Mineral exploration method, mineral exploration system and mineral exploration robot

Abstract

The invention provides a mineral exploration method, a mineral exploration system and a mineral exploration robot, which are applied to a mineral exploration robot, and the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool and is in communication connection with an external upper computer. The method comprises the following steps: acquiring an exploration point image of a target exploration point through visual perception equipment; inputting the exploration point image into a pre-trained geological type identification model, and outputting an actual geological type corresponding to the exploration point image; determining a target sampling tool corresponding to the actual geological type based on the actual geological type and a preset geological type and sampling tool corresponding relationship, and collecting a sample through the target sampling tool; and performing real-time analysis on the collected sample through analysis equipment to obtain real-time exploration data of the target exploration point. In the mode, the geological type can be automatically identified in a complex terrain, a proper sampling tool is selected, and sample components are analyzed in real time, so that the safety and the efficiency of exploration work are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mineral exploration, and in particular to a mineral exploration method, a mineral exploration system and a mineral exploration robot. Background Art

[0002] In the process of mineral resource exploration, prospectors usually need to identify geological types and collect samples in the target area to obtain mineral composition and content data. The existing survey equipment often requires operators to operate on-site, and the difficulty and complexity of operating the survey equipment are relatively high. While affecting the survey efficiency, it also makes the survey operation easily interfered by natural environmental factors, affecting the versatility and environmental adaptability of the survey operation. Summary of the invention

[0003] In view of this, the object of the present invention is to provide a mineral exploration method, a mineral exploration system and a mineral exploration robot, which can automatically identify geological types in complex terrain, select appropriate sampling tools, and analyze sample composition in real time, thereby improving the safety and efficiency of exploration work.

[0004] In a first aspect, an embodiment of the present invention provides a mineral exploration method, which is applied to a mineral exploration robot, wherein the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool, and is communicatively connected to a host computer of an external device; the method comprises: obtaining an exploration point image of a target exploration point through a visual perception device; inputting the exploration point image into a pre-trained geological type recognition model, and outputting the actual geological type corresponding to the exploration point image; determining a target sampling tool corresponding to the actual geological type based on the correspondence between the actual geological type and a preset geological type and a sampling tool, and collecting samples through the target sampling tool; performing real-time analysis on the collected samples through an analysis device to obtain real-time exploration data of the target exploration point.

[0005] Furthermore, the mineral exploration robot is also provided with a positioning device; before the step of obtaining the exploration point image of the target exploration point through the visual perception device, the method also includes: obtaining a coordinate file sent by the host computer; the coordinate file includes multiple exploration points to be measured, the exploration sequence and the longitude and latitude coordinates, elevation data and geological environment remarks corresponding to each exploration point to be measured; obtaining the current position of the mineral exploration robot collected by the positioning device, and setting the current position as the starting point; obtaining the exploration status of each exploration point to be measured, and based on the exploration sequence, determining the next exploration point to be measured with an exploration status of unfinished exploration as the target exploration point; based on the longitude and latitude coordinates, planning a first exploration path from the starting point to the target exploration point; and driving to the target exploration point based on the first exploration path.

[0006] Furthermore, after the step of planning a first exploration path from a starting point to a target exploration point based on longitude and latitude coordinates, the method further includes: acquiring a road condition image sent by a visual perception device; inputting the road condition image into a pre-trained road condition type recognition model, and outputting a road condition analysis result corresponding to the road condition image; identifying the road condition analysis result; if the road condition analysis result shows that an obstacle exists, updating the first exploration path based on a preset obstacle avoidance algorithm; if the road condition analysis result shows that abnormal terrain exists, controlling the running speed based on preset adjustment rules.

[0007] Furthermore, after the step of performing real-time analysis on the collected samples by the analysis equipment to obtain the real-time exploration data of the target exploration point, the method also includes: S1: after obtaining the real-time exploration data of the target exploration point, the target exploration point and the real-time exploration data corresponding to the target exploration point are associated and saved, and uploaded to the host computer; S2: the exploration status of the target exploration point is changed to completed exploration; S3: based on the exploration sequence, the exploration point to be tested whose next exploration status is unfinished exploration is determined as the next target exploration point; S4: based on the longitude and latitude coordinates, a second exploration path from the target exploration point to the next target exploration point is planned; and based on the second exploration path, the vehicle travels to the next target exploration point; S5: repeating steps S1-S4 until the exploration status of each exploration point to be tested is completed exploration.

[0008] Furthermore, the mineral exploration robot is also provided with a power supply device; the method also includes: identifying real-time battery information collected by the power supply device; the real-time battery information includes battery power and battery health status; if it is identified that the battery power is lower than a preset return threshold, return to the starting point; if it is identified that the battery health status is not within the preset battery health threshold range, return to the starting point.

[0009] Furthermore, the geological type recognition model is trained by the following method: obtaining a historical geological image data set; the historical geological image data set includes geological images and geological type annotation information corresponding to the geological images; dividing the historical geological image data set into a training set, a verification set and a test set according to a preset ratio; training an initial convolutional neural network model based on the training set until the preset training requirements are met to obtain a first geological type recognition model; optimizing the first geological type recognition model based on a preset verification algorithm to obtain a second geological type recognition model; verifying the second geological type recognition model based on the verification set until the preset verification requirements are met to obtain a geological type recognition model.

[0010] In a second aspect, an embodiment of the present invention provides a mineral exploration system, which is applied to a mineral exploration robot. The mineral exploration robot is provided with a visual perception device, an analysis device, a sampling tool and a positioning device, and is communicatively connected to a host computer of an external device; the system includes: an exploration point image acquisition module, which is used to obtain an exploration point image of a target exploration point through a visual perception device; a geological type determination module, which is used to input the exploration point image into a pre-trained geological type recognition model, and output the actual geological type corresponding to the exploration point image; a sampling module, which is used to determine the target sampling tool corresponding to the actual geological type based on the actual geological type and the preset correspondence between the geological type and the sampling tool, and collect samples through the target sampling tool; an exploration data generation module, which is used to perform real-time analysis on the collected samples through an analysis device to obtain real-time exploration data of the target exploration point.

[0011] In the third aspect, an embodiment of the present invention provides a mineral exploration robot, comprising: a carrying platform and a control device, a visual perception device, an analysis device, a sampling tool and a positioning device arranged on the carrying platform; the visual perception device, the analysis device, the sampling tool and the positioning device are respectively connected to the control device; and also includes the above-mentioned mineral exploration system, which is arranged in the control device.

[0012] Furthermore, the mineral exploration robot also includes a mechanical arm and a guide mechanism; the mechanical arm is arranged on one side of the carrying platform; and the guide mechanism is arranged below the carrying platform.

[0013] Furthermore, the sampling tools include a bulldozer, a small drill and a sweeper; the sampling tools, the visual perception device and the analysis device are respectively arranged at the front end of the robotic arm; and the analysis device is a laser induced breakdown spectrometer.

[0014] The embodiment of the present invention provides a mineral exploration method, a mineral exploration system and a mineral exploration robot, which are applied to the mineral exploration robot, wherein the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool, and is connected to a host computer of an external device in communication; the method comprises: obtaining an exploration point image of a target exploration point through a visual perception device; inputting the exploration point image into a pre-trained geological type recognition model, and outputting the actual geological type corresponding to the exploration point image; determining the target sampling tool corresponding to the actual geological type based on the actual geological type and the preset geological type and sampling tool correspondence, and collecting samples through the target sampling tool; and performing real-time analysis on the collected samples through an analysis device to obtain real-time exploration data of the target exploration point. In this way, the geological type can be automatically identified in complex terrain, a suitable sampling tool can be selected, and the sample composition can be analyzed in real time, thereby improving the safety and efficiency of the exploration work.

[0015] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0016] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A flow chart of a mineral exploration method provided in Embodiment 1 of the present invention;

[0019] Figure 2 A flow chart of a method for planning an exploration route for a target exploration point provided in the first embodiment of the present invention;

[0020] Figure 3 A flow chart of an obstacle avoidance method provided in Embodiment 1 of the present invention;

[0021] Figure 4 A flow chart of a geological type identification model training method provided in the first embodiment of the present invention;

[0022] Figure 5 A flow chart of a method for planning a next target exploration point provided in Embodiment 1 of the present invention;

[0023] Figure 6 A flow chart of a power supply equipment detection method provided in Embodiment 1 of the present invention;

[0024] Figure 7 A schematic diagram of a mineral exploration system provided in Embodiment 2 of the present invention;

[0025] Figure 8 A schematic diagram of another mineral exploration system provided in Embodiment 2 of the present invention;

[0026] Fig. 9 This is a schematic diagram of a mineral exploration robot provided in Example 3 of the present invention.

[0027] Icons: 1-power supply equipment; 2-guiding mechanism; 3-analysis equipment; 4-sampling tool; 5-visual perception equipment; 6-communication equipment; 7-positioning equipment; 8-control equipment; 9-carrying platform; 10-mechanical arm; 11-exploration point image acquisition module; 12-geological type determination module; 13-sampling module; 14-exploration data generation module; 15-battery information acquisition module. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] To facilitate understanding of this embodiment, the embodiment of the present invention is described in detail below.

[0030] Embodiment 1:

[0031] Figure 1 This is a flow chart of the mineral exploration method provided in Example 1 of the present invention.

[0032] The mineral exploration method is applied to the mineral exploration robot, which is provided with a visual perception device, an analysis device and a sampling tool, and is connected to a host computer of the peripheral device in communication. The mineral exploration robot is also provided with a positioning device.

[0033] Visual perception devices include cameras and lidar sensors, etc.

[0034] Sampling tools include dozers, small drills and sweepers.

[0035] The analysis equipment is a laser induced breakdown spectrometer.

[0036] The mineral exploration robot is connected to the host computer of the peripheral device through the antenna.

[0037] Reference Figure 1 , mineral exploration methods include:

[0038] Step S101: acquiring an exploration point image of a target exploration point through a visual perception device.

[0039] Here, when the mineral exploration robot drives to the target exploration point, the visual camera installed at the front captures the on-site ground image and determines that the captured ground image is the exploration point image.

[0040] In one embodiment, referring to Figure 2 Before step S101, the method further includes:

[0041] Step S201, obtaining a coordinate file sent by a host computer; the coordinate file includes a plurality of exploration points to be measured, an exploration sequence, and the longitude and latitude coordinates, elevation data, and geological environment remark information corresponding to each exploration point to be measured.

[0042] Here, before the exploration mission begins, the exploration team determines several target exploration points in the target area through remote sensing images and preliminary geological surveys. The coordinate data of these exploration points are sorted into coordinate files, and the coordinate files are uploaded to the host computer. After the mineral exploration robot is turned on, it obtains the coordinate file sent by the host computer.

[0043] Step S202, obtaining the current position of the mineral exploration robot collected by the positioning device, and setting the current position as the starting point.

[0044] Here, the positioning device may adopt GPS (Global Positioning System) and RTK (Realtime kinematic, carrier phase difference technology) and the like.

[0045] When the mission starts, the system will set the current geographic coordinates as the starting point of the exploration mission, which will serve as the starting point for path planning.

[0046] Step S203, obtaining the exploration status of each to-be-tested exploration point, and determining the next to-be-tested exploration point whose exploration status is unfinished exploration as the target exploration point based on the exploration sequence.

[0047] Here, the coordinate file also includes the exploration status (completed or uncompleted) of each exploration point to be tested. The mineral exploration robot reads this status information and automatically determines the next exploration point to be explored according to the preset exploration sequence, and sets it as the current target exploration point.

[0048] Step S204: planning a first exploration path from the starting point to the target exploration point based on the longitude and latitude coordinates; and driving to the target exploration point based on the first exploration path.

[0049] Here, the mineral exploration robot converts the environmental data of the starting point into a grid map, and divides the exploration area into several grid cells, each of which represents different terrain features, such as passable areas, obstacle areas, high-risk areas, etc. Each cell of the grid map is assigned a different weight or cost value, reflecting the difficulty of passing different terrains or obstacles. Obstacle areas are marked as inaccessible, and flat areas have a lower cost.

[0050] Based on the preset path planning algorithm, according to the latitude and longitude coordinates of the starting point and the target exploration point, an optimal driving path (based on the shortest distance, safety, obstacle avoidance requirements, etc.) is calculated and planned to ensure that the robot can reach the target location efficiently. The path planning algorithm can be A* search algorithm (A star algorithm), Dijkstra algorithm (Dijkstra algorithm) and DWA (Dynamic Wi ndow Approach, dynamic window method), etc.

[0051] According to the planned path, the wheels, motors and navigation system are controlled to move along the predetermined route. During the driving process, the mineral exploration robot will use the real-time data sent by the camera and sensors to make real-time adjustments to ensure that it reaches the target exploration point safely.

[0052] Specifically, in a mineral exploration mission in a mountainous area, the exploration team needs to explore multiple points within a designated area.

[0053] When the exploration mission starts, the exploration team sends the pre-designed coordinate file to the mineral exploration robot through the host computer. The file contains the longitude and latitude coordinates, elevation data, geological environment notes and the exploration order of each point of the 10 exploration points to be tested.

[0054] The mineral exploration robot starts the positioning device, obtains the longitude and latitude coordinates of the current location (for example, 34.056 degrees north latitude, 118.245 degrees east longitude), and sets this point as the starting point as the starting position of the exploration mission.

[0055] The mineral exploration robot reads the status of each exploration point in the coordinate file and finds that the first point (point A) is incomplete. According to the exploration sequence, the robot automatically sets point A as the current target exploration point.

[0056] The mineral exploration robot uses a path planning algorithm to calculate the optimal driving path based on the latitude and longitude data of the starting point (current location) and the target point (Point A), taking into account factors such as obstacles and slopes on the path to ensure safe arrival.

[0057] The mining exploration robot starts to drive automatically according to the planned path. During the journey, the visual perception device and obstacle avoidance camera monitor the road conditions ahead in real time. When encountering obstacles, the system automatically adjusts the route and continues to move forward after bypassing the obstacles.

[0058] In one embodiment, referring to Figure 3 In step S204, after planning a first exploration path from the starting point to the target exploration point based on the longitude and latitude coordinates, the method further includes:

[0059] Step S301, obtaining a road condition image sent by a visual perception device.

[0060] Here, during the driving process of the mineral exploration robot, the visual perception device (such as a camera or lidar) continuously captures the road conditions ahead and obtains image data in real time. These images contain terrain information, obstacles, road conditions, etc. on the current path, which serve as basic data for judging the driving environment.

[0061] Step S302: input the road condition image into a pre-trained road condition type recognition model, and output the road condition analysis result corresponding to the road condition image.

[0062] Here, the acquired road condition image is input into a pre-trained road condition type recognition model. The road condition type recognition model has been trained with a large amount of road condition image data and can accurately identify various road condition types contained in the image, such as flat areas, obstacles, muddy sections, ramps, etc. After the model analysis, it outputs the corresponding road condition analysis results, such as "obstacles exist" or "abnormal terrain".

[0063] Among them, the road condition type recognition model is trained by the following method.

[0064] A historical road condition image dataset is obtained; the historical road condition image dataset includes road condition images and road condition annotation information corresponding to the road condition images.

[0065] The historical road condition image dataset is divided into a training set, a validation set, and a test set according to a preset ratio.

[0066] An initial convolutional neural network model is trained based on the training set until preset training requirements are met to obtain a first road condition type recognition model.

[0067] Based on a preset verification algorithm, the first road condition type recognition model is optimized to obtain a second road condition type recognition model.

[0068] The second road condition type recognition model is verified based on the verification set until the preset verification requirements are met to obtain the road condition type recognition model.

[0069] During the model training process, the hyperparameters are adjusted through the validation set to prevent overfitting. Finally, the recognition accuracy of the model is evaluated through the test set to ensure its ability to recognize unknown data.

[0070] The trained road condition type recognition model can output the corresponding road condition analysis results in real time based on the input road condition image.

[0071] Step S303, identifying the road condition analysis result.

[0072] Here, the mineral exploration robot parses and judges the road condition analysis results output by the model to identify the specific road conditions on the current driving path, such as whether there are obstacles (such as stones, branches, etc.) or special terrain (such as steep slopes, muddy areas).

[0073] Step S304: If the road condition analysis result shows that there is an obstacle, the first exploration path is updated based on a preset obstacle avoidance algorithm.

[0074] Here, if an obstacle is identified on the path, the optimal path to bypass the obstacle is calculated based on the preset obstacle avoidance algorithm, and the driving route is updated in real time to ensure that the mineral exploration robot can safely avoid the obstacle and continue driving. The preset obstacle avoidance algorithm can be DWA or artificial potential field method, etc.

[0075] Step S305: If the road condition analysis result shows that there is abnormal terrain, the running speed is controlled based on a preset adjustment rule.

[0076] Here, if the road condition analysis results show that there is abnormal terrain ahead, the driving speed is automatically adjusted based on the preset adjustment rules. For example, the speed is reduced before a steep slope to ensure safe climbing, or the speed is reduced on a slippery road to prevent skidding and loss of control.

[0077] Among them, the preset adjustment rules can also include adjusting the center of gravity, etc., to prevent the mineral exploration robot from overturning by adjusting the center of gravity.

[0078] Specifically, on the way to point A, the visual perception device captured a series of road condition images, showing that there was a rock obstacle and a muddy ground about 5 meters ahead.

[0079] These road condition images are input into the road condition type recognition model in real time to identify the presence of rocks and the size of muddy areas, and output the results as the presence of obstacles and abnormal terrain.

[0080] Receive and analyze these analysis results, clearly identify obstacles (rocks) and special terrain (muddy sections) ahead, and make corresponding adjustments to the driving strategy.

[0081] Since the analysis results showed that there were rock obstacles, the optimal detour path was calculated based on the obstacle avoidance algorithm, and the original driving route was updated, so that the mineral exploration robot shifted 1 meter to the left to bypass the rock while maintaining a safe distance.

[0082] After identifying a muddy road section, the system automatically reduces the speed from the normal 4km / h to 2km / h according to the preset adjustment rules to prevent the vehicle from getting stuck in the mud or slipping. At the same time, the torque output of the wheels is adjusted to improve the vehicle's passing ability.

[0083] Step S102: input the exploration point image into a pre-trained geological type recognition model, and output the actual geological type corresponding to the exploration point image.

[0084] Here, the actual geological type may be rock or soil or a mixture of rock and soil.

[0085] If the exploration point image shows a mixture of gravel and soil of varying sizes, the geological type recognition model identifies that the image features best match the rock and soil mixture category in the training data. Therefore, the output result is a mixture of rock and soil. If most of the exploration point image shows a hard, reflective surface, the geological type recognition model identifies that the image features best match the rock in the training data, and the output result is rock. If the exploration point image shows a loose, irregular granular structure, the geological type recognition model identifies that the image features best match the soil in the training data, and the output result is soil.

[0086] In one embodiment, referring to Figure 4 , the geological type recognition model is trained by the following method:

[0087] Step S401, obtaining a historical geological image dataset; the historical geological image dataset includes geological images and geological type annotation information corresponding to the geological images.

[0088] Here, the exploration personnel collected a large amount of historical geological image data from multiple geological exploration missions to form a historical geological image data set, including different types of geological images such as soil, rock, and mineral mixtures. These geological image data are annotated by geologists and classified into different geological type labels, such as "sandstone", "granite", "mineral-rich soil", etc.

[0089] The collected geological images are preprocessed, including image size standardization, denoising, contrast enhancement, etc., to improve the efficiency of model training and recognition effect. At the same time, data enhancement techniques (such as rotation, flipping, brightness adjustment, etc.) are used to increase the diversity of the data set and improve the generalization ability of the model.

[0090] Step S402, dividing the historical geological image data set into a training set, a validation set and a test set according to a preset ratio.

[0091] Here, the preset ratio can be pre-set according to actual conditions, and can be set to training set: validation set: training set = 7:2:1.

[0092] Step S403: training an initial convolutional neural network model based on the training set until a preset training requirement is met to obtain a first geological type recognition model.

[0093] Here, the model structure of the initial convolutional neural network model includes several convolutional layers, pooling layers and fully connected layers.

[0094] The processed image data is input into the initial convolutional neural network model, and the model parameters are optimized through the back propagation algorithm so that the initial convolutional neural network model can learn the characteristics of different geological types from the image features. The cross entropy loss function and Adam optimizer are used to accelerate convergence.

[0095] Step S404: Based on a preset verification algorithm, the first geological type identification model is optimized to obtain a second geological type identification model.

[0096] Here, during the model training process, the hyperparameters are adjusted through the validation set to prevent overfitting. Finally, the recognition accuracy of the model is evaluated through the test set to ensure its ability to recognize unknown data.

[0097] Step S405, verifying the second geological type identification model based on the verification set until the preset verification requirements are met to obtain the geological type identification model.

[0098] Here, the trained geological type recognition model can output the corresponding geological type in real time according to the input geological image.

[0099] Step S103: Based on the actual geological type and the preset correspondence between the geological type and the sampling tool, a target sampling tool corresponding to the actual geological type is determined, and samples are collected by the target sampling tool.

[0100] Here, the mineral exploration robot includes a carrying platform and a mechanical arm, wherein the mechanical arm is arranged on one side of the carrying platform, and a sampling tool, a visual perception device and an analysis device are arranged at the front end of the mechanical arm respectively.

[0101] If the actual geological type is identified as rock, the robotic arm will switch to a small drill to drill samples for in-depth analysis.

[0102] If the actual geological type is identified as soil, the robotic arm switches to a dozer or sweeper to collect surface samples.

[0103] If the actual geological type is identified as a mixture of rock and soil, the mineral exploration robot will decide whether to use a drill or a bulldozer based on the main components of the sample, while cleaning the surface for further testing.

[0104] Step S104: Analyze the collected samples in real time by using analysis equipment to obtain real-time exploration data of the target exploration point.

[0105] Here, the mineral exploration robot collects soil or rock samples at the target exploration point through the corresponding sampling tools. The collected samples are collected in the sample slot or sampling bin of the analysis equipment for elemental composition analysis.

[0106] After sampling, the laser induced breakdown spectrometer performs rapid and non-destructive elemental composition testing on the sample. The laser induced breakdown spectrometer emits high-energy laser pulses to irradiate the sample surface, excite plasma and generate spectral signals.

[0107] The excited plasma will release light with a specific wavelength, and the laser-induced breakdown spectrometer captures these light signals and converts them into electrical signals. These signals are analyzed and the type and content of each element in the sample are identified through the built-in analysis model.

[0108] The laser-induced breakdown spectrometer processes the acquired spectral data and matches it with the built-in standard spectral database to obtain the specific content of each element in the sample through calculation. The final output results include real-time exploration data of the target exploration point, such as the type and concentration of the element and potential mineral composition information.

[0109] The mineral exploration robot will upload the real-time exploration data obtained through analysis to the host computer for geologists to review and make decisions, assisting geologists to quickly determine the value of the mining area and determine subsequent exploration strategies.

[0110] In one embodiment, referring to Figure 5 After step S104, the method further includes:

[0111] Step S1: After acquiring the real-time exploration data of the target exploration point, the target exploration point and the real-time exploration data corresponding to the target exploration point are associated and saved, and uploaded to the host computer.

[0112] Step S2: Change the exploration status of the target exploration point to completed exploration.

[0113] Here, the status of the target exploration point is automatically updated from unfinished exploration to completed exploration. The exploration progress is recorded in real time to avoid duplication of work and ensure that each exploration point to be tested is only explored once.

[0114] Step S3: Based on the exploration sequence, determine the next exploration point to be measured whose exploration status is unfinished as the next target exploration point.

[0115] According to the preset exploration sequence, the next exploration point to be tested with an unfinished exploration status will be automatically found. This point will be set as the target point for the next exploration, ensuring that the exploration work is gradually advanced in the established sequence and covers all exploration points to be tested.

[0116] Step S4: planning a second exploration path from the target exploration point to the next target exploration point based on the longitude and latitude coordinates; and driving to the next target exploration point based on the second exploration path.

[0117] Here, the best driving route is calculated by the path planning algorithm based on the latitude and longitude coordinates of the current target exploration point and the next target exploration point. Factors such as terrain and obstacles are taken into account during the planning process to ensure safe and efficient driving. After the planning is completed, the vehicle moves to the next target exploration point along the calculated path.

[0118] Step S5: Repeat steps S1-S4 until the exploration status of each exploration point to be tested is that the exploration has been completed.

[0119] Here, after reaching the next target exploration point, the steps of sample analysis, data saving and uploading, status update, target point confirmation and path planning are continued. This cycle will continue until the exploration status of all the exploration points to be tested changes to completed exploration, marking the end of this exploration task.

[0120] In one embodiment, the mineral exploration robot is also provided with a power supply device. Figure 6 , the method further comprises:

[0121] Step S601, identifying real-time battery information collected by the power supply device; the real-time battery information includes battery power and battery health status.

[0122] Here, the initial power of the mining exploration robot is fully charged, and the current power is automatically recorded and the minimum power required for the return trip is estimated. When the mining exploration robot is driving and performing exploration tasks, it continuously obtains real-time battery information of the power supply equipment and reads the battery power every few seconds. At the same time, it monitors the battery health status (such as temperature and voltage) to ensure that the equipment operates within a safe power range.

[0123] Step S602: If it is identified that the battery power is lower than the preset return threshold, return to the starting point.

[0124] Here, the preset return threshold is used to represent the minimum return power. For example, when the power is only 25% and it is just enough to return safely, the minimum return power threshold is set to 25%.

[0125] The amount of electricity required for the return trip is dynamically estimated based on the current location, the number of exploration points completed, and the number of remaining task points. Once the current electricity is detected to be close to or below the estimated preset return threshold, an early warning will be triggered to control the mineral exploration robot to return to the starting point.

[0126] When the battery power is lower than the preset return threshold, the mining exploration robot terminates the current mission, automatically abandons the next exploration point it is heading to, plans the shortest route, and starts the return journey. During the return journey, the robot reduces high-power actions (such as obstacle avoidance and quick turns) to save energy.

[0127] During the return journey, the battery power is continuously monitored to ensure that it remains within a safe range. If obstacles or complex terrain are found during the return journey, the robot automatically adjusts the path to ensure that it returns to the starting point in the most power-saving way. Finally, the mineral exploration robot successfully returns and docks at the charging platform at the starting point, completing a safe return.

[0128] Step S603: If it is determined that the battery health status is not within the preset battery health threshold range, return to the starting point.

[0129] Here, the battery health threshold range is preset according to actual conditions.

[0130] If the battery health status is not within the preset battery health threshold, the mineral exploration robot terminates the current mission, automatically abandons the next exploration point it is heading to, plans the shortest route, and starts the return journey.

[0131] The embodiment of the present invention provides a mineral exploration method, which is applied to a mineral exploration robot, wherein the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool, and is connected to a host computer of an external device in communication; the method comprises: obtaining an exploration point image of a target exploration point through a visual perception device; inputting the exploration point image into a pre-trained geological type recognition model, and outputting the actual geological type corresponding to the exploration point image; determining a target sampling tool corresponding to the actual geological type based on the actual geological type and the preset geological type and sampling tool correspondence, and collecting samples through the target sampling tool; and performing real-time analysis on the collected samples through an analysis device to obtain real-time exploration data of the target exploration point. In this method, the geological type can be automatically identified in complex terrain, a suitable sampling tool can be selected, and the sample composition can be analyzed in real time, thereby improving the safety and efficiency of the exploration work.

[0132] Embodiment 2:

[0133] Figure 7 Schematic diagram of a mineral exploration system provided in Embodiment 2 of the present invention.

[0134] The mineral exploration system is applied to the above-mentioned mineral exploration robot.

[0135] Reference Figure 7 , the mineral exploration system includes:

[0136] The exploration point image acquisition module 11 is used to acquire the exploration point image of the target exploration point through a visual perception device.

[0137] The geological type determination module 12 is used to input the exploration point image into a pre-trained geological type recognition model and output the actual geological type corresponding to the exploration point image.

[0138] The sampling module 13 is used to determine the target sampling tool corresponding to the actual geological type based on the actual geological type and the preset correspondence between the geological type and the sampling tool, and collect samples through the target sampling tool.

[0139] The exploration data generation module 14 is used to perform real-time analysis on the collected samples through analysis equipment to obtain real-time exploration data of the target exploration point.

[0140] In one embodiment, the exploration point image acquisition module 11 is further used for:

[0141] Obtain the coordinate file sent by the host computer; the coordinate file includes multiple exploration points to be measured, the exploration sequence and the longitude and latitude coordinates, elevation data and geological environment notes corresponding to each exploration point to be measured.

[0142] Get the current position of the mineral exploration robot collected by the positioning device and set the current position as the starting point.

[0143] The exploration status of each exploration point to be tested is obtained, and based on the exploration sequence, the next exploration point to be tested whose exploration status is unfinished is determined as the target exploration point.

[0144] Based on the longitude and latitude coordinates, a first exploration path from the starting point to the target exploration point is planned; and the vehicle travels to the target exploration point based on the first exploration path.

[0145] In one embodiment, the exploration point image acquisition module 11 is further used for:

[0146] Obtain the road condition image sent by the visual perception device.

[0147] The road condition image is input into a pre-trained road condition type recognition model, and the road condition analysis result corresponding to the road condition image is output.

[0148] Identify traffic analysis results.

[0149] If the road condition analysis result shows that there is an obstacle, the first exploration path is updated based on a preset obstacle avoidance algorithm.

[0150] If the road condition analysis results show that there is abnormal terrain, the operating speed is controlled based on the preset adjustment rules.

[0151] In one embodiment, the exploration data generating module 14 is further configured to:

[0152] S1: After acquiring the real-time exploration data of the target exploration point, the target exploration point and the real-time exploration data corresponding to the target exploration point are associated and saved, and uploaded to the host computer.

[0153] S2: Change the exploration status of the target exploration point to completed exploration.

[0154] S3: Based on the exploration sequence, determine the next exploration point to be tested whose exploration status is not completed as the next target exploration point.

[0155] S4: planning a second exploration path from the target exploration point to the next target exploration point based on the longitude and latitude coordinates; and driving to the next target exploration point based on the second exploration path.

[0156] S5: Repeat steps S1-S4 until the exploration status of each exploration point to be tested is that the exploration is completed.

[0157] In one embodiment, referring to Figure 8 The mineral exploration system further includes: a battery information acquisition module 15; the battery information acquisition module 15 is used to:

[0158] Identify real-time battery information collected by power supply equipment; real-time battery information includes battery power and battery health status.

[0159] If the battery level is detected to be lower than the preset return threshold, the vehicle returns to the starting point.

[0160] If it is determined that the battery health status is not within the preset battery health threshold, return to the starting point.

[0161] In one embodiment, the geological type determination module 12 is further used to:

[0162] A historical geological image dataset is obtained; the historical geological image dataset includes geological images and geological type annotation information corresponding to the geological images.

[0163] The historical geological image dataset is divided into training set, validation set and test set according to preset proportions.

[0164] The initial convolutional neural network model is trained based on the training set until the preset training requirements are met to obtain the first geological type recognition model.

[0165] Based on a preset verification algorithm, the first geological type identification model is optimized to obtain a second geological type identification model.

[0166] The second geological type identification model is verified based on the verification set until the preset verification requirements are met to obtain the geological type identification model.

[0167] An embodiment of the present invention provides a mineral exploration system, which is applied to a mineral exploration robot. In this way, the geological type can be automatically identified in complex terrain, appropriate sampling tools can be selected, and the sample composition can be analyzed in real time, thereby improving the safety and efficiency of the exploration work.

[0168] Embodiment three:

[0169] Fig. 9This is a schematic diagram of a mineral exploration robot provided in Example 3 of the present invention.

[0170] Reference Fig. 9 The mineral exploration robot includes: a carrying platform 9 and a control device 8, a visual perception device 5, an analysis device 3, a sampling tool 4 and a positioning device 7 arranged on the carrying platform 9; the visual perception device 5, the analysis device 3, the sampling tool 4 and the positioning device 7 are respectively connected to the control device 8; and also includes the above-mentioned mineral exploration system (not shown in the figure), and the mineral exploration system is arranged in the control device 8.

[0171] In one embodiment, the mineral exploration robot further includes a mechanical arm 10 and a guide mechanism 2 ; the mechanical arm 10 is disposed on one side of the carrying platform 9 ; and the guide mechanism 2 is disposed below the carrying platform 9 .

[0172] The mineral exploration robot further includes a power supply device 1 and a communication device 6. The power supply device 1 and the communication device 6 are arranged on a carrying platform 9. The communication device 6 is used for communication connection with a host computer of an external device.

[0173] In one embodiment, the sampling tool 4 includes a bulldozer, a small drill and a sweeper; the sampling tool 4, the visual perception device 5 and the analysis device 3 are respectively arranged at the front end of the robotic arm 10; the analysis device 3 is a laser induced breakdown spectrometer.

[0174] Specifically, the power supply equipment is located on the upper part of the carrying platform 9, providing power for the entire mineral exploration robot, ensuring the normal operation of all equipment (such as robotic arms, control equipment, analysis equipment, etc.), and monitoring the battery power and battery health status to realize the automatic return function.

[0175] The guide mechanism 2 is installed at the bottom of the carrying platform 9, and includes a plurality of wheels, which are responsible for supporting the carrying platform 9 and providing stable mobility. The guide mechanism 2 cooperates with the control of the mineral exploration robot to realize navigation and path planning, ensuring that the mineral exploration robot can travel safely in complex terrain.

[0176] The analysis device 3 is fixed at the end of the mechanical arm 10, and its main function is to perform real-time analysis on the collected samples. Common analysis devices such as laser induced breakdown spectrometer can quickly identify the elemental composition in the sample and provide support for the acquisition of real-time exploration data.

[0177] The sampling tool 4 is connected to the mechanical arm 10 and is used to collect geological samples. The tool (such as a drill bit, a shovel, a sweeper, etc.) can be switched according to different geological types to ensure efficient and accurate sampling.

[0178] The visual perception device 5 is installed at the front end of the robotic arm 10 and is responsible for acquiring image data of the exploration point, supporting geological type recognition, road condition monitoring and obstacle detection functions, and providing data support for subsequent analysis and path planning.

[0179] The communication device 6 is located at the base of the mechanical arm 10 and is responsible for wireless communication with the host computer. It can be an antenna to achieve real-time uploading of exploration data and receiving of remote control commands to ensure the timeliness of information interaction.

[0180] The positioning device 7 is installed at the base of the robotic arm 10 to obtain real-time position information, support path planning, navigation and target point confirmation, and ensure accurate positioning and efficient exploration of the mineral exploration robot in complex environments.

[0181] The control device 8 is installed below the power supply device 1 and is responsible for controlling the operation of each device, including robot arm operation, path planning, sampling and analysis, etc.

[0182] The carrying platform 9 provides an installation base for all devices. It has a solid structure and a reasonable design, so that the devices can be firmly connected and work together, ensuring the operation stability of the robot in various environments.

[0183] The robotic arm 10 is fixed to one side of the carrying platform 9. The flexible multi-axis design allows the robotic arm 10 to perform sampling and analysis operations in different orientations, and can accurately locate and execute complex sampling and detection tasks.

[0184] The embodiment of the present invention provides a mineral exploration robot, which realizes the automation, efficiency and intelligence of multi-point exploration by integrating visual perception, real-time analysis, path planning and data uploading functions. At the same time, it can autonomously identify geological types, accurately sample and perform real-time element analysis, generate real-time data of exploration points and upload them to the host computer, ensuring the timeliness and integrity of exploration data. At the same time, it improves the efficiency and safety of mineral exploration and provides strong decision-making support for geologists.

[0185] The computer program product provided in the embodiment of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the method described in the previous method embodiment. The specific implementation can be found in the method embodiment and will not be repeated here.

[0186] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system and device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0187] In addition, in the description of the embodiments of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0188] If the functions are implemented in the form of software functional units 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, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for 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 methods described in various embodiments 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 read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0189] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0190] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A mineral exploration method, characterized in that: Applied to a mineral exploration robot, the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool, and is communicatively connected with a host computer of an external device; the method comprises: Acquire an exploration point image of a target exploration point through a visual perception device; Inputting the exploration point image into a pre-trained geological type recognition model, and outputting the actual geological type corresponding to the exploration point image; Based on the actual geological type and the preset correspondence between the geological type and the sampling tool, determining a target sampling tool corresponding to the actual geological type, and collecting samples through the target sampling tool; The collected samples are analyzed in real time by the analysis equipment to obtain real-time exploration data of the target exploration point.

2. The mineral exploration method according to claim 1, characterized in that: The mineral exploration robot is also provided with a positioning device; Before the step of acquiring the exploration point image of the target exploration point by a visual perception device, the method further includes: Obtaining a coordinate file sent by the host computer; the coordinate file includes a plurality of exploration points to be measured, an exploration sequence, and the longitude and latitude coordinates, elevation data, and geological environment notes corresponding to each of the exploration points to be measured; Acquire the current position of the mineral exploration robot collected by the positioning device, and set the current position as the starting point; Acquire the exploration status of each of the to-be-tested exploration points, and determine, based on the exploration sequence, the next to-be-tested exploration point whose exploration status is that the exploration is not completed as the target exploration point; Based on the longitude and latitude coordinates, a first exploration path from the starting point to the target exploration point is planned; and based on the first exploration path, the vehicle travels to the target exploration point.

3. The mineral exploration method according to claim 2, characterized in that: After the step of planning a first exploration path from the starting point to the target exploration point based on the latitude and longitude coordinates, the method further includes: Acquiring a road condition image sent by the visual perception device; Input the road condition image into a pre-trained road condition type recognition model, and output a road condition analysis result corresponding to the road condition image; Identifying the road condition analysis result; If the road condition analysis result shows that there is an obstacle, updating the first exploration path based on a preset obstacle avoidance algorithm; If the road condition analysis result shows that abnormal terrain exists, the running speed is controlled based on preset adjustment rules.

4. The mineral exploration method according to claim 2, characterized in that: After the step of performing real-time analysis on the collected samples by the analysis device to obtain real-time exploration data of the target exploration point, the method further includes: S1: after acquiring the real-time exploration data of the target exploration point, the target exploration point and the real-time exploration data corresponding to the target exploration point are associated and saved, and uploaded to the host computer; S2: changing the exploration status of the target exploration point to completed exploration; S3: Based on the exploration sequence, determine the next exploration point to be measured whose exploration status is unfinished as the next target exploration point; S4: planning a second exploration path from the target exploration point to the next target exploration point based on the latitude and longitude coordinates; and driving to the next target exploration point based on the second exploration path; S5: Repeat steps S1-S4 until the exploration status of each of the to-be-measured exploration points is that the exploration has been completed.

5. The mineral exploration method according to claim 2, characterized in that: The mineral exploration robot is also provided with a power supply device; the method further comprises: Identify real-time battery information collected by the power supply device; the real-time battery information includes battery power and battery health status; If it is identified that the battery power is lower than a preset return threshold, return to the starting point; If it is identified that the battery health status is not within the preset battery health threshold range, return to the starting point.

6. The mineral exploration method according to claim 1, characterized in that: The geological type recognition model is trained by the following method: Acquire a historical geological image data set; the historical geological image data set includes geological images and geological type annotation information corresponding to the geological images; Dividing the historical geological image data set into a training set, a validation set and a test set according to a preset ratio; Training an initial convolutional neural network model based on the training set until a preset training requirement is met to obtain a first geological type recognition model; Based on a preset verification algorithm, the first geological type identification model is optimized to obtain a second geological type identification model; The second geological type identification model is verified based on the verification set until a preset verification requirement is met to obtain a geological type identification model.

7. A mineral exploration system, characterized in that: Applied to a mineral exploration robot, the mineral exploration robot is provided with a visual perception device, an analysis device and a sampling tool, and is communicatively connected with a host computer of an external device; the system includes: An exploration point image acquisition module, used to obtain an exploration point image of a target exploration point through a visual perception device; A geological type determination module, used for inputting the exploration point image into a pre-trained geological type recognition model, and outputting the actual geological type corresponding to the exploration point image; A sampling module, used to determine a target sampling tool corresponding to the actual geological type based on the actual geological type and a preset correspondence between the geological type and the sampling tool, and collect samples through the target sampling tool; The exploration data generation module is used to perform real-time analysis on the collected samples through the analysis equipment to obtain real-time exploration data of the target exploration point.

8. A mineral exploration robot, characterized in that: include: A carrying platform and a control device, a visual perception device, an analysis device, and a sampling tool positioning device arranged on the carrying platform; the visual perception device, the analysis device, the sampling tool, and the positioning device are respectively connected to the control device; and also includes the mineral exploration system according to claim 7, wherein the mineral exploration system is arranged in the control device.

9. The mineral exploration robot according to claim 8, characterized in that: The mineral exploration robot further comprises a mechanical arm and a guide mechanism; the mechanical arm is arranged on one side of the carrying platform; and the guide mechanism is arranged below the carrying platform.

10. The mineral exploration robot according to claim 9, characterized in that: The sampling tool comprises a dozer, a small drill and a sweeper; the sampling tool, the visual perception device and the analysis device are respectively arranged at the front end of the mechanical arm; the analysis device is a laser induced breakdown spectrometer.

Images