130results about How to "Improve exploration efficiency" patented technology

The invention relates to a construction method of a cooperative combat decision-making agent for ground unmanned equipment, and aims to solve the problem of space dimension disaster of a complex cooperative combat task. The DRL solution space is effectively compressed, the exploration efficiency is improved, and meanwhile intelligent decision making of tasks with different granularities in the whole combat process is achieved; aiming at the problem of generating a'task layer 'combat command action strategy, a cooperative combat decision-making agent network model is constructed to realize the unified coordination control from the effective representation of complex state input to the combat action; by constructing a distributed training framework, the sampling and training efficiency is improved, and high-throughput asynchronous parallel training of a cooperative combat decision-making agent network model is realized; by establishing a combat decision agent learning algorithm based on deep reinforcement learning and solving the combat decision agent learning algorithm, the learning efficiency and final performance of an agent network model are effectively improved.

The invention discloses a multi-unmanned aerial vehicle target exploration and tracking method and system based on multi-agent ant colony reinforcement learning. The method comprises the following steps: dividing an actual environment for target tracking of unmanned aerial vehicles into grid units; performing target state initialization based on the grid units to obtain an initial observation position of the unmanned aerial vehicle; based on an ant colony algorithm and the initial observation position of the unmanned aerial vehicle, the transition probability of the unmanned aerial vehicle among the grid units is calculated; establishing a reward function based on the transition probability and the grid unit; based on the reward function and the existing exploration information and tracking information, performing centralized training on the multiple unmanned aerial vehicles by using a multi-agent near-end strategy optimization algorithm to obtain an optimal joint decision; and based on the optimal joint decision, enabling the unmanned aerial vehicle to complete exploration and tracking of multiple targets to be explored. According to the technical scheme, the possibility that the unmanned aerial vehicle finds the target in the unknown area is increased, and the target can be effectively followed.

The invention belongs to the technical field of automatic driving, and discloses an autonomous driving decision near-end strategy optimization method based on distributed exploration enhancement. According to the invention, a distributed architecture of the reinforcement learning decision module of the autonomous vehicle is designed, through parallelization and distributed processing, multiple agents can execute strategy training in different environments at the same time, and strategy updating and empirical data are shared in real time. In addition, the distributed architecture allows the method to operate in a larger-scale data set and a complex environment, and the adaptability and generalization ability of autonomous driving decision making are improved. In addition, a random network exploration (RND) module is also combined, the diversity and effectiveness of exploration are further enhanced, and a near-end strategy optimization algorithm used by the reinforcement learning decision module is perfected. By combining the distributed architecture and the RND module, the training process is remarkably accelerated, the exploration efficiency is improved, the decision-making quality of the automatic driving system is comprehensively improved, and the reliability and efficiency of automatic driving decision-making are improved.

The invention discloses a three-dimensional environment coverage method for on-line sub-region updating and hierarchical planning, which remarkably improves the exploration efficiency and reduces the calculation consumption through on-line sub-region updating and hierarchical planning. Firstly, a fast environment preprocessing method is used, only updated boundary extraction leading edge points are checked, and therefore the calculation burden is reduced. Secondly, constructing a sparse topological graph, discretizing the space, and storing topological nodes by using a KD tree, so that time-consuming path search on a three-dimensional voxel map is avoided; in addition, hierarchical structure division of sub-regions is provided, the environment is divided into a plurality of adaptive sub-regions, environment information is updated in real time, and subsequent exploration planning is supported. A hierarchical planning strategy is adopted, environment coverage is modeled as a traveling salesman problem, a global coverage path is solved, and the total path length of subarea access is considered. Compared with a greedy strategy, the method provides a long-term perspective and improves the exploration efficiency. Finally, the efficiency and feasibility of the method are proved through simulation and actual measurement verification.

The invention discloses an unmanned aerial vehicle flight path planning method based on an improved Harris eagle optimization algorithm, and the method comprises the steps: setting peak parameter information, constructing a threat equivalent terrain simulation map through a peak function, and carrying out the modeling of a flight environment; establishing a corresponding cost function according to the length cost, the threat cost and the smoothness cost of the flight path, and converting a flight path planning problem into an optimization problem of minimizing the cost function; and solving the optimization problem of the minimum cost function by using an improved Harris eagle optimization algorithm fusing a labor division mechanism and an elimination selection strategy, and carrying out iterative optimization to obtain the optimal flight path of the unmanned aerial vehicle. According to the invention, the feasible track can be explored and developed as much as possible in the flight space, collision and track jitter can be avoided, and smooth execution of the flight task of the unmanned aerial vehicle is facilitated.

The invention provides a sandstone type uranium ore prospecting method based on a tectonic unit boundary, and the method comprises the steps: obtaining the geological data of a sedimentary basin, determining the spatial distribution and the tectonic and stratum characteristics of tectonic units in the sedimentary basin, and determining the types of beneficial metallogenic tectonic units according to the spatial distribution and the tectonic and stratum characteristics of the tectonic units; and determining a first-level exploration target region of the sandstone type uranium mine by combining the gravity data and / or seismic data and / or oil and gas bearing conditions of the favorable metallogenic structure units for different types of favorable metallogenic structure units. According to the method, on the basis of deep analysis of uranium-controlled metallogenic characteristics of the tectonic unit boundary, the upheaval and fracture of the tectonic unit boundary are taken as important marks for prospecting, and geological and geophysical prospecting methods are combined, so that a novel efficient and low-cost prospecting method is formed, an optimal target of a sandstone type uranium ore prospecting area can be accurately locked, and the prospecting efficiency of the sandstone type uranium ore prospecting area is improved. The blindness in the prospecting process can be obviously reduced, and the exploration efficiency is improved.

The invention relates to the technical field of surveying and mapping, and discloses a mine geological drilling three-dimensional visualization method based on deep learning, which comprises the following steps: S1, collecting drilling data, and extracting drilling sampling site data from the drilling data by adopting a three-point positioning method; s2, preprocessing the collected data to form a data set for model training; s3, constructing a three-dimensional deep learning model; s4, training the three-dimensional deep learning model by using the data set to realize optimization operation on the three-dimensional deep learning model; s5, performing data drilling by using the optimized three-dimensional deep learning model to form and generate a three-dimensional geologic model; and S6, inputting the three-dimensional geologic model after drilling into visual equipment, and carrying out display operation on the three-dimensional geologic model. According to the invention, efficient processing and three-dimensional visual display of mine geological drilling data are realized, the efficiency of mine exploration and development is improved, the safety risk is reduced, and powerful support is provided for decision making of mining companies.

The invention discloses a multi-unmanned-vehicle autonomous exploration method and system for a real large-scale complex scene, and the method comprises the steps: carrying out the voxelization of a three-dimensional environment, and carrying out the detection of a leading edge point; constructing an initialized undirected passable topological graph for sharing of multiple vehicles and calculation of path distances from the multiple vehicles to a target point; obtaining the pose of the radar relative to the world coordinate system and the pose of the robot relative to the world coordinate system by using observation data of the radar and the IMU; ground radar data extracted through ground segmentation is combined with obstacle point cloud data to guarantee that the robot operates in a complex environment. A brand-new distributed task allocation method for multiple unmanned vehicles is provided, and the dispersion rate and the exploration efficiency between robots are considered; and obtaining a reachable path from the robot to the target point based on the cost function distribution. The problem that in the field of multi-unmanned-vehicle exploration, multi-unmanned-vehicle exploration based on 3D laser radar is mostly applied to a simulation environment and is difficult to deploy to a mobile robot to explore a real environment is solved.

The invention relates to the technical field of image processing, and discloses an autonomous exploration method, device and equipment of a small unmanned aerial vehicle, and a storage medium. The method comprises the following steps: performing three-dimensional rasterization processing on a closed pipeline environment to obtain a three-dimensional grid map represented by voxels; performing classification marking processing on the three-dimensional grid map to obtain a classification map containing idle voxels, occupied voxels and unknown voxels; performing local exploration planning processing on the classification map to obtain a local unmanned aerial vehicle exploration path; performing global exploration planning processing on the classification map to obtain a global unmanned aerial vehicle exploration path; and performing real-time execution and dynamic adjustment processing on the local unmanned aerial vehicle exploration path and the global unmanned aerial vehicle exploration path to obtain closed pipeline exploration data. According to the invention, the autonomous exploration efficiency and accuracy of the small unmanned aerial vehicle are improved.

The invention provides a multi-mechanical-arm motion planning reinforcement learning training method based on a structured action space and a mechanical arm motion planning method. The multi-mechanical-arm motion planning reinforcement learning training method based on the structured action space comprises the steps that sample state poses of multiple mechanical arms and task target coordinates corresponding to sample tasks needing to be executed for the multiple mechanical arms are obtained; inputting the state pose and the task target coordinate into a high-level learning module of a preset hierarchical reinforcement model to obtain a discrete joint action value; inputting the state pose, the task target coordinate and the discrete joint action value into a low-layer learning module of a preset hierarchical reinforcement model to obtain a continuous joint action value; the poses of the multiple mechanical arms are adjusted based on the discrete joint action values and the continuous joint action values, and the adjusted poses of the multiple mechanical arms are obtained; determining a reward value based on the adjustment poses of the multiple mechanical arms; and adjusting a low-level learning module of a preset hierarchical enhancement model based on the reward value to obtain a trained hierarchical enhancement model.

The invention discloses a multivariable inversion method for measuring the neutron porosity of a buried hill reservoir, and belongs to the technical field of nuclear logging. The multivariable inversion method specifically comprises the following steps: constructing a neutron logging instrument array detector, measuring the buried hill reservoir, calculating a counting ratio R, and defining R1 and R2; constructing a multivariable inversion model, and performing inversion to obtain Por, Por1, Por2, cal and OFFSET; and the relative eccentricity OFFSETRE is defined, iteration inversion is conducted on the OFFSETRE through a dichotomy method, and the well diameter, the relative eccentricity and the porosity are output to serve as inversion results. The method can effectively reduce the negative influence of dual environment effects generated by well diameter fluctuation and instrument eccentric distance change on a logging result, remarkably alleviates the complexity and difficulty of environment correction, greatly improves the accuracy and reliability of neutron porosity measurement of the buried hill reservoir, and has a great significance in improving the oil-gas exploration and development efficiency. And the reservoir evaluation precision is optimized, and the method has extremely important practical significance and application value.

The invention discloses a seabed sediment real-time analysis method based on multi-beam sonar. The method comprises the steps of collecting seabed scattering intensity calibration data and navigation data, constructing a seabed scattering data matrix, extracting a seabed scattering intensity coefficient, evaluating a scattering fitting curve and a model curve, and outputting a sediment analysis result. Compared with a traditional seabed sediment analysis method, the seabed scattering intensity calibration method is established, the fitting problem of a seabed scattering model is solved, the sediment type is obtained in real time, seabed exploration operation and engineering construction can be guided, and seabed exploration efficiency and operation safety are improved.

The invention belongs to but not limited to the technical field of seismic facies intelligent identification, and particularly relates to a three-dimensional seismic facies intelligent identification method and system based on V-Net, and the method comprises the steps: S1, obtaining three-dimensional seismic data and a corresponding seismic facies interpretation result; s2, establishing a learning sample for three-dimensional seismic phase identification; s3, building a seismic facies intelligent identification model based on V-Net; and S4, identifying the three-dimensional seismic facies of unknown data. According to the method, seismic facies recognition is taken as a three-dimensional image segmentation problem, and meanwhile, residual errors are added in modeling, so that the rationality of a model stretching hypothesis space is optimized, and the precision of seismic facies intelligent recognition can be effectively improved. According to the method, the residual layer is added on the basis of three-dimensional U-Net modeling, the structure of the intelligent model is optimized, the reasonability of the model expanding hypothesis space is guaranteed, and the overall feature learning ability of the network is improved.

The invention relates to the technical field of prospecting, in particular to an intelligent prospecting method based on multi-source information fusion and physical-data dual-drive, and the method comprises the steps: obtaining a target multi-source information database based on the granularity and reliability of multi-source information data in the target multi-source information database; performing complementary fusion on the multi-source information data according to the matched fusion level to obtain a multi-source information complementary fusion result; extracting mineral resource distribution information of the target area according to a multi-source information complementary fusion result, and generating an earth crust mineral resource refined distribution diagram of the target area by using the mineral resource distribution information; and based on the target multi-source information database and the earth crust mineral resource refined distribution map, generating a mineral resource information inference result of the target area by using a target identification system. According to the method, the target identification system can be utilized to perform deep mining on the earth crust mineral resource refined distribution map drawn based on the multi-source information, the mineral resource information inference result is obtained, the earth crust mineral resource exploration efficiency can be effectively improved, the resource exploration cost can be reduced, and strategic mineral resources can be accurately locked.

The invention relates to a hierarchical target searching and hunting method for an unmanned vehicle cluster, and the method comprises the steps: dividing the unmanned vehicle cluster into a plurality of groups, and each group comprises a guide vehicle and two member vehicles; in the to-be-searched area, each group of guide vehicles performs target search based on the positions of the guide vehicles; when a target is not found, each group of guide vehicles respectively carry out incremental environment node exploration to continuously carry out target search: storing respective current positions into explored stacks and moving to latest node positions in corresponding to-be-explored stacks to carry out target search until the searched target is found or the nodes in the respective corresponding to-be-explored stacks are empty; wherein the to-be-explored stack is a storage space of a plurality of passable node positions obtained by carrying out radiation type node exploration on the target search position by the guide vehicle; the explored stack is used for storing node positions which the small group guide vehicles have reached; when a target is found, the guide vehicles of the group predict the motion trail of the target; and the three vehicles in the current group carry out triangular surrounding on the target.

The invention relates to a reservoir physical property parameter prediction and evaluation system for oil-gas exploration, and aims to deal with exploration challenges under complex geological conditions. The system comprises a multi-source data acquisition module, a self-adaptive data processing module, a self-learning prediction module, a dynamic model correction module and a visual result display module. By integrating logging, core analysis, earthquake and historical exploration data, the system performs high-precision prediction on the porosity and permeability of the reservoir, and updates the model in real time through a dynamic correction technology to ensure the accuracy and reliability of a prediction result. In addition, the system also provides two-dimensional and three-dimensional visual display and uncertainty analysis functions, and assists a decision maker in identifying a high-potential reservoir area and making an optimal development scheme. The system has the interface connection capability with an external data management system, can realize automatic updating and economic analysis of data, and effectively improves the efficiency and success rate of oil-gas exploration.

The invention provides a robot control method, device and equipment based on multi-agent reinforcement learning. The method comprises the following steps: adaptively resetting strategy parameters of an initial neural network of each robot in a robot group cooperation system according to a preset rule; inputting the sample observation data of each robot at the current moment and the sample action of each robot at the previous moment into an initial neural network of each robot to obtain a local prediction return value of each robot; calculating a global prediction return value according to the local prediction return value of each robot, and fusing the parameters of the initial neural network of each robot; updating the experience pool of each robot, collecting sample data from the experience pool of each robot, and periodically repeating the training process of each robot until a preset convergence condition is reached; and controlling each robot to complete a cooperation task based on the trained neural network of each robot. The diversity and breadth of exploration behaviors can be improved, and the exploration efficiency is improved.

The invention relates to the technical field of reinforcement learning, in particular to a self-learning method and device for a non-antagonistic task, equipment and a medium, and the method comprises the steps: 1, initializing a decision neural network of the non-antagonistic task; step 2, executing an MCTS state search algorithm based on the current decision neural network, obtaining a third-party efficiency score and recording path information; step 3, performing training optimization on a decision neural network according to the third-party efficiency score and the path information; and step 4, repeatedly executing the steps 2-3 in parallel until the decision neural network meets a trigger termination condition. The self-learning method for the non-antagonistic tasks provided by the invention is not limited to task scenes, can obtain better results for all the non-antagonistic tasks, and can perform adaptive modification according to the task scenes; the shortest solution can be quickly and gradually explored from 0 for different non-antagonistic tasks, repeated calculation is avoided to the greatest extent, and the exploration efficiency is improved.

The invention discloses a comprehensive method for analyzing compressibility of a shale reservoir based on machine learning, and relates to the technical field of oil and gas field development, the comprehensive method for analyzing compressibility of the shale reservoir based on machine learning mainly comprises the following steps: obtaining a training data set according to core experiment data and corresponding logging data; constructing a three-parameter prediction model by using a random forest regression model; training the three-parameter prediction model by using the training data set to obtain a trained three-parameter prediction model, and predicting the to-be-predicted logging data to obtain three parameters; and calculating a compressibility coefficient based on the three parameters so as to obtain compressibility classification evaluation of the shale reservoir in the research area. By implementing the comprehensive method for analyzing the compressibility of the shale reservoir based on machine learning provided by the invention, the compressibility of the shale reservoir can be efficiently, accurately and universally evaluated.

The present invention belongs to the field of autonomous driving technology, and discloses a proximal policy optimization method for autonomous driving decision-making based on distributed exploration enhancement. The present invention designs a distributed architecture of the reinforcement learning decision-making module of the autonomous driving vehicle. Through parallelization and distributed processing, multiple agents can simultaneously perform policy training in different environments, and share policy updates and experience data in real time. In addition, the distributed architecture allows the method to operate on larger data sets and complex environments, improving the adaptability and generalization ability of autonomous driving decisions. The present invention also combines a random network exploration RND module to further enhance the diversity and effectiveness of exploration, and improves the proximal policy optimization algorithm used by the reinforcement learning decision-making module. By combining the distributed architecture with the RND module, the present invention significantly accelerates the training process, improves the exploration efficiency, and comprehensively improves the decision-making quality of the autonomous driving system, and improves the reliability and efficiency of autonomous driving decisions.

The invention discloses a layered planning method for indoor autonomous exploration of a mobile robot based on dynamic prediction, and the method comprises the steps: S1, dividing an exploration environment where the robot is located into a plurality of local subspaces, and marking the state of each subspace as an unexplored state, an explored state, an explored state or a semi-explored state according to the data of a sensor and a semantic detection result; s2, establishing a global topological graph connected with each subspace, and planning a global access path by using a traveling salesman problem model; s3, integrating a pedestrian trajectory prediction model in the selected current subspace, and evaluating and selecting a local viewpoint capable of maximizing information gain and minimizing dynamic collision risk based on a prediction result; the method has the beneficial effects that a semi-exploration state and a self-adaptive revisit mechanism are introduced, and when the door is possibly opened, the check is turned back, so that the dead angle of a map is eliminated, and the exploration completeness is improved; pedestrian trajectory prediction is incorporated into viewpoint scoring, and people deadlock is avoided.

The present invention discloses a signal-level radar interference generation method and device in an open environment, the method comprising: obtaining radar interference data information; the radar interference data information comprises radar radiation source description word information, interference pattern action space, interference pattern and interference pattern parameter space; the radar radiation source description word information comprises carrier frequency information, pulse interval information, pulse width information and radar state judgment information; processing the radar interference data information to obtain state information S, interference pattern action space A, interference pattern parameter space B; setting the interference effectiveness reward value and the termination state, using a preset radar interference generation model, processing the radar interference data information to be processed, and obtaining radar interference signal information. The method of the present invention uses a large amount of real-time online radar confrontation training, and the training algorithm is from the radar signal to the underlying logic of the output interference signal, so as to effectively deal with the radar target confrontation problem in an unknown and small sample environment.

The present invention discloses a method and device for constructing a seismic velocity body. The method comprises: selecting the time-depth relationship of the drilled wells in the target area that meets the requirements based on the synthetic seismic records and actual seismic records of the drilled wells in the target area; determining the velocity data of the drilled wells according to the time-depth relationship of the drilled wells; extracting the preferred seismic attributes related to the target layer based on the velocity data, the seismic interpretation layer of the target layer and the seismic data body; fitting the layer velocity body of the target layer based on the time-depth relationship of the drilled wells, the seismic interpretation layer and the seismic attributes related to the target layer and the pre-established time structure model of the target layer; establishing the average velocity body from the surface to the target layer based on the layer velocity body of the target layer to obtain the seismic velocity body of the target area. It can effectively guide the preferred exploration target of the target area, improve the success rate of the deployment of exploration wells, optimize the drilling footage, save drilling costs, and improve the exploration efficiency of the early exploration blocks.

The present invention discloses a drilling device for mineral resource exploration, and relates to the technical field of drilling devices. Its technical key points include a bearing frame and a movable plate which can be lifted and lowered and is installed on the bearing frame, an engine is fixedly installed on the movable plate, a notch is opened on the bearing frame, wherein a gear is fixedly installed on the output end of the engine; and further includes: a gear ring, which is rotatably installed on the movable plate, an annular column is rotatably installed on the gear ring, a second fixed plate is fixedly installed on the inner surface of the annular column, a pair of limit plates are fixedly installed on the inner surface of the gear ring, the second fixed plate is located between the two limit plates, an elastic member is fixedly arranged between the second fixed plate and the gear ring, a plurality of clamping blocks are slidably installed on the annular column, a tension spring is fixedly installed between the clamping block and the annular column, a piston ring is slidably installed inside the gear ring, and a drill rod can be automatically spliced ​​and drilled into the ground, thereby reducing the difficulty of exploration for staff.

The invention discloses a mobile robot autonomous exploration method based on a topological graph, and relates to the field of intelligent robots. According to the method, a framework topological structure of an environment is introduced, exploration behaviors are guided by utilizing topology of the environment, and the method comprises incremental framework extraction, framework topological graph updating and exploration strategy finite-state machine. According to the method, the BFMs phenomenon is relieved to a great extent, complex TSP problem solving is avoided, and efficient exploration incremental skeleton extraction is achieved with extremely low calculation consumption; the exploration behavior of the robot is effectively guided, invalid movement is remarkably reduced, information acquisition is maximized, and resource consumption is minimized.

The invention relates to the field of machine control, and provides a terrain exploration method and device for a ground robot, equipment and a storage medium. The method comprises the following steps: carrying out trafficability state marking on each region of an exploration state map, wherein trafficability states comprise an explored impassable state, an explored trafficable state and an unexplored state; in the exploration state map, determining a maximum square which contains the ground robot and only contains the explored area; safety degree evaluation is carried out on each unknown area; based on a preset exploration target and the safety degree of each unknown area, determining an optimal exploration route of the ground robot; and planning a safe exploration path of the ground robot according to the optimal exploration route, and controlling the ground robot to perform terrain exploration on an area with an unexplored state based on the safe exploration path. The method provided by the invention can improve the exploration efficiency of the ground robot.

This invention discloses a well placement method, equipment, and storage medium for bioherm-type carbonate reservoirs. The well placement method includes the following steps: S1. Obtaining the amplitude spectrum data of the reservoir to be well placed; S2. Classifying the bioherm type and reservoir category, and determining the connectivity of the reservoir; S3. Determining the optimal well placement area and optimal well placement pattern based on the determined bioherm type, reservoir category, and reservoir connectivity; S4. Performing economic calculations on the wells to be placed within the optimal well placement area, establishing an economic model, and determining the optimal well spacing d. i And the optimal number of oil and gas wells N i This method, tailored to the characteristics of bioherm-type carbonate reservoirs, classifies them into different types, distribution ranges, and connectivity, providing a reliable reference for well placement. It designs optimal well placement areas and patterns, effectively improving well development efficiency. An economic model is established through economic accounting to optimize well spacing and the number of oil and gas wells, thereby increasing reservoir development utilization and economic benefits.

The invention provides an exploration method of a robot in an unknown space and the robot, and the method comprises the steps: collecting environment data in the unknown space through a sensor, so as to generate a grid map; inputting the grid map into a preset prediction model, and outputting a target point needing to be explored; generating a path to the target point, and going to the target point along the path; in the process of going to the target point, a sensor is used for collecting surrounding environment data in real time, and the grid map is updated in real time based on the collected environment data; and judging whether exploration of the unknown space is completed or not according to a preset rule, if not, inputting the updated raster map into the prediction model again, outputting a new target point needing to be explored, collecting environment data in the process of going to the new target point to update the raster map again, and repeating the steps until exploration of the unknown space is completed. Compared with the prior art, the method has the advantages that the clear exploration target point is output based on the prediction model, so that the exploration of the robot is clear and high in efficiency.

The invention discloses a mineral resource exploration optimization method based on gravity forward and reverse modeling, and relates to the technical field of Internet of Things. Multi-source conversion is carried out on target gravity data, a target gravity factor is obtained, a target optimization coefficient is set for parameter variation and series division, and a target coefficient slice is obtained; performing detail optimization on the target gravity factor through the target coefficient slice to obtain an optimized gravity coefficient; generating an optimized coefficient spectrogram according to the optimized gravity coefficient, and adjusting and judging the gravity coefficient curve according to a dynamic judgment axis to obtain a section range coefficient; according to the section range coefficients, performing range distinguishing and peer linking on the virtual space coordinates to obtain peer coefficient link circles, and according to the peer coefficient link circles, performing anomaly survey on the virtual target area to obtain an abnormal gravity area; the accuracy of the gravity inversion result is greatly improved, the exploration efficiency is improved, resources and time required by exploration are reduced, and the exploration cost is reduced.

The present invention relates to the technical field of intelligent robots, and in particular to a system and method for autonomous source search and exploration of robots in an unknown environment. An environment perception module and an environment explorer in a control module jointly explore the environment; a source search judge and a source direction guided exploration decision in the source search perception module and the control module complete the source search decision and guide the work of the environment exploration system; a motion system follows the instructions sent by the environment explorer to a path planner to go to a target exploration point; a communication component is used to receive instructions and report source search results; for autonomous source search of robots in an unknown environment, the present invention adopts source prediction to guide environmental exploration and simultaneously carry out source information judgment, adopts global and local multi-level boundary exploration methods for unknown environments, adopts boundary maximum source search value judgment for local non-source extreme point dilemmas, selects the optimal target exploration point, jumps out of the current local interval, and carries out source search and exploration in a new area until the real source point is found.