A mine ecological detection method and system based on big data analysis

Abstract

The present application provides a kind of mine ecological detection method and system based on big data analysis belongs to mine ecological monitoring technical field, the present application improves the precision and timeliness of mine ecological detection, realizes the accurate capture of multidimensional data;And for the defect that traditional detection relies on experience judgment cannot be quantified, based on historical data set training correlation analysis model of multiple sub-models, using LSTM neural network and DQN reinforcement learning algorithm training governance repair optimization model, by updating function trade-off immediate and future repair income;The stage dust source positioning result, dust dispersion prediction result and repair effect evaluation data set are input into the optimization model, to generate the optimization scheme including dust-settling equipment parameter adjustment and ecological repair auxiliary measures, realize the dynamic optimization closed loop from data monitoring to accurate governance;It can link detection results and repair measures, so as to realize the optimization of mine dust control strategy according to mine ecological feedback analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mine ecological monitoring, and particularly relates to a mine ecological detection method and system based on big data analysis. BACKGROUND

[0002] The drilling, blasting, ore crushing and ore dressing in the process of mine exploitation can produce a large amount of production dust (such as PM10 and PM2.5 with a diameter of less than 10 μm), which can cover the leaves of the vegetation and form a dust film after being diffused to the ecological restoration area by the wind, thereby hindering the sunlight irradiation to cause the photosynthesis efficiency to decrease, blocking the stomata of the leaves to affect the respiration, and long-term causing the vegetation to wither and fall off. Meanwhile, the dust deposited on the soil surface can change the pH value of the soil and reduce the air permeability, thereby affecting the seed germination and root growth, and the interference of the dust on the mine ecological restoration has the characteristics of strong concealment and long duration.

[0003] The traditional detection mainly relies on manual sampling and observation, and can only obtain the local dust concentration or the growth state of the vegetation. The manual sampling has a long cycle and a large error, and is difficult to capture the dynamic change of the dust. The dust source and diffusion trend are only determined by experience, the dust contribution rate of each operation link cannot be quantified, and the influence range of the dust on the ecological restoration area cannot be accurately predicted. Moreover, the existing system mainly has the data monitoring function, cannot link the detection result with the restoration measure, and is difficult to analyze and optimize the mine dust control strategy according to the feedback of the mine ecology. Therefore, the mine ecological detection method based on big data analysis is provided. SUMMARY

[0004] The present application provides a mine ecological detection method based on big data analysis, which can link the detection result with the restoration measure, thereby realizing the analysis and optimization of the mine dust control strategy according to the feedback of the mine ecology.

[0005] The technical scheme adopted by the present application is as follows:

[0006] A mine ecological detection method based on big data analysis comprises the following steps:

[0007] S1: obtaining historical basic data of mine ecological detection, wherein the historical basic data comprises historical dust data, historical associated ecological data and historical associated operation data;

[0008] S2: performing data calibration on the historical dust data, the historical associated ecological data and the historical associated operation data in step S1;

[0009] S3: training a big data analysis model based on the historical dust data, the historical associated ecological data and the historical associated operation data in step S2 to generate an associated analysis model, wherein the associated analysis model comprises:

[0010] The dust source positioning sub-model: a random forest algorithm is used to construct a multi-classification model by integrating decision trees, historical dust data and historical associated operation data are used as feature vectors to form a model training sample set, and the dust contribution rate of each operation link is output through the decision tree voting mechanism to realize the quantitative positioning of the dust source;

[0011] The dust diffusion prediction sub-model: a computational fluid dynamics numerical simulation method is used to construct a physical and mathematical coupling model combined with the atmospheric diffusion equation, historical dust data and historical associated operation data are input, boundary conditions are set, and the dust diffusion trajectory at different times is output by solving the continuity equation, momentum equation and diffusion equation;

[0012] The repair effect evaluation sub-model uses an LSTM neural network to build a regression evaluation model using its time series data processing capability, and uses the time series sequence of historical associated ecological data as training samples to label the repair effect level at the corresponding period;

[0013] S4: Real-time acquisition of dynamic data of mine ecological detection, dynamic data includes real-time dust data, real-time associated ecological data and real-time associated operation data, and 1 day is preset as 1 time period to be summarized and stored as a time period sub-data set;

[0014] S5: Summarize the time period sub-data set in step S4 in a 7-day cycle to generate a dynamic data set, and then calibrate the dynamic data set;

[0015] S6: Input the dynamic data set in step S5 into the associated analysis model generated in step S3 to generate the stage dust source positioning result, the stage dust diffusion prediction result and the stage repair effect evaluation data set in a 7-day cycle;

[0016] S7: Based on the stage repair effect evaluation data set in step S6, combined with the historical associated operation data in step S1, an LSTM neural network and a DQN reinforcement learning algorithm are used to train a governance and repair optimization model;

[0017] S8: Input the stage dust source positioning result, the stage dust diffusion prediction result and the stage repair effect evaluation data set in step S6 into the governance and repair optimization model in step S7 to generate a measure optimization scheme including dust reduction equipment parameter adjustment and ecological repair auxiliary measures.

[0018] Further, the real-time dust data includes TSP, PM2.5 and PM10 concentration data, which are obtained by a laser scattering dust sensor, the laser scattering dust sensor is arranged along the periphery of the mine repair area and within 50-100m of the dust source, and the distance between two adjacent laser scattering dust sensors is not more than 200m.

[0019] Further, in steps S2 and S4, the historical dust data, the historical associated ecological data, the historical associated operation data, the real-time dust data, the real-time associated ecological data and the real-time associated operation data are all converted into JSON format.

[0020] Further, the dust data calibration in step S2 includes:

[0021] The calibration sample is synchronously obtained by weight method every month, and is compared with the synchronous data of the laser scattering dust sensor to calculate the calibration coefficient K:

[0022] K = weight method determination value / sensor determination value, when When the value is greater than the calibration threshold, the historical dust data is corrected, and the corrected data = sensor original data * K.

[0023] Further, in step S4,

[0024] The real-time dust data is obtained by the laser scattering dust sensor deployed in the mine, and the data is transmitted to the edge computing node in real time;

[0025] The real-time associated ecological data is obtained by the mobile monitoring equipment once a day, covering all fixed sample plots in the restoration area.

[0026] The real-time associated operation data is synchronously obtained by the weather station, the equipment control system and the vehicle positioning system, and is automatically summarized into a time period sub-data set every day.

[0027] Further, in step S6,

[0028] The phased dust source positioning result outputs the dust contribution rate of each operation link in the form of a pie chart, and marks the key sources with a contribution rate of ≥20%;

[0029] The phased dust diffusion prediction result outputs the 7-day average diffusion range, marks the affected ecological restoration area and the diffusion direction.

[0030] Further, in step S1, the historical associated ecological data includes vegetation survival rate, leaf chlorophyll content, soil pH value and soil surface dust deposition amount, and the historical associated operation data includes wind speed, wind direction, mine daily mining amount, dust reduction measure intensity, historical restoration measure data and historical restoration effect data.

[0031] A mine ecological detection system based on big data analysis, comprising:

[0032] A first data acquisition module for acquiring historical basic data of mine ecological detection, the historical basic data including historical dust data, historical associated ecological data and historical associated operation data, and generating a historical data set;

[0033] The calibration module calibrates the historical dust data, the historical associated ecological data and the historical associated operation data in step S1.

[0034] The first generation module is configured to train a big data analysis model based on the historical data set and generate an associated analysis model.

[0035] The second data acquisition module is configured to acquire dynamic data of mine ecological detection in real time, wherein the dynamic data includes real-time dust data, real-time associated ecological data and real-time associated operation data, and 1 day is set as 1 time period for storage as a time period sub-data set.

[0036] The summary module is configured to summarize the time period sub-data set in step S4 in a 7-day cycle and generate a dynamic data set.

[0037] The second generation module is configured to input the dynamic data set into the associated analysis model generated in step S3 to generate a phased dust source positioning result, a phased dust diffusion prediction result and a phased repair effect evaluation data set in the 7-day cycle.

[0038] The third generation module is configured to train a governance and repair optimization model based on the phased repair effect evaluation data set and in combination with the historical associated operation data in step S1 by using an LSTM neural network and a DQN reinforcement learning algorithm.

[0039] The fourth generation module is configured to input the phased dust source positioning result, the phased dust diffusion prediction result and the phased repair effect evaluation data set into the governance and repair optimization model in step S7 to generate a measure optimization scheme including dust-settling equipment parameter adjustment and ecological repair auxiliary measures.

[0040] The present application has the following advantages:

[0041] The present application improves the precision and timeliness of mine ecological detection and realizes accurate capture of multi-dimensional data.

[0042] The phased dust source positioning result, the dust diffusion prediction result and the repair effect evaluation data set are then input into the optimization model to generate an optimization scheme including dust-settling equipment parameter adjustment and ecological repair auxiliary measures, thereby realizing dynamic optimization of a closed loop from data monitoring to accurate governance. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 A flow chart of a mine ecological detection method based on big data analysis for an embodiment of the present application;

[0044] Figure 2 A flow chart of historical data processing and correlation analysis model training for an embodiment of the present application;

[0045] Figure 3 A flow chart of dynamic data processing and stage analysis for an embodiment of the present application;

[0046] Figure 4 A flow chart of governance repair optimization model training and measure generation for an embodiment of the present application;

[0047] Figure 5 A block schematic diagram of a mine ecological detection system module for an embodiment of the present application. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0049] As shown in Figures 1-5 , a mine ecological detection method based on big data analysis according to an embodiment of the present application comprises the following steps:

[0050] S1: Obtain historical basic data for mine ecological detection, including historical dust data, historical associated ecological data, and historical associated operation data;

[0051] Specifically, the historical associated ecological data includes vegetation survival rate, leaf chlorophyll content, soil pH value, and soil surface dust deposition amount, and the historical associated operation data includes wind speed, wind direction, mine daily mining amount, dust control measure intensity, historical repair measure data, and historical repair effect data.

[0052] S2: Calibrate the historical dust data, historical associated ecological data, and historical associated operation data in step S1;

[0053] The historical dust data calibration comprises: synchronously obtaining calibration samples every month by using the gravimetric method, comparing the calibration samples with the same period data of the laser scattering dust sensor, and calculating a calibration coefficient When is greater than a threshold value , the historical dust data is corrected, and the corrected data ​

[0054] wherein, is ; : threshold value (preset 0.1, i.e. when the deviation from 1 exceeds 10%, triggering correction);

[0055] In addition, the historical associated ecological data and the historical associated operation data are data measured and stored in the system log, and only data type calibration is required. After the historical associated ecological data and the historical associated operation data are uniformly converted into a data type, the training of the big data analysis model is facilitated.

[0056] S3: training a big data analysis model based on the historical dust data, the historical associated ecological data and the historical associated operation data in step S2, to generate an associated analysis model, which includes:

[0057] A dust source positioning sub-model is generated by using a random forest algorithm. The random forest is composed of n decision trees (n is preset as 100), and the historical dust data feature vector ( TSP, PM2.5 and PM10 concentration and spatiotemporal distribution) and the historical associated operation data feature vector ( mining daily production and dust reduction measure intensity) are inputted, and the dust contribution rate of each operation link is outputted by integration, which is the output vector of the kth decision tree.

[0058] The historical dust data and the historical associated operation data in the historical data set are inputted, and the dust contribution rate of each operation link is outputted as the input of the stage source positioning result in step S6. When the sample amount of the historical data set is increased by more than 10% of the total sample amount, the model is automatically triggered for retraining.

[0059] A dust diffusion prediction sub-model is generated by using CFD numerical simulation. Based on the continuity equation , the momentum equation and the diffusion equation , the historical dust data and the environmental data are inputted, and the dust diffusion trajectory is outputted.

[0060] wherein, is the air density, is the time, is the real-time wind speed and direction, is the air pressure, is the air dynamic viscosity, is the gravity acceleration vector; is the dust concentration, is the dust diffusion coefficient;​​ The dust source item is calculated from the phased dust source positioning result of step S6. Among them, The PM2.5 is 1.5x10⁻ 5 m 2 / s, and the PM10 is 8x10⁻ 6 m 2 / s.

[0061] The real-time dust data and real-time associated operation data in the dynamic data set are input; and the dust diffusion trajectory, including the spatial coordinates (x, y, z) and the concentration distribution C(x, y, z, t) of time t and the 7-day average diffusion range, are output.

[0062] It should be noted that the time sequence of the historical associated ecological data includes the monthly change data of the vegetation survival rate, the leaf chlorophyll content, the soil pH value, and the soil surface dust deposition amount.

[0063] The LSTM neural network is used to generate a repair effect evaluation sub-model, a memory unit is constructed through a forgetting gate , an input gate and an output gate , the historical associated ecological data time sequence is input, and the 1~5 level repair effect grades are output through ;

[0064] Among them, the repair effect 1 level: vegetation survival rate < 50%, soil pH value < 5.0 or > 8.5, soil surface dust deposition amount > 500g / m 2 ;

[0065] Repair effect 2 level: vegetation survival rate 50%~69%, soil pH value 5.0~5.5 or 8.0~8.5, soil surface dust deposition amount 300~500g / m 2 ;

[0066] Repair effect 3 level: vegetation survival rate 70%~84%, soil pH value 5.5~6.5 or 7.0~8.0, soil surface dust deposition amount 100~300g / m 2 ;

[0067] Repair effect 4 level: vegetation survival rate 85%~94%, soil pH value 6.5~7.0, soil surface dust deposition amount 50~100g / m 2 ;

[0068] Repair effect 5 level: vegetation survival rate ≥ 95%, soil pH value 6.5~7.0, soil surface dust deposition amount < 50g / m 2 .

[0069] ​Specifically, input the associated ecological data time series in the historical data set or the dynamic data set ; output the stage evaluation results after generating the 7-day dynamic data set.

[0070] S4: Real-time acquisition of dynamic data of mine ecological detection, the dynamic data including real-time dust data, real-time associated ecological data and real-time associated operation data, and preset 1 day as 1 time period to be summarized and stored as a time period sub-data set;

[0071] S5: Summarize the time period sub-data set in step S4 in a 7-day cycle to generate a dynamic data set, and then perform data calibration on the dynamic data set;

[0072] Wherein, the data calibration on the dynamic data set is the same as the data calibration step in step S2.

[0073] S6: Input the dynamic data set in step S5 into the associated analysis model generated in step S3 to generate the stage dust source positioning results, the stage dust diffusion prediction results and the stage repair effect evaluation data set in the 7-day cycle;

[0074] Wherein, the stage dust source positioning results are output in the form of a pie chart to show the dust contribution rate of each operation link, and the key sources with a contribution rate of ≥20% are marked; the stage dust diffusion prediction results output the 7-day average diffusion range, and mark the affected ecological restoration area and the diffusion direction.

[0075] S7: Based on the stage repair effect evaluation data set in step S6, combining the historical associated operation data in step S1, using LSTM neural network and DQN reinforcement learning algorithm to train a governance and repair optimization model,

[0076] DQN updates the Q value function for the ecological state, for the repair measure, for the reward value; In the formula, is the learning rate, which controls the update amplitude,

[0077] is the discount factor, which weighs the immediate reward and the future reward; is the optimal action value of the next state. The governance and repair optimization model is trained by historical repair measure data and historical repair effect data, and iterates value, and finally selects the action with the maximum value as the optimal measure in the current state .

[0078] ​S8: input the results of the periodic dust source positioning, the results of the periodic dust diffusion prediction, and the periodic repair effect evaluation dataset in step S6 into the treatment and repair optimization model in step S7 to generate a measure optimization scheme including dust reduction equipment parameter adjustment and ecological repair auxiliary measures.

[0079] Specifically, the dynamic data of the mine ecology detection is acquired in real time in step S4, specifically including: acquiring TSP, PM2.5, and PM10 concentration data through a laser scattering dust sensor with a precision of 0.1 μg / m 3 The laser scattering dust sensor is arranged along the periphery of the mine repair area and within 50-100 m of the dust source, and the distance between two adjacent laser scattering dust sensors is not more than 200 m.

[0080] Specifically, the real-time acquisition of dynamic data in step S4 includes: real-time dust data is acquired through the Internet of Things sensor network deployed in the mine, the acquisition frequency is set to once every 2 minutes, and the data is transmitted to the edge computing node in real time; real-time associated ecological data is acquired once a day by a mobile monitoring device, covering all fixed sample plots in the repair area; real-time associated operation data is synchronously acquired through a weather station, a device control system, and a vehicle positioning system, and is automatically summarized into a time period sub-dataset every day.

[0081] According to the dust contribution rate from high to low, the sources with a contribution rate of ≥30% are preferentially processed, and the specific processing operation is to adjust the dust reduction equipment parameters to the upper limit: spray pressure 1.2 MPa, flow rate 20 L / min, and then process the sources with a contribution rate of 20%-29%, and adjust the parameters to the upper limit: pressure 0.8-1.0 MPa, flow rate 15-18 L / min. Synchronous supporting artificial dust cleaning measures.

[0082] In addition, the vegetation survival rate is reviewed by random sampling, for example, 30 plants are extracted from each sample plot, the deviation between the actual survival number and the number counted by the vegetation monitoring sensor is calculated, and when the deviation is >5%, the actual value is corrected.

[0083] The soil pH value is measured and calibrated by the laboratory potential method every month, and the on-site sensor data is compared, the calibration coefficient K is the laboratory value divided by the sensor value, and when K-1>0.05, it is corrected. The daily mining amount of the mine is compared by comparing the weighing data of the weighbridge and the statistical data of the equipment control system, and when the deviation is >3%, the weighing data of the weighbridge is used as the correction.

[0084] A mine ecology detection system based on big data analysis, comprising:

[0085] A first data acquisition module for acquiring historical basic data of mine ecology detection, the historical basic data including historical dust data, historical associated ecological data, and historical associated operation data, to generate a historical dataset;

[0086] The calibration module calibrates the historical dust data, the historical associated ecological data and the historical associated operation data in step S1.

[0087] The first generation module is configured to train a big data analysis model based on the historical data set and generate an associated analysis model.

[0088] The second data acquisition module is configured to acquire dynamic data of mine ecological detection in real time, wherein the dynamic data comprises real-time dust data, real-time associated ecological data and real-time associated operation data, and 1 day is set as 1 time period for storage as a time period sub-data set.

[0089] The summary module is configured to summarize the time period sub-data set in step S4 in a 7-day cycle to generate a dynamic data set.

[0090] The second generation module is configured to input the dynamic data set into the associated analysis model generated in step S3 to generate a phased dust source positioning result, a phased dust diffusion prediction result and a phased repair effect evaluation data set in the 7-day cycle.

[0091] The third generation module is configured to train a governance and repair optimization model based on the phased repair effect evaluation data set and in combination with the historical associated operation data in step S1 by using an LSTM neural network and a DQN reinforcement learning algorithm.

[0092] The fourth generation module is configured to input the phased dust source positioning result, the phased dust diffusion prediction result and the phased repair effect evaluation data set into the governance and repair optimization model in step S7 to generate a measure optimization scheme comprising dust-settling equipment parameter adjustment and ecological repair auxiliary measures.

[0093] In the description of the present application, the terms "first", "second" are only used for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. The meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0094] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific feature, structure, material or characteristic being described in connection with the embodiment or example is included in at least one embodiment or example of the present application. Descriptive terms of the above-mentioned terms do not necessarily refer to the same embodiment or example in the specification. Also, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples, without contradiction.

[0095] Any process or method descriptions or descriptions of the flow of processes or methods described in flow diagrams or otherwise described herein can be understood as representing code modules, segments, or portions of code which include one or more executable instructions for implementing specific logic functions (or steps) of the process or method. The various embodiments of the present application can include additional or fewer steps or processes or methods, and the order of those that are included can be different from that shown. It is also possible for some steps or portions of the described processes or methods to be executed in parallel. It is also possible for some steps or portions of the described processes or methods to be executed in reverse order.

[0096] The logic and / or steps represented in flow diagrams or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be specifically embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, apparatus or device and execute the instructions. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport a program for use by or in connection with an instruction execution system, apparatus or device or in conjunction with these instruction execution systems, apparatus or devices. More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connections having one or more wires (electronic devices), portable computer disks (magnetic devices), random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memories), fiber optic devices, and portable compact disc read-only memories (CD ROMs). In addition, a computer-readable medium can even be paper or other suitable medium on which a program can be printed, as the program can be electronically obtained, for example, by optical scanning of the paper or other medium, followed by electronic conversion of the optical scanning into a form that can be edited, interpreted or otherwise processed as appropriate, and then stored in a computer storage medium.

[0097] It should be understood that parts of the present application can be implemented in hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and as in another embodiment, it can be implemented using any or a combination of the following technologies, which are well known in the art: discrete logic circuitry having logic gates for implementing logic functions upon data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0098] Those skilled in the art can understand that all or part of the steps carried out by the above-mentioned embodiment methods can be completed by programs instructing relevant hardware, and the programs can be stored in a computer readable storage medium, and when the programs are executed, one or a combination of the steps of the method embodiments is included.

[0099] In addition, each functional unit in each embodiment of the present application can be integrated into one processing module, or each unit can exist physically, or two or more units can be integrated into one module. The above integrated module can be realized in the form of hardware or in the form of a software functional module. The integrated module, if realized in the form of a software functional module and sold or used as an independent product, can also be stored in a computer readable storage medium.

[0100] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A mine ecological detection method based on big data analysis, characterized in that, Comprise the following steps: S1: obtain the historical basic data of mine ecological detection, the historical basic data includes historical dust data, historical associated ecological data, historical associated operation data; S2: the historical dust data, historical associated ecological data, historical associated operation data in step S1 are calibrated; S3: based on the historical dust data, historical associated ecological data, historical associated operation data in step S2, train a big data analysis model, generate an associated analysis model, the associated analysis model includes: Dust source positioning submodel: adopt random forest algorithm, construct multi-classification model through integrated decision tree, use historical dust data and historical associated operation data as feature vector, form model training sample set, output dust contribution rate of each operation link through decision tree voting mechanism, realize quantitative positioning of dust source; Dust diffusion prediction submodel: adopt computational fluid dynamics numerical simulation method, combine atmospheric diffusion equation to construct physical and mathematical coupling model, input historical dust data and historical associated operation data, set boundary conditions, output dust diffusion trajectory at different time periods by solving continuity equation, momentum equation and diffusion equation; Repair effect evaluation submodel, adopt LSTM neural network, utilize its time series data processing capability to construct regression evaluation model, use time series sequence of historical associated ecological data as training sample, label repair effect grade of corresponding period; S4: real-time acquisition of dynamic data of mine ecological detection, dynamic data includes real-time dust data, real-time associated ecological data and real-time associated operation data, preset 1 day as 1 period to be stored as period sub-data set; S5: the period sub-data set in step S4 in 7-day cycle is summarized, a dynamic data set is generated, and then the dynamic data set is calibrated; S6: the dynamic data set in step S5 is input into the associated analysis model generated in step S3, and stage dust source positioning result, stage dust diffusion prediction result and stage repair effect evaluation data set in 7-day cycle are generated; S7: based on the stage repair effect evaluation data set in step S6, combined with the historical associated operation data in step S1, adopt LSTM neural network and DQN reinforcement learning algorithm to train governance and repair optimization model; S8: the stage dust source positioning result, the stage dust diffusion prediction result and the stage repair effect evaluation data set in step S6 are input into the governance and repair optimization model in step S7, and a measure optimization scheme including dust reduction equipment parameter adjustment and ecological repair auxiliary measures is generated. 2.The mine ecological detection method based on big data analysis according to claim 1, characterized in that, The real-time dust data includes TSP, PM2.5 and PM10 concentration data, which are obtained by a laser scattering dust sensor, the laser scattering dust sensor is arranged along the periphery of the mine repair area and within 50-100m of the dust source, the distance between two adjacent laser scattering dust sensors is not more than 200m. 3.The mine ecological detection method based on big data analysis according to claim 1, characterized in that, In steps S2 and S4, the historical dust data, historical associated ecological data, historical associated operation data and real-time dust data, real-time associated ecological data and real-time associated operation data are all converted into JSON format. 4.The mine ecological detection method based on big data analysis according to claim 2, characterized in that, The dust data calibration in step S2 includes: Synchronously obtain calibration samples every month by weight method, compare with the synchronous data of the laser scattering dust sensor, and calculate the calibration coefficient K: K = gravimetrically determined value / sensor determined value, when correcting the historical dust data, corrected data = sensor raw data x K, when K > calibration threshold value. 5.The mine ecological detection method based on big data analysis according to claim 2, characterized in that, In step S4, Real-time dust data is obtained by the laser scattering dust sensor deployed in the mine, and the data is transmitted to the edge computing node in real time; Real-time associated ecological data is obtained once a day by a mobile monitoring device, covering all fixed sample plots in the restoration area; Real-time associated operation data is synchronously obtained by a weather station, a device control system, and a vehicle positioning system, and is automatically summarized into a time period sub-data set every day. 6.The mine ecological detection method based on big data analysis according to claim 1, characterized in that, In step S6, The phase dust source positioning result is output in the form of a pie chart, and the key sources with a contribution rate of more than or equal to 20% are marked; The phase dust diffusion prediction result outputs the 7-day average diffusion range, and marks the affected ecological restoration area and the diffusion direction. 7.The mine ecological detection method based on big data analysis of claim 1, wherein, In step S1, the historical associated ecological data includes vegetation survival rate, leaf chlorophyll content, soil pH value, and soil surface dust deposition amount, and the historical associated operation data includes wind speed, wind direction, mine daily mining amount, dust suppression measure intensity, historical restoration measure data, and historical restoration effect data.

8. A mine ecological detection system based on big data analysis, applied to the method of any one of claims 1-7, characterized in that, It includes: A first data acquisition module is configured to acquire historical basic data of mine ecological detection, wherein the historical basic data includes historical dust data, historical associated ecological data, and historical associated operation data, and a historical data set is generated; A calibration module is configured to calibrate the historical dust data, the historical associated ecological data, and the historical associated operation data in step S1; A first generation module is configured to train a big data analysis model based on the historical data set, and generate an associated analysis model; A second data acquisition module is configured to acquire dynamic data of mine ecological detection in real time, wherein the dynamic data includes real-time dust data, real-time associated ecological data, and real-time associated operation data, and 1 day is preset as a time period to be summarized and stored as a time period sub-data set; A summary module is configured to summarize the time period sub-data set in step S4 in a 7-day cycle, and generate a dynamic data set; A second generation module is configured to input the dynamic data set into the associated analysis model generated in step S3, and generate phase dust source positioning result, phase dust diffusion prediction result, and phase restoration effect evaluation data set in the 7-day cycle; A third generation module is configured to train a governance and restoration optimization model based on the phase restoration effect evaluation data set and the historical associated operation data in step S1, using an LSTM neural network and a DQN reinforcement learning algorithm; A fourth generation module is configured to input the phase dust source positioning result, the phase dust diffusion prediction result, and the phase restoration effect evaluation data set into the governance and restoration optimization model in step S7, and generate a measure optimization scheme including dust suppression device parameter adjustment and ecological restoration auxiliary measures.

Images