41results about How to "Avoid falling into" patented technology

This invention relates to the field of graphite waste treatment technology, and more particularly to a method for preparing special graphite materials from isostatic pressing graphite waste, comprising the following steps: waste collection: firstly, the waste generated from isostatic pressing graphite production is collected and processed centrally; waste treatment: the collected waste is centrally fed into a purification module, where impurities are screened and removed; waste crushing: after the isostatic pressing graphite waste has been screened, it is pushed into a crushing device for crushing; material proportioning, pressing, primary calcination, and graphitization treatment; this invention, by processing the waste from isostatic pressing graphite production through the screening and purification processes described above, followed by crushing, can then be proportioned, fed into a mixing pot, and then pressed, calcined, and graphitized to produce special graphite materials, thus avoiding environmental pollution from waste and achieving full utilization of resources.

This utility model discloses a multi-functional fire-fighting intercom, including a main body and an intercom module, a temperature detection module, a control module, a gas detection module, a stationary detection module, an alarm module, and a battery integrated on the main body. The intercom module has a microphone speaker on its left or right side. The temperature display panel of the temperature detection module is located on the top of the main body. The gas detection module has a gas detection probe and an information display window on its front, displaying the concentration of the detected gas. The stationary detection module includes an accelerometer sensor to sense the speed of the main body. The alarm module includes an alarm speaker and several alarm indicator lights. The control module is electrically connected to the gas detection module, the stationary detection module, and the alarm module. The battery is electrically connected to each module. This significantly expands the uses of the intercom and effectively improves the safety of firefighters during fire rescue operations.

The utility model discloses a safety protection device for chemical engineering, which comprises a feed hopper and a safety filter screen, the top end of the feed hopper is detachably provided with a protective cover, the top end of the protective cover is communicated with a feed pipe, the top end of the feed pipe is in threaded connection with a sealing cover, and a fixed cylinder is inserted into the feed hopper. The safety filter screen is installed on the bottom side of the interior of the fixing cylinder, a top ring is arranged at the position, below the fixing cylinder, in the feeding hopper, the bottom end of the top ring is fixedly connected with a plurality of connecting rods, the connecting rods are all in sealed sliding fit with the feeding hopper, and the bottom ends of the connecting rods are fixedly connected with a lifting ring; and the fixed cylinder and the safety filter screen can be conveniently ejected out from the interior of the feeding hopper, and people can conveniently take out the safety filter screen from the interior of the feeding hopper.

This invention discloses a container ship trajectory prediction method based on EMD-PSO-GRU-RBFNN, belonging to the field of intelligent shipping technology. The method takes ship AIS data as input, including time, longitude, latitude, speed, and heading angle information; preprocesses the AIS data to divide it into training and test sets; constructs a trajectory prediction model, calculates the difference between the predicted values ​​of the training set and the actual ship position coordinates to obtain the residual signal; adaptively decomposes the residual signal using empirical mode decomposition to obtain intrinsic mode components and residual components; constructs prediction models for high-frequency and mid-to-low-frequency components, reconstructs the prediction results of each component to obtain the residual prediction value, and superimposes it with the baseline prediction results of the test set to obtain the corrected test set output; the corrected test set output is then evaluated. This invention can model and compensate for trajectory errors at different frequency scales, improving the stability and reliability of trajectory prediction.

The application discloses a transformer integrated cooler group and a transfer device thereof, and relates to the technical field of transformer manufacturing equipment.The transformer integrated cooler group comprises cooler group bodies, a frame and mounting supports, the cooler group bodies are provided in plurality, the plurality of cooler group bodies are fixedly arranged on the mounting supports, the mounting supports are fixedly arranged on the frame, and the cooler group bodies are located on the side of the frame which is far away from the transformer body when the frame is located at the mounting position.The plurality of cooler group bodies are pre-installed on the mounting supports of the frame, and the frame is integrally carried to be further installed on the transformer body, so that the installation site of the transformer is saved, the installation process of the transformer cooler group is simplified, the input of the workers is reduced, and the installation process performance is improved.

The application provides a power grid fault traveling wave positioning method based on a fusion optimization algorithm, comprising: establishing a dynamic virtual fault framework model of a power grid, randomly generating a plurality of virtual fault particles containing position coordinates, and constructing an initial virtual fault particle cluster; using a shortest path algorithm to calculate the topological path length of each virtual fault particle in the virtual fault particle cluster to all measurement points in the power grid, and then calculating a virtual fault space-time feature vector; setting the traveling wave arrival time when a real fault occurs as a target clustering centroid in a feature space, and calculating a feature matching fitness; constructing a non-convex optimization problem with the optimal feature matching fitness as a target, solving the problem using a fusion optimization algorithm, and thus obtaining a fault positioning result and a traveling wave velocity, so as to realize traveling wave positioning. The application can solve the problems of local optimum, slow convergence speed and high calculation cost in the prior art.

The application discloses a point pair registration method and system based on feature guidance and double constraints, and belongs to the technical field of computer vision and robot grabbing. The method aims to solve the problems of recognition failure and insufficient estimation accuracy of traditional methods when parts are blocked and stacked, and point cloud features are not obvious. The scheme adopts a two-stage optimization strategy. Firstly, in the coarse registration stage, an adaptive sampling PPF algorithm guided by geometric features is proposed. By preferentially sampling points in the edge, curvature and other feature obvious areas, the robustness and efficiency of matching are improved. Secondly, in the fine registration stage, an ICP algorithm with multi-constraint optimization is proposed. The pose result obtained by coarse registration is used as the initial value. Through rotation and angle constraint, a multi-objective optimization function is constructed to guide the fine registration process to converge in the correct direction and avoid falling into local optimum, thereby ensuring the accuracy and physical feasibility of the final pose estimation. The application significantly improves the accuracy of industrial part pose recognition.

The embodiment of the application discloses a cluster artificial potential field improvement method and device based on a region threat degree; the method comprises the following steps: acquiring a static obstacle position; establishing a multi-agent cluster structure, and adopting a second-order consistency algorithm to form following; when a moving queue is a desired queue, obstacle avoidance control is performed through an artificial potential field method to obtain an overall following path of the cluster; according to the obstacle avoidance path, the second-order consistency algorithm is adopted to perform path following to control the cluster to avoid obstacles and reach a target point; by introducing second obstacle position information and threat weight information, the size and direction of repulsion force suffered by the cluster are adjusted; by introducing position information and threat weight information of obstacles in a farther range, the size and direction of repulsion force suffered by the cluster are adjusted, so that the moving direction of the cluster is changed, and the cluster is prevented from falling into a local minimum value trap region.

This invention relates to the field of electronic information technology and discloses a zero-carbon airport BIPV photovoltaic direct-drive low-voltage lighting system. It addresses the problem in traditional methods where the MPPT algorithm easily gets trapped in local optima under dynamic shading and irradiance variation conditions, failing to achieve efficient global power point tracking. The method includes: a data acquisition module that collects photovoltaic array voltage, current, and power parameters; an environmental monitoring module that detects environmental changes and generates variation signals using irradiance sensors and cameras; an algorithm initialization module that sets the search range and divides the power curve region based on data and signals; a global search module that scans multiple peaks and traverses the global space to identify the maximum power point; a local optimization module that finely tracks and corrects the step size for peak regions; and a control output module that generates commands to drive the inverter to adjust its operating point and provides status feedback. Through module collaboration, efficient operation of the photovoltaic direct-drive low-voltage lighting is achieved, making it suitable for zero-carbon airport environments.

The utility model discloses an OCA optical cement defoaming device which comprises a vacuum box, a cover plate is hinged to the top of the vacuum box, a connecting pipe is arranged on the left side of the vacuum box, a vacuum pump is fixedly connected to the end of the connecting pipe, a connecting plate is fixedly connected to the upper portion of the side face of the vacuum box, and a dust removal mechanism is arranged on the surface of the connecting plate. The dust removal mechanism comprises a supporting plate, a first motor is fixedly installed on the outer side wall of the supporting plate, and a threaded lead screw is arranged on the inner side wall of the supporting plate. By means of the dust removal mechanism, solid dust particles left on the surface of an OCA raw material can be effectively removed before vacuum bubble removal is conducted on the OCA raw material, the raw material is in a clean state to be subjected to vacuum machining treatment, the treatment effect of follow-up vacuum machining is guaranteed, the dust removal mechanism can be pushed to ascend in cooperation with an air cylinder, and the dust removal efficiency is improved. And therefore, OCA adhesive tape raw materials with different thicknesses can be conveniently cleaned, and the adaptability of the dust removal mechanism to the raw materials is improved when the dust removal mechanism is used.

The invention discloses underground air dust removal equipment for phosphorite mining, and relates to the technical field of air dust removal. The fixed end of an elastic telescopic column is fixedly mounted at the bottom of a bottom plate, a fixed plate is fixedly mounted on the front face of the fixed end of the elastic telescopic column, and one end of a long-strip-shaped rod is fixedly mounted at the bottom of the fixed plate; a round hole is formed in the front face of the square box, the square box is used for controlling the descending speed of a bottom plate at the top of the elastic telescopic column, the situation that the height of the spraying machine deviates from the preset requirement due to the fact that the elastic telescopic column descends too fast is avoided, and it is guaranteed that the spraying machine is always at the optimal dust removal height; the moving range of the spraying machine along with the carrier plate is expanded, multi-direction and wider-area spraying dust falling can be achieved, the problem that dust removal is not thorough in a single fixed range when the underground space is complex is solved, and the comprehensiveness and flexibility of dust removal are improved.

The application belongs to the technical field of power equipment fault diagnosis, and specifically discloses a power transmission line fault identification method and a direct current test instrument. The power transmission line fault identification method comprises the following steps: S1, obtaining fault signals and environmental temperature and humidity data of a to-be-tested line and performing pretreatment to obtain fault characteristic data; S2, inputting the fault characteristic data as test samples into a WOA-KFCM fault detection model, wherein the WOA-KFCM fault detection model comprises a WOA algorithm unit and a KFCM algorithm unit; and S3, calculating the initial clustering center of the KFCM algorithm unit through the WOA algorithm unit, and outputting the line fault category through the KFCM algorithm unit according to the initial clustering center and the test samples. Through the direct current test instrument and the above-mentioned fault identification program, the content of the power transmission line fault can be quickly identified, and the troubleshooting efficiency and the identification accuracy of the complex fault type of the power transmission line are effectively improved.

This invention relates to the field of neural network architecture search and industrial intelligent processing technology, and discloses a structure search method for multi-output dendritic neuron models for industrial classification tasks. The method includes: initializing the synaptic connection weights and dendritic threshold parameters of the multi-input multi-output dendritic neuron model to generate an initial population; uniformly dividing the initial population into several subpopulations of equal size; calculating the temporal and spatial criteria for each subpopulation after task allocation and optimization, and dynamically allocating evolutionary computational resources for each subpopulation based on the fitness change rate represented by the temporal criterion and the population distribution state represented by the spatial criterion; implementing an accelerated sharing penalty mechanism to adjust fitness values ​​according to the crowding degree among individuals to maintain solution set diversity; and updating each subpopulation. This invention solves the problem of low convergence efficiency caused by uniform resource allocation in traditional large-scale multi-objective evolutionary algorithms during neural network architecture search.

The application provides a runoff sediment content prediction method, system, device and storage medium, belongs to the river sediment prediction technical field, and includes the following steps: introducing the position updating mechanism of the spotted hyena optimization algorithm and the position updating mechanism of the gradient optimization algorithm in the position updating mechanism of the gorilla troop optimization algorithm to obtain an improved gorilla troop optimization algorithm; the model parameters of a BiLSTM model are optimized by using the improved gorilla troop optimization algorithm, and a runoff sediment prediction model is obtained by training the BiLSTM model after parameter optimization. The runoff sediment content prediction result and the prediction accuracy are obtained based on the runoff sediment prediction model. Strong support is provided for ecological protection and engineering planning.

The invention provides a blasting vibration peak velocity prediction method fusing parameter uncertainty and data-driven optimization, which comprises the following steps of: firstly, acquiring data such as blasting parameters, lithologic indexes, geological conditions and actually measured vibration peak velocity (PPV), and establishing a basic database; a prior probability model is constructed, a joint uncertainty model is established in combination with probability disturbance and a fuzzy triangular number, a multi-dimensional disturbance sample is generated through a joint central value and a joint standard deviation and is fused with original data, and an extended database is formed. Feature analysis is carried out on the fused data, and input variables which have obvious influence on the vibration peak velocity (PPV) are screened; and constructing a BP neural network model based on the screened features, carrying out global optimization on a network weight and a threshold by adopting a PSO algorithm, and then carrying out local fine tuning by utilizing Adam. Finally, through training and verification, prediction of the vibration peak velocity (PPV) is realized, and model precision is evaluated through RMSE, MAE, MAPE, Rand other indexes. The influence of rock and soil parameter uncertainty on prediction precision can be effectively processed, and the reliability and applicability of blasting vibration prediction are improved.

This invention discloses a method for constructing a road cavity detection model based on 3D ground-penetrating radar (GPR). The method includes constructing a hybrid spatiotemporal detection model, which comprises a feature preprocessing module, a soft clipping module, a spatiotemporal feature extraction module, and an output module. The feature preprocessing module performs feature pre-extraction on preprocessed and labeled data, outputting a pre-extracted feature map. The soft clipping module performs spatiotemporal perception on the pre-extracted feature map, outputting a clipped feature map. The spatiotemporal feature extraction module analyzes the clipped feature map, outputting feature information with global spatiotemporal information. The output module integrates the feature information and predicts the existence of cavities and the range of their occurrence channels. By combining the correlations between data in the spatiotemporal dimension, it accurately detects underground cavities, solving the problems of high computational load and difficulty in adjusting model parameters in existing 3D GPR high-resolution data interpretation. This provides an efficient and reliable technical solution for intelligent road cavity detection and infrastructure safety maintenance.

The utility model relates to the technical field of greenhouse gas collection, and discloses a rice field greenhouse gas collecting device which comprises a mounting plate and four inserting holes formed in the periphery of the top end of the mounting plate, an inserting pipe is fixedly communicated with the middle of the bottom end of the mounting plate, and a water collecting assembly used for collecting water drops is arranged in the middle of the top end of the mounting plate. The top end of the water collecting assembly is fixedly connected with a collecting barrel used for collecting gas, the middle of the top end of the collecting barrel is fixedly connected with a sealing connecting block, the water collecting assembly comprises a gas collecting pipe fixedly connected to the top end of the mounting plate, and the top end of the gas collecting pipe is fixedly connected with a water collecting ring; water drops are formed on the inner wall of the water collection ring in the gas collection process, flow into the water collection tank through the inclined inner wall to be collected and are intensively discharged through the water drainage pipe, the situation that the water drops flow into soil to affect the growth environment of the rice field is avoided, and the balance degree and stability of the device are guaranteed through the mounting plate.

The utility model belongs to the field of pesticide spraying device, concretely is a kind of lotus root planting organic fertilizer pesticide spraying device, the bottom of spraying device is provided with locating plate, the upper portion of locating plate is inserted with antiskid rod, the bottom fixedly connected with locating plate has connecting column, the bottom fixedly connected with connecting column has bearing plate, when the spraying device of connecting column upper portion is moved by bearing plate, the pesticide spraying pipe of spraying device side portion can carry out pesticide application to lotus root pond, and the bottom of bearing plate is in contact with lotus root pond silt ground area, avoid spraying device in the process of moving to sink into silt ground, to reduce work intensity, while the end of bearing plate is provided with pry plate, so that in the process of bearing plate movement, when the protruding soil layer is encountered in moving path, it can also pass smoothly, the stop block can prevent steering rod from continuing to rotate when steering rod is in vertical state, so that the steering rod can be kept in place, and multiple antiskid rods can form a fence to avoid the spraying device from falling during movement.

This invention discloses an improved near-infrared spectral wavelength selection and modeling method for artificial bee colony algorithms, comprising the following steps: data preparation, chaotic initialization, fitness evaluation, establishing a PLS model based on wavelength subsets, using the regularized fitness function of RMSECV as the fitness Fitnessxi of nectar source xi; searching for foraging bees; searching for observation bees; searching for scout bees; outputting the optimal wavelength subset Sbest and establishing the final PLS model. The beneficial effects of this invention are: the method effectively improves convergence speed and accuracy; chaotic initialization improves the quality of the initial population, avoiding the algorithm from prematurely falling into local extrema; the adaptive perturbation factor balances the algorithm's global search and local exploitation, improving the ability to find the optimal wavelength combination in high-dimensional discrete space, and the selected wavelength subset has stronger representativeness, resulting in a final PLS model with lower RMSEP (root mean square error of prediction) and better robustness.

The utility model provides an elephant trunk structure of a bionic elephant robot. The elephant trunk structure comprises a body, the front end of the body is provided with an elephant trunk mechanism, the elephant trunk mechanism comprises a flexible shell and a plurality of soft body units, each soft body unit is provided with an inertia measurement module, each soft body unit is provided with four SMA spring drivers, and each SMA spring driver is provided with an SMA spring. When the service object exceeds the monitored electronic fence area, the whole mechanism moves to the vicinity of the service object and drives the trunk mechanism, so that the trunk mechanism straightens and blocks the service object to prevent the service object from crossing the electronic fence area, thereby preventing the service object from being caught in a dangerous environment, and reducing the probability of safety accidents. And the whole robot adapts to the requirements of different service objects, the combination of science and technology and humanistic care is promoted, and the popularization value is high.

The utility model discloses a plate radiator. A first sensor and a fan are limited by the bottom and the top of the radiator body through first connecting pieces respectively, a power box is arranged on the side face, a baffle is connected to the outer side, the baffle comprises a top plate and side plates on the two sides, and second connecting pieces are hung on the side plates and hook a water pipe at the bottom of the radiator body. According to the plate type radiator provided by the scheme, the simplification of a mounting mode and the convenience of integration can be realized.

The present application relates to the technical field of Internet of Things security, and more particularly to an intelligent testing method and system for Internet of Things protocol vulnerabilities based on machine learning, which generates an initial seed corpus of state transition path coverage by constructing a protocol state machine and a semantic constraint model, generates a mutated test message by weighting and mutating the seed message according to state coverage feedback, forms an abnormal message feature set by executing fuzzy testing and extracts a vulnerability feature fingerprint library, dynamically adjusts the network access control strategy based on the vulnerability feature fingerprint library to realize real-time interception rule updating and priority sorting, thereby improving the fuzzy testing efficiency and vulnerability discovery capability and realizing continuous dynamic defense for Internet of Things devices.

The application discloses an electric scooter, a control method thereof, a storage medium and a domain controller, relates to the technical field of electric scooter steering control, and the method comprises the steps of: obtaining a rocker signal of the electric scooter; normalizing the rocker signal into an X-axis component and a Y-axis component; according to the Y-axis component, mapping the rocker signal into a first target steering angle of a left front wheel, a second target steering angle of a right front wheel, a first target driving speed of a left rear wheel and a second target driving speed of a right rear wheel by using an Ackerman steering model or a model of turning around in place; controlling steering motors of the left front wheel and the right front wheel according to the first target steering angle and the second target steering angle respectively, and controlling driving motors of the left rear wheel and the right rear wheel according to the first target driving speed and the second target driving speed respectively. The application sets a steering motor on the universal wheel, and actively controls the steering angle of the universal wheel through the steering motor, so that the universal wheel is prevented from being trapped in a stuck angle, and the steering smoothness is improved.

The utility model discloses a poultry egg husking device, which belongs to the technical field of poultry egg processing equipment, and comprises a pressure conveying roller and two husking rollers which are respectively and rotatably arranged on a rack, each husking roller comprises a core shaft which is rotatably arranged on the rack, each core shaft comprises a transmission end and a mounting end, a transmission gear is fixedly arranged at each transmission end, and a transmission shaft is fixedly arranged at each mounting end. The transmission gear is sleeved on the mandrel, the mounting end is sleeved with a rubber sleeve, the rubber sleeve and the pressing conveying roller jointly form an effective poultry egg shelling area, and the shaft diameter of non-shelling effective areas on two sides of the rubber sleeve is larger than the diameter of the mandrel at the position of the rubber sleeve. The poultry egg shelling roller solves the technical problems that a shelling roller is poor in bending resistance, and poultry eggs are damaged due to the fact that the poultry egg shelling roller is prone to deformation during shelling, and is widely applied to poultry egg shelling.

The utility model discloses a convenient to install's high -efficient energy -conserving oxygen -increasing machine structure, including floating device, driving device and installation component, driving device adopts frequency conversion drive motor, the driving motor's driving shaft bottom end is installed with helical impeller, floating device adopts three arc float, installation component includes circular protective cover, conical fairing, conical base, triangular support frame, inner support sleeve, the cavity between circular protective cover, inner support sleeve between arc float installation saves space, stable support is not easy to roll over, and arc float automatically adapts to water level, and oxygen -increasing machine installation is convenient, and personnel can complete installation just by placing oxygen -increasing machine in the corresponding position of aquaculture water area and connecting power supply, greatly reduce the labor intensity of user, through the conical base installation in inner support sleeve bottom end center, effectively avoid the water level of aquaculture water area too low and cause helical impeller to sink into silt and sand, and the structural design is reasonable, and the practicability of oxygen -increasing machine has been increased.

This application provides a parameter tuning method, apparatus, device, and computer storage medium for multiple objectives. The method includes: collecting historical behavior information of multiple users regarding recommended materials; determining the revenue parameter vector corresponding to each of the N sets of fusion parameter vectors in the fusion network based on the historical behavior information; determining the correlation coefficient between each set of fusion parameter vectors and the revenue parameter vector corresponding to each set of fusion parameter vectors; determining the increment vector between the first revenue parameter vector and each second revenue parameter vector in the N sets of revenue parameter vectors corresponding to the N sets of fusion parameter vectors; updating the mean vector of the fusion parameter vector corresponding to each second revenue parameter vector based on each set of increment vectors; and updating the N sets of fusion parameter vectors based on the updated mean vector and the correlation coefficient. This application improves the efficiency and speed of fusion parameter vector optimization, achieving more accurate material recommendations.

The invention relates to the technical field of mining dump trucks, in particular to an off-highway mining dump truck and an unloading method. The vehicle comprises a vehicle frame, a vehicle hopper and a transfer device, wherein the transfer device is located below the vehicle hopper; the car hopper comprises a car hopper body and a discharging door assembly, the car hopper body is hinged to the car frame in a turnover mode, and the discharging door assembly is rotatably installed at the rear end of the car hopper body so as to open or close a discharging opening; the transfer device comprises a conveying mechanism which can horizontally move in the front-back direction and can be horizontally and rotatably connected to the frame. The transfer device is arranged, and the conveying mechanism in the transfer device can horizontally move in the front-back direction and rotate in the horizontal direction, so that materials in the car hopper body are conveyed into the tunnel through conveying of the conveying mechanism, the materials can be accurately poured into the target tunnel without manual work, manual secondary carrying is omitted, and the working efficiency is improved. The labor intensity of workers is greatly reduced, and the discharging accuracy and the discharging efficiency are improved.

This invention discloses a reinforcement learning-based satellite pursuit and escape method and system based on orbital maneuver prediction. The method constructs a SAC policy network, designs a suitable loss function for backpropagation, and trains the initial policy of the model. A curiosity mechanism is introduced into the SAC algorithm, using the error between the prediction results and the actual values ​​from a fully connected network as a reward signal to encourage the agent to explore the action space. An input vector format and dataset for Transformer training are constructed, and the input vectors are positionally encoded upon input. A suitable loss function is designed for backpropagation of the Transformer. The Transformer's output is used as the input to the SAC algorithm, and the SAC observations are reconstructed, forming an integrated decision network. This invention solves the game problem of satellite pursuit and escape using pulse maneuvers.

The utility model relates to the related technical field of leveling rod supports, and one embodiment of the utility model provides a leveling rod support which comprises a main frame and a leveling rod body, the leveling rod body is arranged in the main frame, a supporting and stabilizing assembly is arranged at the bottom of the main frame and the bottom of the leveling rod body, and a sleeving assembly is arranged on the main frame. The supporting and stabilizing assembly comprises a plurality of supporting legs which are arranged at the bottom of the main frame, moving wheels are arranged at the lower ends of the supporting legs, a pair of connecting rods are vertically and movably connected into the main frame in a sleeved mode, a bracket is arranged at the lower ends of the connecting rods, and the lower end of the leveling rod body is inserted into the bracket. By means of the technical scheme, the problems that in the prior art, the contact area between a supporting leg and the ground is small, and a self-adaptive adjusting function is lacked are solved. In the prior art, the support is easy to sink into the ground or incline under the soft ground or the uneven measurement environment, and even if the initial calibration and measurement process is slightly disturbed by external force, the measurement data are not accurate due to shaking.

This invention discloses an odor source localization method based on adaptive spatial perception information. The method includes the following steps: First, the search environment is divided into a fine grid, and a gas diffusion model that fits the actual scene is established to obtain the predicted concentration for each grid. Then, the real-time gas sampling data from the robot is binarized, and a sensor response model based on Poisson distribution is established to obtain the sampling concentration. The robot performs Bayesian inference using the predicted and sampled concentrations to update the likelihood function, thereby obtaining the posterior probability density function of the entire map. The reward function for each movable direction in the allowable action set is calculated using the posterior probability density function. The robot finds the direction with the largest change in the reward function, calculates the movement step size based on the current information entropy, and moves the robot a specified step size in the direction with the largest change in the reward function. This method solves the problems of existing odor source localization strategies easily getting trapped in local optima and the low search efficiency of fixed step sizes.