CN114723131A - Mine environment intelligent prediction system based on artificial intelligence - Google Patents
Mine environment intelligent prediction system based on artificial intelligence Download PDFInfo
- Publication number
- CN114723131A CN114723131A CN202210350937.6A CN202210350937A CN114723131A CN 114723131 A CN114723131 A CN 114723131A CN 202210350937 A CN202210350937 A CN 202210350937A CN 114723131 A CN114723131 A CN 114723131A
- Authority
- CN
- China
- Prior art keywords
- environment
- prediction
- heavy metal
- environmental
- training 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.)
- Pending
Links
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 25
- 230000007613 environmental effect Effects 0.000 claims abstract description 70
- 238000012545 processing Methods 0.000 claims abstract description 35
- 230000002265 prevention Effects 0.000 claims abstract description 21
- 238000012544 monitoring process Methods 0.000 claims abstract description 9
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims description 67
- 229910001385 heavy metal Inorganic materials 0.000 claims description 38
- PXHVJJICTQNCMI-UHFFFAOYSA-N Nickel Chemical compound [Ni] PXHVJJICTQNCMI-UHFFFAOYSA-N 0.000 claims description 28
- 238000013528 artificial neural network Methods 0.000 claims description 22
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 claims description 14
- HCHKCACWOHOZIP-UHFFFAOYSA-N Zinc Chemical compound [Zn] HCHKCACWOHOZIP-UHFFFAOYSA-N 0.000 claims description 14
- 229910052785 arsenic Inorganic materials 0.000 claims description 14
- RQNWIZPPADIBDY-UHFFFAOYSA-N arsenic atom Chemical compound [As] RQNWIZPPADIBDY-UHFFFAOYSA-N 0.000 claims description 14
- 229910052793 cadmium Inorganic materials 0.000 claims description 14
- BDOSMKKIYDKNTQ-UHFFFAOYSA-N cadmium atom Chemical compound [Cd] BDOSMKKIYDKNTQ-UHFFFAOYSA-N 0.000 claims description 14
- 229910052802 copper Inorganic materials 0.000 claims description 14
- 239000010949 copper Substances 0.000 claims description 14
- 239000011133 lead Substances 0.000 claims description 14
- 229910052759 nickel Inorganic materials 0.000 claims description 14
- 229910052725 zinc Inorganic materials 0.000 claims description 14
- 239000011701 zinc Substances 0.000 claims description 14
- 230000007774 longterm Effects 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 10
- 230000015654 memory Effects 0.000 description 8
- 230000008569 process Effects 0.000 description 6
- 238000005065 mining Methods 0.000 description 5
- 241000282414 Homo sapiens Species 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000004880 explosion Methods 0.000 description 3
- 230000008033 biological extinction Effects 0.000 description 2
- 230000001934 delay Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 229910052500 inorganic mineral Inorganic materials 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 239000011707 mineral Substances 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 239000002893 slag Substances 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- -1 landforms Substances 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 239000005416 organic matter Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Biophysics (AREA)
- Tourism & Hospitality (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (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)
Abstract
The invention provides an artificial intelligence-based mine environment intelligent prediction system, which comprises: the system comprises an environmental parameter acquisition module, a data processing module, a mine environment intelligent prediction module and a prevention alarm module; the environment parameter acquisition module is used for acquiring environment parameters in a monitoring area; the data processing module is used for preprocessing the environment parameters; the mine environment intelligent prediction module is used for predicting the future trend according to the preprocessed environment parameters; and the prevention alarm module alarms according to the predicted result. The intelligent prediction system based on the mine environment carries out intelligent prediction on the environmental data, and realizes intelligent prediction and advanced prevention and treatment on the mine environment change trend.
Description
Technical Field
The invention belongs to the technical field of mine restoration, and particularly relates to an artificial intelligence-based mine environment intelligent prediction system.
Background
The coverage of the mineral resource centralized distribution area in China overlaps with the population dense area, the grain main production area and the ecological fragile area, so that the national grain safety, food safety, ecological safety and human health are seriously influenced by the heavy metal-organic matter compound pollution of the mine. .
The LSTM (Long Short-Term Memory network) is a time-cycle neural network and is specially designed for solving the problems of gradient extinction and gradient explosion of RNN. Compared with RNN, LSTM has a unique design structure, which adds an input gate, an output gate and a forgetting gate in a hidden layer, and uses a memory state unit to store and process long-time sequence information, and is very suitable for processing and predicting important events with very long intervals and delays in time sequence.
The mining, dressing and metallurgy process of nonferrous metal mines can cause serious damage to soil, landforms, water resources, vegetation and the like in areas, meanwhile mining abandoned lands such as slag discharge pit abandoned lands, stope abandoned lands, tailing slag abandoned lands and the like can seriously damage the surrounding ecological environment, if ecological restoration is not carried out, the environment deterioration process can be accelerated, mineral resource loss and frequent geological disasters can be caused, and disastrous consequences can be caused to human beings and natural environments. And the ecological restoration of the mine is difficult to meet the ecological restoration requirement in a short time only by natural succession, so that the artificial intelligent early warning of the mine waste land can effectively implement the measures of 'one mine and one strategy', the cost is reduced, the restoration of the ecological environment of the mine is accelerated more efficiently, the water and soil loss of a mining area is reduced, the deterioration of the ecological environment of the mining area is restrained, and finally, the ecological system of the mining area meets the requirements of reasonable structure, high function efficiency and continuous and stable development. However, the conversion of real-time site environment detection data subjected to intelligent analysis between the data processing module and the intelligent prediction system is limited, and the practical application of the environment intelligent prediction system in the mine waste site early warning field still faces huge challenges.
Therefore, a novel intelligent prediction system special for the complex pollution environment of the abandoned mine land is urgently needed.
Disclosure of Invention
In order to solve the technical problems, the invention provides an artificial intelligence-based intelligent mine environment prediction system, which utilizes an LSTM-based intelligent prediction system to intelligently and real-timely predict the future trend of environmental parameter data of a monitored area, thereby really realizing intelligent prediction and advanced prevention and treatment of the mine environment change trend.
In order to achieve the aim, the invention provides an intelligent mine environment prediction system based on artificial intelligence; the method comprises the following steps: the system comprises an environmental parameter acquisition module, a data processing module, a mine environment intelligent prediction module and a prevention alarm module;
the environment parameter acquisition module is used for acquiring environment parameters in a monitoring area;
the data processing module is used for preprocessing the environment parameters;
the mine environment intelligent prediction module is used for predicting the future trend according to the preprocessed environment parameters;
and the prevention alarm module alarms according to the predicted result.
Optionally, the environmental parameters include: pH data, the content of an effective state of the heavy metal and the content of an occurrence form of the heavy metal;
the environmental parameter acquisition module comprises: a pH sensor, a heavy metal effective state content sensor and a heavy metal occurrence state content sensor;
the heavy metal available state content sensor is used for collecting the available state content of the heavy metal;
the heavy metal occurrence form content sensor is used for collecting occurrence form contents of arsenic, copper, cadmium, lead, zinc and nickel.
Optionally, the preprocessing in the data processing module is: and sequencing the environmental parameters according to the time sequence, and carrying out mean value processing on the environmental parameters of the same type.
Optionally, the mine environment intelligent prediction module includes: the system comprises an LSTM neural network, a training database and a trend prediction model;
the training database is used for storing environmental training data and marking the environmental training data;
the LSTM neural network is used for carrying out iterative training based on the environment training data after labeling to obtain the trend prediction model;
the trend prediction model is used for predicting future trends according to the environment parameters.
Optionally, the LSTM neural network comprises: an input layer, a hidden layer and an output layer;
the input layer is used for inputting environmental training data;
the hidden layer is used for iteratively learning short-range and long-range semantic features of the time sequence data;
the output layer is used for outputting the prediction result.
Optionally, the trend prediction model for future trend prediction comprises: and carrying out short-term trend prediction, medium-term trend prediction and long-term trend prediction.
Optionally, the environmental training data are arranged in a time sequence;
the environmental training data includes: pH data, the content of an effective state of the heavy metal and the content of an occurrence form of the heavy metal;
the effective state content of the heavy metal is the effective state content of arsenic, copper, cadmium, lead, zinc and nickel;
the occurrence form content of the heavy metal is the occurrence form content of arsenic, copper, cadmium, lead, zinc and nickel.
Optionally, the LSTM neural network further comprises LSTM network parameters;
the LSTM network parameters comprise learning rate, iteration times and stepsize;
the stepsize takes a value between 1 and 24.
Optionally, the manner of labeling the environmental training data is as follows:
when stepsize is equal to 1, taking the environment training data at the n + x time as a label of the environment training data at the n time; when stepsize is 2, taking the environment training data at the nth + x moment as the labels of the environment training data at the nth and the (n-1) th moments; when stepsize is 3, taking the environment training data at the nth + x moment as the labels of the environment training data at the nth, the n-1 and the n-2 moments, and the rest are analogized in sequence;
the n-th time environmental training data is the average value of the environmental training data in a time period, and n is an arbitrary integer not less than 0; x is a prediction step length parameter, and x is an arbitrary integer not less than 0;
the value of the prediction step size parameter x depends on short-term trend prediction, medium-term trend prediction or long-term trend prediction.
Optionally, the trend prediction model comprises a pH trend prediction model, a heavy metal effective state content trend prediction model and a heavy metal occurrence form content trend prediction model;
the prevention alarm module is specifically used for pH alarm, heavy metal effective state content alarm and heavy metal occurrence state content alarm.
Compared with the prior art, the invention has the following advantages and technical effects:
the intelligent prediction system based on artificial intelligence predicts the future trend of the environmental data of the monitored area, really realizes advanced prevention and treatment before the environmental parameters are deteriorated, and avoids high cost, long time consumption and large amount of manpower and material resources generated by treatment after the environmental parameters are deteriorated;
the intelligent prediction and advanced prevention management have real-time performance and synchronism, and the first time advanced discovery and the first time advanced management are really realized;
the intelligent prediction system provided by the invention is developed and designed completely based on artificial intelligence and an artificial neural network LSTM, and has higher intelligent degree.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application. In the drawings:
FIG. 1 is a schematic structural diagram of an artificial intelligence-based mine environment intelligent prediction system according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an intelligent mine environment prediction module according to an embodiment of the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer-executable instructions and that, although a logical order is illustrated in the flowcharts, in some cases, the steps illustrated or described may be performed in an order different than presented herein.
Examples
As shown in fig. 1, the present embodiment provides an intelligent mine environment prediction system based on artificial intelligence, which is characterized by including: the system comprises an environmental parameter acquisition module, a data processing module, a mine environment intelligent prediction module and a prevention alarm module;
the environment parameter acquisition module is used for acquiring environment parameters in a monitoring area;
the data processing module is used for preprocessing the environmental parameters;
the mine environment intelligent prediction module is used for predicting the future trend according to the preprocessed environment parameters;
and the prevention alarm module alarms according to the predicted result.
Further, the environmental parameter acquisition module includes a plurality of sensors, and the sensor sets up one or more collection site in the environmental monitoring area, and every collection site sets up one or more sensors according to environment, condition, demand etc. to detect one or more parameters in the environmental monitoring area. For example, the environmental parameters include pH sensor, sensor for effective state content of heavy metal (such as arsenic, copper, cadmium, lead, zinc, nickel, etc.), and sensor for occurrence state content of heavy metal (such as arsenic, copper, cadmium, lead, zinc, nickel, etc.).
Furthermore, the environment parameter acquisition module is connected with the data processing module, acquires environment parameter data in the monitoring area, and transmits the environment parameter data to the data processing module in real time for data preprocessing. The data preprocessing mode is as follows: the different types of data collected by each collection site are sorted according to time sequence, and multiple sets of parameter data of the same type collected by a plurality of collection sites are subjected to mean value processing, wherein the mean value processing requires that the data of the same type with the same time sequence number are subjected to mean value processing. If only one part of data of a certain type is available, the data is not subjected to mean value processing and is only sorted according to time sequence. And then transmitting the sequenced data to a corresponding prediction model according to the data type to predict the future trend.
Artificial Intelligence (AI) is a machine learning technique that involves the study, design, and application of intelligent machines. The artificial intelligence models the complex relation among the data through an artificial neural network, and forms more abstract high-level characteristics by combining low-level characteristics, so that the data characteristics are extracted, the artificial intelligence has stronger modeling and reasoning capabilities, and the artificial intelligence can simulate the human brain to work. Unlike the traditional method, artificial intelligence can autonomously learn useful characteristics from data only through training and learning of the own neural network without determining a mathematical equation of a mapping relation between input and output in advance, so that the output result which is closest to the expected output value can be obtained when the input value is given. The artificial neural network is a neural network formed by connecting a large number of processing units, has strong self-learning capability and can automatically summarize the data rule characteristics from the existing data.
In a traditional RNN neural network, the RNN training method adds a time consideration on the basis of a traditional back propagation algorithm, but when the propagation time is longer, the residual error needing to be returned is exponentially reduced, so that the network weight is slowly updated, the long-term memory effect of the RNN cannot be reflected, and at the moment, a gradient signal becomes very tiny and nearly zero or is scattered all at once, so that the problems of gradient disappearance and gradient explosion in the RNN are caused. Therefore, a memory unit is required to store memory, and the LSTM model is proposed.
The LSTM (Long Short-Term Memory network) is a time-cycle neural network and is specially designed for solving the problems of gradient extinction and gradient explosion of RNN. Compared with RNN, LSTM has unique design structure, which adds input gate, output gate and forgetting gate (three gates can pass information selection mode) in hidden layer, and uses memory state unit to store and process long time sequence information, wherein the memory gate is used to select forgetting some information, the input gate is used to memorize some information, the information is merged with the memory through the input gate and the memory gate, and the output gate outputs the information finally. LSTM is therefore well suited to handle and predict significant events in time series with very long intervals and delays.
Further, the mine environment intelligent prediction module is shown in fig. 2 and comprises an LSTM neural network, a training database and a trend prediction model, and the mine environment intelligent prediction module can perform short-term, medium-term or long-term future trend prediction on the environmental parameters, wherein the short term means 4-48 h, the medium term means 2-7 days, and the long term means 7-15 days. The intermediate-term and long-term future tendency predictions are not limited to the above-described large and small intervals, but may be set to have a larger interval value than the above-described interval, but the longer the predicted future time is, the larger the corresponding error is.
The LSTM neural network comprises an input layer, a hidden layer and an output layer, wherein the input layer inputs training data; the hidden layer iteratively learns short-range and long-range semantic features of the time sequence data; the output layer outputs the prediction result. The LSTM network parameters comprise learning rate, iteration times, stepsize and the like, wherein the key network parameter stepsize takes a value between 1 and 24, and the specific value is determined according to the scale of the environmental parameter training data and the actual conditions and requirements.
The training database comprises environmental training data collected by a pH sensor, a heavy metal active state content (such as arsenic, copper, cadmium, lead, zinc, nickel and the like) sensor and a heavy metal occurrence state content (such as arsenic, copper, cadmium, lead, zinc, nickel and the like) sensor; each type of environmental training data in the training database needs to be labeled, and the labeling processing mode meets the following requirements:
when stepsize is equal to 1, the labeling processing mode is that the environment training data at the n + x th moment is used as a label of the environment training data at the n th moment; when stepsize is 2, the labeling processing mode is that the environment training data at the nth + x moment is used as the labels of the environment training data at the nth and the (n-1) th moments; when stepsize is equal to 3, the labeling processing mode is to use the environment training data at the nth + x moment as labels of the environment training data at the nth, the n-1 and the n-2 moments, and the rest is analogized in sequence, wherein x is a prediction step size parameter and is any integer which is greater than or equal to 0.
The value of the prediction step length parameter x is related to short-term prediction, medium-term prediction and long-term prediction, if the prediction is carried out in a short term, the value of x is smaller, if the prediction is carried out in a long term, the value of x is larger, and if the prediction is carried out in a medium term, the value of x is between the short-term prediction and the medium-term prediction. By adjusting the prediction step size parameter x, the prediction environment parameter data corresponding to the near or far future time can be obtained. For example, when the value of x is 3 (the time span between the nth and the (n + 1) th time is set to be 4 hours), the trend prediction model can obtain the predicted environment data after 12 hours; when x is 24, the trend prediction model can obtain prediction environment data after 96 hours (namely after 4 days); when x is 72, the trend prediction model can obtain the predicted environment data after 12 days.
And each type of environmental parameter training data in the training database are arranged according to a time sequence, wherein the environmental parameter data at the nth time refers to the average value of the environmental parameters in a certain time period, but not the environmental parameter value at a time point. Moreover, the time periods corresponding to the environmental parameter data at all times have the same unified interval size. In addition, the time span between the nth and the (n + 1) th time (where n is an arbitrary integer of 0 or more) is preferably 4 or 6 or 8 hours, and the maximum time does not exceed 48 hours.
The trend prediction model comprises a trend prediction model corresponding to environmental training data acquired by a pH sensor, a heavy metal effective state content (such as arsenic, copper, cadmium, lead, zinc, nickel and the like) sensor and a heavy metal occurrence form content (such as arsenic, copper, cadmium, lead, zinc, nickel and the like) sensor. Each type of trend prediction model is obtained by utilizing an LSTM network to perform training and learning based on corresponding environmental training data in a training database, and performs future trend prediction on corresponding environmental parameters.
Each type of environmental training data in the training database needs to update and supplement the latest environmental parameters to the corresponding environmental training data in time sequence at regular intervals, so that the validity and the continuity of the data can be ensured. And each type of trend prediction model also needs retraining after the environmental training data is regularly updated and updating the trained model, so that the accuracy, reliability and effectiveness of each type of trend prediction model on future trend prediction can be ensured.
The training process of the LSTM neural network is as follows:
carrying out iterative training by turns by the LSTM neural network, in each round of training, taking the environmental training data at the n, n-1, … … and n-stepsize +1 moments as input data, outputting predicted environmental parameter data aiming at the n + x moment, then matching the predicted environmental parameter data at the n + x moment with the actual environmental parameter data at the n + x moment, if the matching error does not meet the preset requirement, correcting and adjusting the weight value parameters of each neural unit of the neural network according to the matching error, then continuing taking the environmental parameter data at the n, n-1, … … and n-stepsize +1 moments as input data, starting the next round of iterative training until the matching error between the predicted environmental parameter data at the n + x moment and the actual environmental parameter data at the n + x moment is smaller than a specified threshold value, the neural network training is complete.
The data processing module preprocesses the acquired environmental parameters, and then correspondingly transmits each type of environmental parameters to a corresponding trend prediction model for future trend prediction.
The mine environment intelligent prediction module and the data processing module are deployed on a computing processing server, the computing processing server needs to meet 64G memory and 4T hard disk space so as to ensure that an artificial intelligent computing processor has enough computing processing capacity and data storage space, and the computing processing server is preferably provided with a plurality of GPUs, so that the requirement of parallel processing computing can be met.
The prevention alarm module comprises a pH alarm module, an alarm module for the content of effective states of heavy metals (such as arsenic, copper, cadmium, lead, zinc, nickel and the like), and an alarm module for the content of occurrence forms of heavy metals (such as arsenic, copper, cadmium, lead, zinc, nickel and the like), and the alarm modes of different submodules are different.
The prevention alarm module is connected with the mine environment intelligent prediction module and receives the prediction result sent by the mine environment intelligent prediction module in real time. As long as the prediction result is obviously or more obviously worse than the current environmental parameter, the prevention alarm module alarms, even if the prediction result is the standard environmental parameter; if the prediction is worse than the current environmental parameters but not significant, it is likely to be caused by normal parameter fluctuations, and no alarm is given.
The detection process of the invention comprises the following steps:
the environment parameter acquisition module acquires one or more environment parameter data in the monitoring area through one or more sensors at one or more acquisition sites in the monitoring area, and transmits the acquired environment parameter data to the data processing module in real time; the data processing module carries out real-time preprocessing on the acquired environmental data and then transmits each type of processed environmental parameters to the mine environment intelligent prediction module in real time; the mine environment intelligent prediction module comprises a plurality of trend prediction models, each type of trend prediction model carries out future trend prediction on corresponding type of environment parameter data transmitted by the data processing module in real time, and transmits the prediction result to a corresponding prevention alarm submodule under the prevention alarm module in real time; if the prediction result is obviously or more obviously worse than the current environmental parameter, the prevention alarm submodule alarms even if the prediction result is the standard environmental parameter, and the alarm modes of different prevention alarm submodules are different; if the prediction is worse than the current environmental parameter data but not significant, it is likely to be caused by normal parameter fluctuations, and no alarm is given.
The above description is only for the preferred embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. The utility model provides a mine environment intelligence prediction system based on artificial intelligence which characterized in that includes: the system comprises an environmental parameter acquisition module, a data processing module, a mine environment intelligent prediction module and a prevention alarm module;
the environment parameter acquisition module is used for acquiring environment parameters in a monitoring area;
the data processing module is used for preprocessing the environment parameters;
the mine environment intelligent prediction module is used for predicting the future trend according to the preprocessed environment parameters;
and the prevention alarm module alarms according to the predicted result.
2. The artificial intelligence based mine environment intelligent prediction system of claim 1,
the environmental parameters include: pH data, the content of an effective state of the heavy metal and the content of an occurrence form of the heavy metal;
the environmental parameter acquisition module comprises: a pH sensor, a heavy metal effective state content sensor and a heavy metal occurrence state content sensor;
the heavy metal effective state content sensor is used for collecting the effective state content of the heavy metal;
the heavy metal occurrence form content sensor is used for collecting occurrence form contents of arsenic, copper, cadmium, lead, zinc and nickel.
3. The artificial intelligence based mine environment intelligent prediction system of claim 2,
the preprocessing in the data processing module is as follows: and sequencing the environmental parameters according to the time sequence, and carrying out mean value processing on the environmental parameters of the same type.
4. The artificial intelligence based mine environment intelligent prediction system of claim 3,
the intelligent mine environment prediction module comprises: the system comprises an LSTM neural network, a training database and a trend prediction model;
the training database is used for storing environmental training data and marking the environmental training data;
the LSTM neural network is used for carrying out iterative training based on the environment training data after labeling to obtain the trend prediction model;
the trend prediction model is used for predicting future trends according to the environment parameters.
5. The artificial intelligence based mine environment intelligent prediction system of claim 4,
the LSTM neural network comprises: an input layer, a hidden layer and an output layer;
the input layer is used for inputting environmental training data;
the hidden layer is used for iteratively learning short-range and long-range semantic features of the time series data;
the output layer is used for outputting a prediction result.
6. The artificial intelligence based mine environment intelligent prediction system of claim 4,
the trend prediction model for predicting the future trend comprises the following steps: and carrying out short-term trend prediction, medium-term trend prediction and long-term trend prediction.
7. The artificial intelligence based mine environment intelligent prediction system of claim 6,
the environment training data are arranged according to a time sequence;
the environmental training data includes: pH data, the content of effective state of heavy metal and the content of occurrence form of heavy metal;
the effective state content of the heavy metal is the effective state content of arsenic, copper, cadmium, lead, zinc and nickel;
the occurrence form content of the heavy metal is the occurrence form content of arsenic, copper, cadmium, lead, zinc and nickel.
8. The artificial intelligence based mine environment intelligence prediction system of claim 7, wherein the LSTM neural network further comprises LSTM network parameters;
the LSTM network parameters comprise learning rate, iteration times and stepsize;
the stepsize takes a value between 1 and 24.
9. The artificial intelligence based mine environment intelligent prediction system of claim 8, wherein the environmental training data is labeled in a manner of:
when stepsize is equal to 1, taking the environment training data at the n + x time as a label of the environment training data at the n time; when stepsize is 2, taking the environment training data at the nth + x moment as the labels of the environment training data at the nth and the (n-1) th moments; when stepsize is 3, taking the environment training data at the nth + x moment as the labels of the environment training data at the nth, the n-1 and the n-2 moments, and the rest are analogized in sequence;
the n-th time environmental training data is the average value of the environmental training data in a time period, and n is an arbitrary integer not less than 0; x is a prediction step length parameter, and x is an arbitrary integer not less than 0;
the value of the prediction step size parameter x depends on short-term trend prediction, medium-term trend prediction or long-term trend prediction.
10. The artificial intelligence based mine environment intelligent prediction system of claim 4,
the trend prediction model comprises a pH trend prediction model, a heavy metal effective state content trend prediction model and a heavy metal occurrence form content trend prediction model;
the prevention alarm module is specifically used for pH alarm, heavy metal effective state content alarm and heavy metal occurrence state content alarm.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210350937.6A CN114723131A (en) | 2022-04-02 | 2022-04-02 | Mine environment intelligent prediction system based on artificial intelligence |
ZA2023/02949A ZA202302949B (en) | 2022-04-02 | 2023-02-28 | Mine environment intelligent prediction system based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210350937.6A CN114723131A (en) | 2022-04-02 | 2022-04-02 | Mine environment intelligent prediction system based on artificial intelligence |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114723131A true CN114723131A (en) | 2022-07-08 |
Family
ID=82242338
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210350937.6A Pending CN114723131A (en) | 2022-04-02 | 2022-04-02 | Mine environment intelligent prediction system based on artificial intelligence |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN114723131A (en) |
ZA (1) | ZA202302949B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115249331A (en) * | 2022-09-22 | 2022-10-28 | 山东世纪矿山机电有限公司 | Mine ecological safety identification method based on convolutional neural network model |
-
2022
- 2022-04-02 CN CN202210350937.6A patent/CN114723131A/en active Pending
-
2023
- 2023-02-28 ZA ZA2023/02949A patent/ZA202302949B/en unknown
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115249331A (en) * | 2022-09-22 | 2022-10-28 | 山东世纪矿山机电有限公司 | Mine ecological safety identification method based on convolutional neural network model |
CN115249331B (en) * | 2022-09-22 | 2022-12-20 | 山东世纪矿山机电有限公司 | Mine ecological safety identification method based on convolutional neural network model |
Also Published As
Publication number | Publication date |
---|---|
ZA202302949B (en) | 2023-04-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110633855B (en) | Bridge health state detection and management decision making system and method | |
Alarifi et al. | Earthquakes magnitude predication using artificial neural network in northern Red Sea area | |
Hassanien et al. | Machine learning in telemetry data mining of space mission: basics, challenging and future directions | |
Esmaeili et al. | Multiple regression, ANN and ANFIS models for prediction of backbreak in the open pit blasting | |
Fdez‐Riverola et al. | Forecasting red tides using an hybrid neuro‐symbolic system | |
CN103186575B (en) | A kind of clustering method of sensing data and system | |
CN114626627A (en) | Monitoring and early warning system for carbon emission in area | |
Lai et al. | Intelligent weather forecast | |
Song et al. | A continuous space location model and a particle swarm optimization-based heuristic algorithm for maximizing the allocation of ocean-moored buoys | |
CN105574604A (en) | Power network operation event-oriented monitoring, pre-judging and analyzing system | |
CN114723131A (en) | Mine environment intelligent prediction system based on artificial intelligence | |
Qin et al. | Trajectory prediction based on long short-term memory network and Kalman filter using hurricanes as an example | |
Ahmed et al. | Hybrid deep learning model for wave height prediction in Australia's wave energy region | |
Yang et al. | Studies on fault diagnosis of dissolved oxygen sensor based on GA-SVM | |
Zhan et al. | A new prediction method for surface settlement of deep foundation pit in pelagic division based on Elman-Markov model | |
CN115169718A (en) | Cellular automaton-based regional landslide risk dynamic prediction method and device | |
Gai et al. | Research on water quality spatiotemporal forecasting model based on ST-BIGRU-SVR neural network | |
Turarbek et al. | 2-d deep convolutional neural network for predicting the intensity of seismic events | |
Rana et al. | A comparative analysis of ANN and ANFIS approaches for earthquake forecasting | |
Pape et al. | Multivariate analysis of nonlinearity in sandbar behavior | |
Baloyi et al. | Internet of things: opportunity for disaster risk reduction | |
Gokila et al. | Clustering and Classification in Support of Climatology to mine Weather Data–A Review | |
Greşiţă et al. | Expert system used for monitoring the behavior of hydrotechnical constructions | |
da Silva et al. | Enhancing hydroelectric inflow prediction in the Brazilian power system: A comparative analysis of machine learning models and hyperparameter optimization for decision support | |
Kanwar et al. | An Approach For Short Term Electricity Load Forecasting |
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 |