Intelligent decision-making method for vegetation planting of opencast coal mine in sandstorm beach area

Abstract

The invention discloses an intelligent decision-making method for vegetation planting of an opencast coal mine in a sandstorm beach area. The method comprises the following steps: constructing a vegetation planting file data set of the open pit coal mine in the sandstorm beach area; constructing an experience knowledge base of vegetation planting of the open pit coal mine in the sandstorm beach area; the AlexNet convolutional neural network is optimized on the basis of an ANPL-layer model, and an ANPL-Net convolutional neural network model is obtained; and based on an ANPL-Net convolutional neural network model, in combination with the archive data set and the experience knowledge base, performing decision making on vegetation planting of the open pit coal mine. The objective of the invention is to overcome the technical problems in the prior art, improve the reliability and scientificity of vegetation planting decision, make up for the deficiency of a single experience method, and provide a basis for selection of open pit coal mine planting in a sandstorm grass beach area.

Description

Technical Field

[0001] The present invention relates to the field of mining area ecology, and in particular to an intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas. Background Art

[0002] Sand dunes in the windy and sandy grassland area are distributed continuously over a large area, the vegetation is seriously degraded, the land is becoming more and more desertified, the groundwater level is unstable, and the ecological environment is extremely fragile; in open-pit coal mines, water and soil resources are seriously scarce, and the area of ​​soil erosion is huge, causing serious damage to the ecological environment. Existing studies have shown that vegetation restoration in open-pit coal mines has a positive effect on environmental governance. It can be seen that it is very important to seek vegetation planting methods for open-pit coal mines in windy and sandy grassland areas. The existing planting decision cases show the following defects:

[0003] (1) In terms of vegetation planting decision-making, first of all, the intelligence performance is insufficient, and most of them use expert experience or directly draw on the experience of predecessors; secondly, some intelligent decision-making methods are mainly based on decision trees and gray theory. Through research, it is found that the volume of data processed and the accuracy of processing results are far lower than the algorithm capability of neural networks; finally, in the decision-making process of neural networks, its weight initialization method is mainly random initialization and transfer learning initialization, while random initialization without any prior knowledge increases the tediousness of model training; the initialization method of using pre-trained models for transfer learning has certain constraints on the model structure, and the weight distribution law needs to be extracted so that the model weight has a higher probability of approaching the global optimal solution.

[0004] (2) In terms of the dataset construction of vegetation planting in open-pit coal mines in sandy grassland areas, there is a lack of long-term monitoring of planting area data, and the vegetation growth characteristics and regional adaptability cannot be scientifically evaluated.

[0005] (3) The results of vegetation planting decisions are mainly based on vegetation variety decisions, which is not comprehensive enough. It is also necessary to supplement vegetation planting density, pest and disease control, irrigation and fertilization, etc. Summary of the invention

[0006] In view of this, the present invention provides an intelligent decision-making method for open-air vegetation planting in sandy grassland areas. The method of the present invention improves the reliability and scientificity of vegetation planting decisions, makes up for the shortcomings of a single empirical method, and provides a basis for the selection of open-air coal mine planting in sandy grassland areas.

[0007] Specifically, the present application is implemented through the following technical solutions:

[0008] A first aspect of the present invention provides an intelligent decision-making method for vegetation planting in an open-pit coal mine in a sandy grassland area, the method comprising:

[0009] Construct a dataset of vegetation planting archives in open-pit coal mines in sandy grassland areas;

[0010] Constructing an experience knowledge base for vegetation planting in open-pit coal mines in sandy grassland areas;

[0011] Optimize the AlexNet convolutional neural network based on the ANPL-layer model to obtain the ANPL-Net convolutional neural network model;

[0012] Based on the ANPL-Net convolutional neural network model, combined with archival datasets and empirical knowledge base, decisions are made on vegetation planting in open-pit coal mines.

[0013] According to one embodiment of the present invention, a method for constructing a vegetation planting archive dataset for open-pit coal mines in a sandy grassland area includes:

[0014] Collect ecological and environmental factor data of open-pit coal mine areas and remote sensing image data of planting areas;

[0015] Construct a data acquisition network based on Internet of Things technology based on ecological environment factor data and remote sensing image data;

[0016] Through the data acquisition network, the data of each platform is transmitted in real time and stored centrally to obtain real-time data;

[0017] The environmental data are sorted and separated based on real-time data, and a vegetation planting archive dataset of open-pit coal mines in sandy grassland areas is generated.

[0018] According to one embodiment of the present invention, a method for optimizing the AlexNet convolutional neural network based on the ANPL-layer model includes:

[0019] Calculate the number of weights, kernel size, and number of channels for each layer in the AlexNet convolutional neural network;

[0020] Optimize the weight distribution of AlexNet convolutional neural network based on symmetric power law distribution function;

[0021] Generate optimized weights based on the ANPL-layer model and apply them to the convolutional layer of the AlexNet convolutional neural network;

[0022] The convolutional layers of the optimized AlexNet convolutional neural network are reintegrated to form the ANPL-Net convolutional neural network model.

[0023] According to one embodiment of the present invention, a method for calculating the number of weights, convolution kernel size, and number of channels of each layer in an AlexNet convolutional neural network includes:

[0024] Statistics of the convolution structure of the AlexNet convolutional neural network and the corresponding number of weights are calculated as follows:

[0025] ki w=ks×ks×c i ×s i ;

[0026] Among them, k i w represents the weight of the convolution kernel of the i-th layer, ks represents the size of the convolution kernel, c i Represents the number of layer channels, s i Represents the number of convolution kernels in the layer;

[0027] All data of the convolutional layer K of the AlexNet convolutional neural network are represented as:

[0028] K i ={δ(Ni,C i , H i , W i )|(1≤N i ≤n i , 1≤C i ≤c i , i≤H i ≤h i , 1≤W i ≤w i )}

[0029] Among them, K i Represents the weight data of the i-th convolutional layer in the network model, n i represents the number of convolution kernels in the i-th convolutional layer, c i represents the number of channels of the convolution kernel in the i-th convolutional layer, (h i , w i ) represents the format of the convolution kernel in the i-th convolutional layer.

[0030] According to one embodiment of the present invention, a method for optimizing the weight distribution of an AlexNet convolutional neural network based on a symmetric power-law distribution function includes:

[0031] The power law distribution test is performed on the weight data of each layer, based on the weight distribution law obtained from the pre-training weight study, and the power law distribution function is normalized. The specific formula is as follows:

[0032]

[0033] Based on the derived standardized symmetric power-law distribution function, the ANPL-layer convolutional layer initialization method for the AlexNet convolutional neural network is constructed.

[0034] According to one embodiment of the present invention, the ANPL-layer initialization method includes the following steps:

[0035] When the number of weights required by the convolutional layer is num, for the Xi value, num*Yi data are generated accordingly;

[0036] Based on the random number generation interval (-r, r), the data with num*Yi value of Xi is processed with random number fluctuation to ensure that there is no continuous identical data in the layer Data generated by the ANPL-layer algorithm;

[0037] According to the format of the current i-th convolutional layer (n i , c i ,h i , w i ), format the weight data generated by the model;

[0038] The final output data of the ANPL-layer model algorithm is: (n i , c i ,h i , w i )’s tensor data layer Data.

[0039] According to one embodiment of the present invention, the AlexNet convolutional neural network includes five convolutional layers and three connection layers.

[0040] According to one embodiment of the present invention, based on the ANPL-Net convolutional neural network model, combined with the archive data set and the experience knowledge base, the method for making decisions on vegetation planting in open-pit coal mines includes:

[0041] The ANPL-Net convolutional neural network model is used to decide vegetation planting varieties, vegetation planting density, planting area, water and fertilizer capacity, pest and disease control plan, and irrigation and fertilization plan.

[0042] A second aspect of the present invention provides a non-transitory computer-readable storage medium storing a computer program, which, when executed by a processor, implements a method for intelligent decision-making on vegetation planting in open-pit coal mines in sandy grassland areas.

[0043] The third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements a method for intelligent decision-making for vegetation planting in open-pit coal mines in sandy grassland areas when executing the computer program.

[0044] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0045] (1) According to the characteristics of the open-pit coal mine planting area, a mining area planting data acquisition network is constructed. Based on the Internet of Things technology, various sensors and implementation monitoring systems are linked. The data of each platform are finally concentrated on the data network created by the present invention by network data capture, forming an archival data set framework for open-pit coal mines in windy and sandy grassland areas, so that vegetation planting decisions have a more complete and real-time data set, providing better data support for subsequent decisions, and improving the regional characteristics and accuracy of the convolutional neural network input data in decision-making.

[0046] (2) The functions and weights of the convolutional layer are relatively independent. The overall weight of the convolutional layer in the AlexNet convolutional neural network is taken as the research object, and the ANPL-layer model is constructed according to its symmetric power-law distribution function. Then the ANPL-layer model is used to initialize the weights of the AlexNet convolutional neural network to form the ANPL-Net convolutional neural network model, so that it can achieve faster convergence and good generalization ability.

[0047] (3) In terms of vegetation planting decision-making, the present invention uses the ANPL-Net convolutional neural network to make decisions on the selection of vegetation varieties, planting density, pest and disease control, and irrigation and fertilization, making the results of vegetation planting intelligent decision-making systematic and comprehensive, thereby improving the comprehensiveness, integrity, and referenceability of the decision-making results.

[0048] (4) The present invention can make up for the defects of subjectivity based on expert experience, and can make decisions on vegetation planting in open-pit coal mine areas quickly and in real time, thereby improving the effectiveness and accuracy of decision-making results, which is of great significance for reducing the cost of planting and ecological restoration of mining areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of an intelligent decision-making method for planting vegetation in an open-pit coal mine in a sandy grassland area provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of the overall technical route provided for an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the ANPL-Net structure of the present invention;

[0052] Figure 4 A schematic diagram of the planting data acquisition network technology route of the present invention;

[0053] Figure 5 Schematic diagram of the internal structure of the convolutional layer of the AlexNet model of the present invention;

[0054] Figure 6 This is a schematic diagram of ANPL-layer initializing the weights of AlexNet convolutional layer; DETAILED DESCRIPTION

[0055] The present invention is further described in detail below in conjunction with embodiments:

[0056] Example

[0057] According to one embodiment of the present invention, Figure 1-6 For the method of intelligent decision-making for vegetation planting in open-pit coal mines in windy and sandy grassland areas of the present invention, refer to Figure 1 As shown, the method of the present invention comprises:

[0058] Construct a dataset of vegetation planting archives in open-pit coal mines in sandy grassland areas;

[0059] Constructing an experience knowledge base for vegetation planting in open-pit coal mines in sandy grassland areas;

[0060] Optimize the AlexNet convolutional neural network based on the ANPL-layer model to obtain the ANPL-Net convolutional neural network model;

[0061] Based on the ANPL-Net convolutional neural network model, combined with archival datasets and empirical knowledge base, decisions are made on vegetation planting in open-pit coal mines.

[0062] Specifically, the method for constructing the vegetation planting archive data set of open-pit coal mines in windy and sandy grassland areas includes: collecting ecological and environmental factor data of the open-pit coal mine area and remote sensing image data of the planting area;

[0063] Construct a data acquisition network based on Internet of Things technology based on ecological environment factor data and remote sensing image data;

[0064] Through the data acquisition network, the data of each platform is transmitted in real time and stored centrally to obtain real-time data;

[0065] The environmental data are sorted and separated based on real-time data, and a vegetation planting archive dataset of open-pit coal mines in sandy grassland areas is generated.

[0066] Specifically, it is mainly aimed at collecting and acquiring data such as ecological environmental factors of open-pit coal mines and remote sensing images of planting areas, and establishing a mining area planting data acquisition network, which is specifically based on the Internet of Things technology, links various sensors and implementation monitoring systems, and uses network data capture to finally concentrate the data of each platform on the data network created by the present invention, forming an archival data set of open-pit coal mines in windy and sandy grassland areas.

[0067] Investigate the parameters required for vegetation planting in open-pit coal mine areas, and collect and acquire data such as ecological environmental factors and remote sensing images of planting areas. First, build a mining area planting data acquisition network, which is specifically based on the Internet of Things technology, links various sensors and implementation monitoring systems, and uses network data capture to finally concentrate the data of each platform on the data network created by the present invention, forming an archive data set framework for open-pit coal mines in windy and sandy grassland areas.

[0068] Preferably, a meteorological detection system is linked to obtain meteorological data, including but not limited to the following meteorological data information, such as temperature, humidity, wind speed, wind direction, air quality, air pressure, light and heat conditions, rainfall and snowfall, ultraviolet rays, evaporation, surface radiation, negative oxygen ions, carbon monoxide, ozone, oxygen, carbon dioxide and other air component contents; the above monitored information is stored by name, and is summarized and classified according to a certain numerical range to reduce the confusion caused by the large amount of data for subsequent analysis, and classified data is used.

[0069] Preferably, a soil detection system is linked to collect soil data of open-pit coal mines in sandy grassland areas, including but not limited to soil moisture, soil temperature, soil texture, soil component content (sodium, nitrogen, potassium, organic matter), and soil pH.

[0070] Preferably, remote sensing images are used in combination with drones. The basic conditions of the planting area can be observed on a large scale through remote sensing image data, and the vegetation coverage, land use and land use type transfer conditions of the area can be inverted through analysis; drones can improve the accuracy of regional observations, and the topographic conditions, surrounding environment, surrounding water-containing areas, existing vegetation coverage, planted vegetation types and vegetation growth conditions of the proposed planting area can be determined through drone photography.

[0071] The methods for constructing the experience knowledge base of vegetation planting in open-pit coal mines in sandy grassland areas include:

[0072] The empirical knowledge is based on the regional characteristics expressed by the acquisition of basic data, and is widely collected for vegetation planting case data with similar regional ecological and environmental conditions. At the same time, a complete knowledge base management mechanism is established to combine the knowledge base with empirical cases to form a more complete and reliable reference data set.

[0073] The method of optimizing the AlexNet convolutional neural network based on the ANPL-layer model to obtain the ANPL-Net convolutional neural network model includes:

[0074] As a classic network, AlexNet convolutional neural network has been widely used in a large number of data sets. First, the AlexNet convolutional neural network is predicted. The input of the AlexNet convolutional neural network is an image of 224×224×3 size with three RGB channels. In this embodiment, the AlexNet convolutional neural network includes 5 convolutional layers (including 3 pooling) and 3 fully connected layers. The 1st, 2nd, and 5th convolutional layers are followed by a maximum pooling layer, and the last three layers are fully connected layers. Finally, the network output is converted into a probability value. For the AlexNet convolutional neural network, in order to facilitate the rough extraction of the weight information of its pre-trained model, the convolution structure of the convolutional neural network and the corresponding number of weights are counted. The specific calculation is as follows:

[0075] k i w=ks×ks×c i ×s i ;

[0076] where k i w represents the weight of the convolution kernel of the i-th layer, ks represents the size of the convolution kernel, c i Represents the number of channels in this layer, s i Represents the number of convolution kernels in this layer.

[0077] All data of the convolutional layer K of the AlexNet convolutional neural network are represented as:

[0078] K i ={δ(N i , C i , H i , W i )|(1≤N i ≤n i , 1≤C i ≤c i , i≤H i ≤h i , 1≤W i ≤w i )};

[0079] Among them, K i Represents the weight data of the i-th convolutional layer in the network model, n i represents the number of convolution kernels in the i-th convolutional layer, c i represents the number of channels of the convolution kernel in the i-th convolutional layer, (h i , w i ) represents the format of the convolution kernel in the i-th convolutional layer.

[0080] Then, the power law distribution test is performed on the weight data of each layer. Based on the weight distribution law obtained from the pre-training weight study, the power law distribution function is further normalized. The specific formula is as follows:

[0081]

[0082] Based on the derived standardized symmetric power-law distribution function, the ANPL-layer convolutional layer initialization method for the AlexNet convolutional neural network is constructed. The main creation principle of the ANPL-layer initialization data generation algorithm is: when the number of weights required for a certain convolutional layer is num, for the Xi value, num*Yi data should be generated accordingly. The detailed algorithm process can be summarized as follows:

[0083] First, based on the function bundle (X, Y), we can calculate the data needcout that each Xi should generate. However, relative to the weight, it is not advisable to have a large number of continuous and identical data for initialization. Therefore, the random parameter r is introduced. Based on the random number generated with a value interval of (-r, r), the num*Yi data with the value of Xi are subjected to random number fluctuation processing to ensure that there is no continuous identical data in the data layer Data generated by the ANPL-layer algorithm. Finally, according to the format of the current i-th convolutional layer (n i , c i ,h i , w i ), format the weight data generated by the model. Therefore, the final output data of the ANPL-layer model algorithm should be: the format is (n i , c i ,h i , w i )’s tensor data layer Data.

[0084] Based on the ANPL-Net convolutional neural network model, combined with archival data sets and empirical knowledge base, the decision-making methods for vegetation planting in open-pit coal mines include:

[0085] Based on the ANPL-Net convolutional neural network model, the meteorological data, soil data, image data, and expert experience and knowledge of the open-pit mine area are first cropped and input in a size of 224×224×3. The optimized ANPL-Net convolutional neural network model is used to make decisions on vegetation varieties, and the geographical conditions, climate conditions, and vegetation growth conditions of the open-pit coal mine in the sandy grassland area are analyzed in detail. According to the actual situation, the appropriate varieties are determined. In the process of using the convolutional neural network model, it is not only necessary to use the weight to initialize the model, but also to assign weight parameters to the various types of data collected according to the size of the impact on the vegetation growth conditions and continuously adjust them. And the parameters of the model can be adjusted and optimized according to different decision-making methods. The following lists several different conditions based on which the varieties are selected. And the parameters of the model can be adjusted and optimized according to different decision-making methods. The following lists several different conditions based on which the varieties are selected:

[0086] According to the geographical decision of the planting area, the climatic conditions show great differences in the length of the frost-free period, sunshine time, and heat resources due to the geographical latitude and altitude. Therefore, when making decisions on vegetation varieties, it is necessary to fully understand the latitude and altitude conditions of the source of the vegetation. Select varieties with similar conditions to the open-pit coal mine distribution area in the sandy grassland area, and the vegetation growth period is close to or shorter than the frost-free period of the planting area. Reduce the difficulty of vegetation growth due to differences in the geographical conditions required for growth, resulting in economic losses.

[0087] The variety is determined based on the ecological conditions of the planting area; ecological conditions refer to the environmental conditions required for vegetation growth, including rainfall conditions, evaporation, light and heat resources, soil inorganic components, soil water and fertilizer conditions, regional geological morphology, etc. Varieties suitable for the ecological conditions of open-pit coal mines in windy and sandy grassland areas should be selected as much as possible, and vegetation with strong resistance to barrenness and drought should be considered.

[0088] The species is determined based on the growth characteristics of vegetation; different vegetation types have different growth characteristics, especially in terms of sensitivity to light conditions. They are usually divided into types that require long sunshine and types that require short sunshine. When choosing a vegetation type, you should understand the sunshine conditions required by the vegetation to avoid failure in planting. Pay attention to the impact of regional changes in vegetation introduction. Vegetation that requires long sunshine hours will have a longer growth period from north to south, and vice versa.

[0089] Preferably, the ANPL-Net convolutional neural network model is used to decide the planting density; the planting area and water and fertilizer capacity of the open-pit coal mine in the sandy grassland area are analyzed in detail. The planting density needs to fully consider the actual regional conditions of the open-pit coal mine in the sandy grassland area, and a multi-dimensional analysis is performed to assign reasonable weight values ​​to the decision factors.

[0090] The following principles can be used to make decisions on planting density: 1) Decisions based on climate type; vegetation wind and sand grassland areas and open-pit coal mines are concentrated in semi-arid and arid areas, with little rainfall concentrated in summer, a long frost-free period, and a high planting density is appropriate. 2) Decisions based on soil and water and fertilizer conditions; semi-arid areas have high evaporation, and the mining of open-pit coal mines has destroyed the structure and texture of the soil, resulting in poor fertility conditions, so a high density is appropriate. 3) Decisions based on vegetation species; if the vegetation is tall, a high density is appropriate; if the vegetation is short, the planting density can be appropriately reduced. 4) Decisions based on vegetation cultivation techniques and measures; early sowing time means high density; late sowing time means low density; spring sowing density means high density, and summer sowing density means low density.

[0091] Preferably, the ANPL-Net convolutional neural network model is used to make decisions on pest and disease control. Pest and disease decision-making also requires the use of deep learning models. Overall, the climate characteristics and regional conditions in the planting archive data set are required as basic input data. At the same time, monitoring points are established to comprehensively supervise and obtain real-time data, which can be summarized and analyzed in a timely manner to determine the occurrence characteristics of pests and diseases. In the future, it is necessary to continuously improve the pest and disease data archive information and establish a complete electronic archive or data set. Develop a digital monitoring system for pests and diseases: use intelligent technologies and systems, such as climate analyzers, imagers and other equipment, to monitor pests and diseases. Through data analysis and computer technology, a complete pest and disease prediction function is established to realize networked prediction and forecasting work.

[0092] Preferably, the ANPL-Net convolutional neural network model is used for irrigation and fertilization decisions. First, the transmitted data is monitored and analyzed in real time, including vegetation growth information and planting area conditions. At the same time, a comprehensive evaluation is conducted based on data such as rainfall, vegetation water consumption, soil fertility and water content. Irrigation decision: According to the results of data analysis, scientific irrigation decisions are made based on the ANPL-Net convolutional neural network model, including irrigation water volume, fertilizer amount and irrigation timing.

[0093] Archival datasets constructed based on data crawling networks, expert experience knowledge, and an intelligent planting decision-making model based on ANPL-Net.

[0094] The method of the present invention first constructs a planting archive data set; collects and acquires data such as ecological environmental factors of open-pit coal mines and remote sensing images of planting areas, and establishes a mining area planting data acquisition network, which is specifically based on the Internet of Things technology, links various sensors and implementation monitoring systems, and uses network data capture to finally concentrate the data of each platform on the data network created by the present invention to form an archive data set of open-pit coal mines in windy and sandy grassland areas. Empirical knowledge mainly refers to the collection of existing similar planting cases and expert knowledge fields in order to combine multi-dimensional analysis with environmental data; the vegetation planting decision function specifically makes decisions from the perspectives of planting varieties, planting density, pest and disease control, and irrigation and fertilization. The main idea of ​​the planting decision part is to use the ANPL-Net convolutional neural network model proposed in the present invention to make decisions on vegetation planting, propose an ANPL-layer weight initialization method to optimize the AlexNet convolutional neural network, make planting decisions based on this and the above data, and output the optimal solution for the decision results.

[0095] The present invention also provides a non-transitory computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, a method for intelligent decision-making on vegetation planting in open-pit coal mines in sandy grassland areas is implemented.

[0096] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a method for intelligent decision-making for vegetation planting in open-pit coal mines in sandy grassland areas is implemented.

[0097] Those of ordinary skill in the art will appreciate that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0098] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and equipment can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0099] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0100] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0101] In addition, each functional module in the embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0102] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the energy-saving signal sending / receiving method of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

[0103] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the present application.

[0104] It should be understood that the size of the sequence number of each step in the content of the invention and the embodiments of the present invention does not absolutely mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. For the purpose of example and description, the foregoing description of the implementation of the present disclosure has been given. The foregoing description is not exhaustive and is not intended to limit the present disclosure to the exact form disclosed. Various deformations and modifications may exist according to the above teachings, or various deformations and modifications may be obtained from the practice of the present disclosure. These embodiments are selected and described to illustrate the principles of the present disclosure and its practical application, so that those skilled in the art can use the present disclosure in various embodiments and various modifications suitable for the specific purpose conceived.

Claims

1. An intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas, characterized in that: include: Construct a dataset of vegetation planting archives in open-pit coal mines in sandy grassland areas; Constructing an experience knowledge base for vegetation planting in open-pit coal mines in sandy grassland areas; Optimize the AlexNet convolutional neural network based on the ANPL-layer model to obtain the ANPL-Net convolutional neural network model; Based on the ANPL-Net convolutional neural network model, combined with archival datasets and empirical knowledge base, decisions are made on vegetation planting in open-pit coal mines.

2. According to the intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas as claimed in claim 1, it is characterized by: The methods for constructing the vegetation planting archive dataset of open-pit coal mines in sandy grassland areas include: Collect ecological and environmental factor data of open-pit coal mine areas and remote sensing image data of planting areas; Construct a data acquisition network based on Internet of Things technology based on ecological environment factor data and remote sensing image data; Through the data acquisition network, the data of each platform is transmitted in real time and stored centrally to obtain real-time data; The environmental data are sorted and separated based on real-time data, and a vegetation planting archive dataset of open-pit coal mines in sandy grassland areas is generated.

3. The intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas according to claim 1 is characterized by: Methods for optimizing AlexNet convolutional neural network based on ANPL-layer model include: Calculate the number of weights, kernel size, and number of channels for each layer in the AlexNet convolutional neural network; Optimize the weight distribution of AlexNet convolutional neural network based on symmetric power law distribution function; Generate optimized weights based on the ANPL-layer model and apply them to the convolutional layer of the AlexNet convolutional neural network; The convolutional layers of the optimized AlexNet convolutional neural network are reintegrated to form the ANPL-Net convolutional neural network model.

4. According to the intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas as described in claim 3, it is characterized in that: Methods for calculating the number of weights, convolution kernel size, and number of channels in each layer of the AlexNet convolutional neural network include: Statistics of the convolution structure of the AlexNet convolutional neural network and the corresponding number of weights are calculated as follows: k i w=ks×ks×c i ×s i ; Among them, k i w represents the weight of the convolution kernel of the i-th layer, ks represents the size of the convolution kernel, c i Represents the number of layer channels, s i Represents the number of convolution kernels in the layer; All data of the convolutional layer K of the AlexNet convolutional neural network are represented as: K i ={δ(Ni,C i ,H i ,IN i )|(1≤N i ≤n i ,1≤C i ≤c i ,1≤H i ≤h i ,1≤W i ≤in i )}; Among them, K i Represents the weight data of the i-th convolutional layer in the network model, n i represents the number of convolution kernels in the i-th convolutional layer, c i represents the number of channels of the convolution kernel in the i-th convolutional layer, (h i , w i ) represents the format of the convolution kernel in the i-th convolutional layer.

5. According to the intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas as described in claim 3, it is characterized in that: The method of optimizing the weight distribution of AlexNet convolutional neural network based on symmetric power law distribution function includes: The power law distribution test is performed on the weight data of each layer, based on the weight distribution law obtained from the pre-training weight study, and the power law distribution function is normalized. The specific formula is as follows: Based on the derived standardized symmetric power-law distribution function, the ANPL-layer convolutional layer initialization method for the AlexNet convolutional neural network is constructed.

6. The intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas according to claim 3 is characterized in that: The ANPL-layer initialization method includes the following steps: When the number of weights required by the convolutional layer is num, for the Xi value, num*Yi data are generated accordingly; Based on the random number generation interval (-r, r), the data with num*Yi value of Xi is processed with random number fluctuation to ensure that there is no continuous identical data in the layer Data generated by the ANPL-layer algorithm; According to the format of the current i-th convolutional layer (n i , c i ,h i , w i ), format the weight data generated by the model; The final output data of the ANPL-layer model algorithm is: (n i , c i ,h i , w i )’s tensor data layerData.

7. The intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas according to claim 1 is characterized by: The AlexNet convolutional neural network includes five convolutional layers and three connection layers.

8. The intelligent decision-making method for vegetation planting in open-pit coal mines in sandy grassland areas according to claim 1 is characterized by: Based on the ANPL-Net convolutional neural network model, combined with archival data sets and empirical knowledge base, the decision-making methods for vegetation planting in open-pit coal mines include: The ANPL-Net convolutional neural network model is used to decide vegetation planting varieties, vegetation planting density, planting area, water and fertilizer capacity, pest and disease control plan, and irrigation and fertilization plan.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores a computer program, which, when executed by a processor, implements a method for intelligent decision-making for vegetation planting in an open-pit coal mine in a sandy grassland area as described in any one of claims 1 to 8.

10. An electronic device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for intelligent decision-making on vegetation planting in open-pit coal mines in sandy grassland areas as described in any one of claims 1 to 8.

Images