the step 1 comprises the following steps:
step 1.1: establishing a grinding virtual object model to simulate the data acquisition condition of an industrial field, and generating real-time values of a group of grinding parameters every 5S;
step 1.2: and the data processing module is used for carrying out standardization processing on the generated real-time value of the grinding parameter and converting the real-time value into standard voltage data of [0,5V ].
In the online real-time grinding particle size prediction method based on the Internet of things, the step 3 comprises the following steps:
step 3.1: converting the received standard voltage data into standard voltage signal labels for display;
step 3.2: reducing the received standard voltage quantity into a real-time value of the grinding parameter;
step 3.3: the real-time values of the milling parameters are stored.
In the online real-time grinding particle size prediction method based on the internet of things, the step 5 comprises the following steps:
step 5.1: establishing an initial model based on a BP neural network, wherein the initial model comprises an input layer, an implicit layer and an output layer;
step 5.2: setting parameters of the initial model;
step 5.3: training an initial model of the BP neural network by taking a historical value of the grinding parameter as a training set of the initial model of the BP neural network, and continuously adjusting weights and thresholds of all layers to obtain a trained BP neural network model;
step 5.4: and taking the real-time value of the grinding parameter as a test set of the trained BP neural network model, and outputting the real-time value as the grinding granularity predicted value.
In the online real-time grinding particle size prediction method based on the Internet of things, the step 5.1 comprises the following steps:
step 5.1.1: setting the number of nodes of the input layer to be 4, wherein the nodes correspond to 4 grinding parameters and comprise: the ore feeding amount of the ball mill, the water feeding amount of an inlet of the ball mill, the water supplementing amount of a pump pool and the ore feeding concentration of a cyclone;
step 5.1.2: the number of hidden layers is set to be 2, and the number of hidden layer neurons is respectively 30 and 20.
In the online real-time grinding particle size prediction method based on the Internet of things, the step 5.2 comprises the following steps:
step 5.2.1: setting a hidden layer weight w ij Hidden layer threshold θ i Output layer weight w ki Output layer threshold a k ；
Step 5.2.2: setting the minimum value of the expected target error to be 1e-5, and setting the learning rate to be 0.05;
step 5.2.3: the excitation functions of the hidden layer and the output layer of the BP neural network select the logarithmic function of the S type, namely
Where n is 4 s-dimensional input column vectors.
The invention provides an online real-time grinding particle size prediction system and method based on the Internet of things. The grinding parameters are acquired in real time through the Internet of things, so that the production flow is optimized; and the BP neural network model is used for carrying out online real-time prediction on the grinding granularity, so that the instantaneity and the accuracy of the grinding granularity prediction are improved on the premise that a mechanism model is not allowed to be established.
Drawings
FIG. 1 is a block diagram of an online real-time grinding particle size prediction system based on the Internet of things;
fig. 2 is a flow chart of an online real-time grinding particle size prediction method based on the internet of things.
Detailed Description
The online real-time grinding particle size prediction system and method based on the Internet of things are described in detail below with reference to the accompanying drawings.
The invention predicts the grinding granularity on line in real time based on the Internet of things and the BP neural network. The BP neural network can learn and store a large number of input-output pattern mappings without revealing beforehand mathematical equations describing such mappings. Mathematical theory has demonstrated that it has the function of implementing any complex nonlinear mapping, which makes it particularly suitable for solving problems of complex internal mechanisms. Its learning rule is to use the steepest descent method to continuously adjust the weight and threshold of the network by back propagation to minimize the sum of squares of errors of the network.
The online real-time grinding particle size prediction system based on the Internet of things, as shown in fig. 1, comprises: the system comprises a data acquisition unit 1, a wireless data transmission unit 2, a data management unit 3, an MES data reading unit 4, an ore grinding granularity prediction unit 5 and a BP neural network model evaluation unit 6.
The data acquisition unit 1 is used for simulating and generating real-time values of grinding parameters and real-time values of grinding granularity, and carrying out data processing on the real-time values of the grinding parameters to obtain standard voltage data. The wireless data transmission unit 2 is used for transmitting standard voltage data to the data management unit. The data management unit 3 is configured to receive the standard voltage data, restore the standard voltage data to a real-time value of the grinding parameter, and store the real-time value. The MES data reading unit 4 is configured to read the real-time value of the grinding parameter stored in the data management unit, and store the historical value of the grinding parameter. The grinding particle size prediction unit 5 is configured to train the BP neural network model by using the historical value of the grinding parameter as an input of the BP neural network, and input the real-time value of the grinding parameter to the trained BP neural network model to predict the grinding particle size. The BP neural network model evaluation unit 6 is configured to compare the predicted real-time value of the grinding grain size with the real-time value of the grinding grain size generated by simulation to evaluate the BP neural network model.
The data acquisition unit 1 includes: a real-time data generation module 11 and a data processing module 12. The real-time data generation module 11 is configured to build a virtual object model of grinding to simulate the data collection of the industrial field, and generate real-time values of a set of grinding parameters at intervals of 5S. The ore grinding parameters include: the ore feeding amount of the ball mill, the water feeding amount of the inlet of the ball mill, the water supplementing amount of the pump pool and the ore feeding concentration of the cyclone. The milling parameters are stored in an oracle local database. The data processing module 12 is used for performing standardization processing on the acquired real-time value of the grinding parameter, and converting the real-time value into standard voltage data of [0,5V ].
The data management unit 3 includes: a data display module 31, a data conversion module 32 and a data storage module 33. The data display module 31 is configured to display a standard voltage signal tag converted from the received standard voltage data, so as to confirm that the data of the wireless transmission is received successfully. The data conversion module 32 is configured to restore the received standard voltage quantity to a real-time value of the grinding parameter. The data storage module 33 is used for storing real-time values of the grinding parameters. In the specific implementation, the data management unit 3 is an industrial personal computer.
The grinding particle size prediction unit 5 includes: an initialization module 51, a parameter setting module 52, a learning module 53, and a grinding particle size prediction module 54. The initialization module 51 is configured to establish an initial model based on a BP neural network including an input layer, an implicit layer, and an output layer. The parameter setting module 52 is used for setting parameters of the initial model. The learning module 53 trains the initial model of the BP neural network by using the historical value of the grinding parameter as the training set of the initial model of the BP neural network, and obtains a trained BP neural network model. The grinding grain prediction module 54 takes the real-time value of the grinding parameter as the test set of the trained BP neural network model, and outputs the real-time value as the grinding grain predicted value.
Fig. 2 shows an online real-time grinding particle size prediction method based on the internet of things, which comprises the following steps:
step 1: simulating to generate real-time values of grinding parameters and real-time values of grinding granularity, and performing data processing on the real-time values of the grinding parameters to obtain standard voltage data;
step 2: transmitting the standard voltage data to a data management unit;
step 3: receiving standard voltage data, and restoring the standard voltage data into real-time values of grinding parameters for storage;
step 4: reading a real-time value of the grinding parameter stored by the data management unit, and storing a historical value of the grinding parameter;
step 5: training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding granularity;
step 6: the predicted real-time value of the grind size is compared with the simulated real-time value of the grind size to assess the BP neural network model.
The step 1 comprises the following steps:
step 1.1: a virtual object model of ore grinding is established to simulate the data acquisition situation of the industrial field, and real-time values of a group of ore grinding parameters are generated every 5S.
In specific implementation, matlab is adopted to build a grinding virtual object model. The ore grinding parameters include: the ore feeding amount of the ball mill, the water feeding amount of the inlet of the ball mill, the water supplementing amount of the pump pool and the ore feeding concentration of the cyclone. The milling parameters are stored in an oracle local database.
Step 1.2: and the data processing module is used for carrying out standardization processing on the generated real-time value of the grinding parameter, converting the real-time value into standard voltage data of [0,5V ], and uploading the standard voltage data to the wireless node.
Step 2: the standard voltage data is sent to a data management unit, specifically:
step 2.1: the wireless node uploads standard voltage data to the intelligent wireless gateway through wireless transmission of the data;
step 2.2: and after receiving the standard voltage data, the intelligent wireless gateway synchronizes the real-time data to the industrial Personal Computer (PC) through the local area network.
The step 3 comprises the following steps:
step 3.1: the industrial personal computer converts the received standard voltage data into a standard voltage signal label for display so as to confirm that the data of wireless transmission is successfully received;
step 3.2: restoring the received standard voltage data of [0,5V ] into real-time values of grinding parameters;
step 3.3: and storing the real-time value of the ore grinding parameter into an oracle database of the industrial personal computer.
Step 4: reading a real-time value of the grinding parameter stored in the industrial personal computer, and storing a historical value of the grinding parameter, wherein the real-time value is specifically as follows: the method for reading the real-time value of the grinding parameter on the industrial Personal Computer (PC) by adopting an MES (manufacturing execution system) comprises the following steps:
step 4.1: connecting a PC (personal computer) where the MES system is located with an industrial Personal Computer (PC) through local area network segment setting;
step 4.2: an ODBC (data source manager) is configured on a PC where an MES system is located, so that the MES system can read data of an oracle database of an industrial personal computer, and the method comprises the following steps: real-time values of the milling parameters and historical values of the milling parameters.
The step 5 comprises the following steps:
step 5.1: establishing an initial model based on the BP neural network, which comprises an input layer, an implicit layer and an output layer, wherein the step 5.1 further comprises:
step 5.1.1: setting the number of nodes of the input layer to be 4, wherein the nodes correspond to 4 grinding parameters and comprise: the ore feeding amount of the ball mill, the water feeding amount of an inlet of the ball mill, the water supplementing amount of a pump pool and the ore feeding concentration of a cyclone;
step 5.1.2: the number of hidden layers is set to be 2, and the number of hidden layer neurons is respectively 30 and 20.
Step 5.2: parameter setting is carried out on the initial model, and the method further comprises the following steps:
step 5.2.1: setting a hidden layer weight w ij Hidden layer threshold θ i Output layer weight w ki Output layer threshold a k The method comprises the steps of carrying out a first treatment on the surface of the Wherein j represents the j-th input sample, i represents the i-th intermediate hidden layer node, and k represents the k-th output layer node;
step 5.2.2: setting the minimum value epsilon of the expected target error to be 1e-5, and setting the learning rate to be 0.05;
step 5.2.3: the excitation functions of the hidden layer and the output layer of the BP neural network select the logarithmic function of the S type, namely
Wherein n is 4 s-dimensional input column vectors;
step 5.3: training an initial model of the BP neural network by taking a historical value of the grinding parameter as a training set of the initial model of the BP neural network, and continuously adjusting weights and thresholds of all layers to obtain a trained BP neural network model; further comprises:
step 5.3.1: the method for acquiring the historical value of the ore grinding parameter as the training set of the initial model of the BP neural network comprises the following steps: the historical values of the ore feeding amount of the ball mill, the water feeding amount of an inlet of the ball mill, the water feeding amount of a pump pool and the ore feeding concentration of a cyclone;
step 5.3.2: normalizing the training data of the training set;
step 5.3.3: by the formula
Calculating the input of the ith node of the hidden layer through the formula
Calculating the output of the ith node of the hidden layer; m is the dimension of the input sample;
step 5.3.4: by the formula
Acquiring the input of a kth node of an output layer, and calculating the output of the kth node of the output layer through a formula psi (netk); q represents the input dimension (i.e., is the hidden layer dimension) acting on the kth output layer node;
step 5.3.5: from the formula
Calculating the output layer errors of the P training samples; l is the dimension of the output layer sample;
step 5.3.6: calculating hidden layer errors by referring to the mode of the step 5.3.5;
step 5.3.7: sequentially correcting the weight w of the output layer according to an error gradient descent method ki Output layer threshold a k Implicit layer weight w ij Implicit layer threshold θ i ；
Step 5.3.8: step 5.3.5 is executed again, the error of the output layer is calculated, and if the error is smaller than epsilon, the training process of the BP neural network is ended; otherwise, continuing to execute the step 5.3.2.
In specific implementation, 30 groups of data are selected to form a training set, and training data of the training set are shown in the following table:
table 1 historical values of grinding parameters.
| Ore feeding amount of ball mill | Inlet water supply of mill | The pump pool is supplemented with water | Cyclone ore feeding concentration | Grinding particle size |
| 11.80408021 | 4 | 7.869386805 | 36.25384938 | 31.47754722 |
| 18.96361676 | 4 | 12.64241118 | 65.93599079 | 50.5696447 |
| 23.30609519 | 4 | 15.5373968 | 90.23767278 | 62.14958719 |
| 25.9399415 | 4 | 17.29329433 | 110.1342072 | 69.17317734 |
| 27.53745004 | 4 | 18.35830003 | 126.4241118 | 73.43320011 |
| 28.50638795 | 4 | 19.00425863 | 139.7611576 | 76.01703453 |
| 29.0940785 | 4 | 19.39605233 | 150.6806072 | 77.58420933 |
| 29.45053083 | 4 | 19.63368722 | 159.6206964 | 78.534744889 |
| 29.6667301 | 4 | 19.77782007 | 166.9402224 | 79.11128028 |
| 29.79786159 | 4 | 19.86524106 | 172.9329434 | 79.460966424 |
| 29.87739686 | 4 | 19.91826457 | 177.8393683 | 79.67305828 |
| 29.92563743 | 4 | 19.95042496 | 181.a564093 | 79.80169983 |
| 29.95489682 | 4 | 19.96993122 | 185.1452844 | 79.87972486 |
| 29.97264354 | 4 | 19.98176236 | 187.8379875 | 79.92704944 |
| 29.98340747 | 4 | 19.98893831 | 190.0425863 | 79.95575325 |
| 29.98993612 | 4 | 19.99329075 | 191.8475592 | 79.97316299 |
| 29.99389595 | 4 | 19.99593063 | 193.325346 | 79.98372253 |
| 29.999629771 | 4 | 19.9975318 | 194.5352555 | 79.99012722 |
| 29.99775445 | 4 | 19.99850296 | 195.5258456 | 79.991401185 |
| 29.998638 | 4 | 19.999092 | 196.3368722 | 79.99636801 |
| 29.99917391 | 4 | 19.99944927 | 197.0008846 | 79.99779708 |
| 29.99949895 | 4 | 19.99966597 | 197.544532 | 79.99866386 |
| 29.9996961 | 4 | 19.9997974 | 197.9896329 | 79.99918959 |
| 29.99981567 | 4 | 19.99987712 | 198.3540506 | 79.99950846 |
| 29.9998882 | 4 | 19.99992547 | 198.6524106 | 79.99970187 |
| 29.99993219 | 4 | 19.99995479 | 198.8966871 | 79.99981917 |
| 29.99995887 | 4 | 19.99997258 | 199.0966838 | 79.99989032 |
| 29.99997505 | 4 | 19.99998337 | 199.2604273 | 79.99993348 |
| 29.99998487 | 4 | 19.99998991 | 199.3944891 | 79.99995965 |
| 30 | 4 | 19.99998991 | 199.4564891 | 79.99995988 |
Step 5.4: and taking the real-time value of the grinding parameter as a test set of the trained BP neural network model, and outputting the real-time value as the grinding granularity predicted value. The step 5.4 specifically comprises the following steps:
step 5.4.1: taking real-time values of grinding parameters from an industrial Personal Computer (PC) as a test set, wherein the test set comprises real-time values of the feeding amount of the ball mill, the inlet feeding amount of the ball mill, the additional feeding amount of the pump pool and the feeding concentration of the cyclone;
step 5.4.2: normalizing the test set data;
step 5.4.3: inputting the normalized test set into the BP neural network model trained in the step 5.3 to obtain the grinding particle size prediction data.
Table 2 real-time values of grinding parameters and grinding particle sizes and predicted values of grinding particle sizes generated by simulation
Table 2 compares the predicted value of the grinding granularity predicted by the BP neural network with the real-time value, and the comparison result shows that the predicted effect basically meets the actual production requirement. And from the prediction effect, the larger the data volume of the prediction method is, the prediction effect and stability are improved.
The foregoing description of the preferred embodiments of the invention is not intended to limit the scope of the invention, but rather to enable any modification, equivalent replacement, improvement or the like to be made without departing from the spirit and principles of the invention.
Claims (6)
1. Online real-time grinding particle size prediction system based on Internet of things, which is characterized by comprising:
the data acquisition unit is used for simulating and generating real-time values of grinding parameters and real-time values of grinding granularity, and carrying out data processing on the real-time values of the grinding parameters to obtain standard voltage data;
the wireless data transmission unit is used for transmitting the standard voltage data to the data management unit;
the data management unit is used for receiving the standard voltage data, restoring the standard voltage data into real-time values of grinding parameters and storing the real-time values;
the MES data reading unit is used for reading the real-time value of the grinding parameter stored by the data management unit and storing the historical value of the grinding parameter;
the grinding particle size prediction unit is used for training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding particle size;
the BP neural network model evaluation unit is used for comparing the predicted real-time value of the grinding granularity with the real-time value of the grinding granularity generated by simulation so as to evaluate the BP neural network model;
the data acquisition unit includes:
the real-time data generation module is used for establishing a grinding virtual object model to simulate the data acquisition condition of an industrial field, and generating real-time values of a group of grinding parameters every 5S;
the data processing module is used for carrying out standardized processing on the generated real-time value of the grinding parameter and converting the real-time value into standard voltage data of [0,5V ];
the grinding particle size prediction unit includes:
the initialization module is used for establishing an initial model based on the BP neural network, wherein the initial model comprises an input layer, an implicit layer and an output layer;
the parameter setting module is used for setting parameters of the initial model;
the learning module is used for training the initial model of the BP neural network by taking the historical value of the ore grinding parameter as a training set of the initial model of the BP neural network to obtain a trained BP neural network model;
and the grinding granularity prediction module takes the real-time value of the grinding parameter as a test set of the trained BP neural network model, and outputs the real-time value as the grinding granularity prediction value.
2. The internet of things-based grinding grain online real-time prediction system of claim 1, wherein the data management unit comprises:
the data display module is used for converting the received standard voltage data into standard voltage signal labels for display;
the data conversion module is used for reducing the received standard voltage quantity into a real-time value of the grinding parameter;
and the data storage module is used for storing real-time values of the grinding parameters.
3. An online real-time grinding particle size prediction method based on the Internet of things is characterized by comprising the following steps of:
step 1: simulating to generate real-time values of grinding parameters and real-time values of grinding granularity, and performing data processing on the real-time values of the grinding parameters to obtain standard voltage data;
step 2: transmitting the standard voltage data to a data management unit;
step 3: receiving standard voltage data, and restoring the standard voltage data into real-time values of grinding parameters for storage;
step 4: reading a real-time value of the grinding parameter stored by the data management unit, and storing a historical value of the grinding parameter;
step 5: training the BP neural network model by taking the historical value of the grinding parameter as the input of the BP neural network, and inputting the real-time value of the grinding parameter into the trained BP neural network model to predict the grinding granularity;
step 6: comparing the predicted real-time value of the grinding grain size with the real-time value of the grinding grain size generated by simulation to evaluate the BP neural network model;
the step 5 comprises the following steps:
step 5.1: establishing an initial model based on a BP neural network, wherein the initial model comprises an input layer, an implicit layer and an output layer;
step 5.2: setting parameters of the initial model;
step 5.3: training an initial model of the BP neural network by taking a historical value of the grinding parameter as a training set of the initial model of the BP neural network, and continuously adjusting weights and thresholds of all layers to obtain a trained BP neural network model;
step 5.4: taking the real-time value of the grinding parameter as a test set of the trained BP neural network model, and outputting the real-time value as a grinding granularity predicted value;
the step 1 comprises the following steps:
step 1.1: establishing a grinding virtual object model to simulate the data acquisition condition of an industrial field, and generating real-time values of a group of grinding parameters every 5S;
step 1.2: and the data processing module is used for carrying out standardization processing on the generated real-time value of the grinding parameter and converting the real-time value into standard voltage data of [0,5V ].
4. The online real-time grinding particle size prediction method based on the internet of things as set forth in claim 3, wherein the step 3 includes:
step 3.1: converting the received standard voltage data into standard voltage signal labels for display;
step 3.2: reducing the received standard voltage quantity into a real-time value of the grinding parameter;
step 3.3: the real-time values of the milling parameters are stored.
5. The online real-time grinding particle size prediction method based on the internet of things as set forth in claim 3, wherein the step 5.1 includes:
step 5.1.1: setting the number of nodes of the input layer to be 4, wherein the nodes correspond to 4 grinding parameters and comprise: the ore feeding amount of the ball mill, the water feeding amount of an inlet of the ball mill, the water supplementing amount of a pump pool and the ore feeding concentration of a cyclone;
step 5.1.2: the number of hidden layers is set to be 2, and the number of hidden layer neurons is respectively 30 and 20.
6. The online real-time grinding particle size prediction method based on the internet of things according to claim 3, wherein the step 5.2 comprises:
step 5.2.1: setting a hidden layer weight w ij Hidden layer threshold θ i Output layer weight w ki Output layer threshold a k Where j represents the j-th input sample, i represents the i-th hidden layer node, and k represents the k-th output layer node;
step 5.2.2: setting the minimum value of the expected target error to be 1e-5, and setting the learning rate to be 0.05;
step 5.2.3: the excitation functions of the hidden layer and the output layer of the BP neural network select the logarithmic function of the S type, namely
Where n is 4 s-dimensional input column vectors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710747634.7A CN107506862B (en) | 2017-08-28 | 2017-08-28 | Online real-time grinding particle size prediction system and method based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710747634.7A CN107506862B (en) | 2017-08-28 | 2017-08-28 | Online real-time grinding particle size prediction system and method based on Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107506862A CN107506862A (en) | 2017-12-22 |
| CN107506862B true CN107506862B (en) | 2023-07-04 |
Family
ID=60693640
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710747634.7A Active CN107506862B (en) | 2017-08-28 | 2017-08-28 | Online real-time grinding particle size prediction system and method based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107506862B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108993751B (en) * | 2018-09-25 | 2021-01-15 | 首钢集团有限公司 | Method for optimizing ore grinding granularity |
| CN110180683B (en) * | 2019-06-21 | 2021-01-26 | 安徽国兰智能科技有限公司 | Flotation system based on big data analysis |
| CN113296561B (en) * | 2020-08-20 | 2022-06-14 | 中冶长天国际工程有限责任公司 | Control system and control method for pelletizer |
| CN112317110A (en) * | 2020-09-27 | 2021-02-05 | 鞍钢集团矿业有限公司 | Ore grinding granularity prediction system and method based on deep learning |
| CN114095803B (en) * | 2022-01-19 | 2022-05-31 | 高视科技(苏州)有限公司 | Data acquisition optimization method and product of Internet of things information management system |
| CN114112819B (en) * | 2022-01-25 | 2022-05-10 | 北矿智云科技(北京)有限公司 | Method and device for measuring ore grinding granularity |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103412489A (en) * | 2013-08-12 | 2013-11-27 | 东北大学 | Ore grinding granularity online prediction system and method |
-
2017
- 2017-08-28 CN CN201710747634.7A patent/CN107506862B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103412489A (en) * | 2013-08-12 | 2013-11-27 | 东北大学 | Ore grinding granularity online prediction system and method |
Non-Patent Citations (1)
| Title |
|---|
| 模糊神经网络在磨矿控制系统中的研究应用;李明旭;《中国优秀硕士学位论文全文数据库》;20160331;第4-22、42-54页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107506862A (en) | 2017-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107506862B (en) | Online real-time grinding particle size prediction system and method based on Internet of things | |
| Qin et al. | A dual-stage attention-based recurrent neural network for time series prediction | |
| CN113642225B (en) | CNN-LSTM short-term wind power prediction method based on attention mechanism | |
| CN104850901B (en) | A kind of flexible measurement method and system based on multi-model prediction gas density | |
| CN109472057B (en) | Product processing quality prediction device and method based on cross-process implicit parameter memory | |
| CN107230113A (en) | A kind of house property appraisal procedure of multi-model fusion | |
| CN106355499A (en) | Stock price trend forecasting and trading method | |
| CN110766212A (en) | Ultra-short-term photovoltaic power prediction method for historical data missing electric field | |
| CN105589806A (en) | SMOTE+Boosting algorithm based software defect tendency prediction method | |
| CN104616078A (en) | Spiking neural network based photovoltaic system generation power prediction method | |
| Williams et al. | Exploring deep learning for dig-limit optimization in open-pit mines | |
| CN113326960A (en) | Subway traction energy consumption prediction method based on particle swarm optimization LSTM | |
| CN108762503A (en) | A kind of man-machine interactive system based on multi-modal data acquisition | |
| CN113627070A (en) | Short-term photovoltaic power prediction method | |
| Xing-yu et al. | RNN-based method for fault diagnosis of grinding system | |
| CN118397429A (en) | Integrated system and method for checking intelligent sorting of cargoes based on computer vision | |
| Wibawa et al. | Deep Learning Approaches with Optimum Alpha for Energy Usage Forecasting. | |
| CN113554213A (en) | Natural gas demand prediction method, system, storage medium and equipment | |
| CN109657938B (en) | Rail transit system cost influence element identification and analysis method | |
| Gu et al. | Employment Quality Evaluation Model Based on Hybrid Intelligent Algorithm | |
| CN103942421A (en) | Method for predicting testing data on basis of noise disturbance | |
| Chen et al. | Transfer learning and LSTM to predict stock price | |
| CN114625831A (en) | Classification evaluation feedback method for load identification of smart power grid | |
| CN111626376A (en) | Domain adaptation method and system based on discrimination joint probability | |
| CN111177975A (en) | Aviation equipment availability prediction method based on artificial intelligence |
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 |