116 results about "Approximation algorithm" patented technology

The invention discloses a multi-modal fusion lithium iron phosphate battery thermal runaway early warning method and system, and the method comprises the steps: collecting multi-source heterogeneous data, such as temperature, voltage, gas concentration and shell strain pressure, in real time through deploying a heterogeneous sensor network, and carrying out the noise reduction and time sequence feature extraction through employing a sub-linear time low-rank approximation algorithm of a Hankel matrix; constructing a cross-modal feature association network by applying a secondary time algorithm of a maximum weight sparse subgraph problem, inputting a fused feature vector into a Bayesian network health degree evaluation model for probabilistic reasoning calculation to obtain a battery health degree score and a thermal runaway risk level, generating a graded early warning signal through multi-level early warning threshold comparison, and performing early warning on the battery health degree score and the thermal runaway risk level. And corresponding prevention and control suggestions are matched. The method solves the technical problems that single physical quantity monitoring is difficult to comprehensively reflect the complex change in the battery and the response delay of a centralized processing architecture causes the early warning lag, and achieves the timely capture and accurate early warning of the early weak characteristics of thermal runaway.

The invention relates to the field of power system planning, in particular to a micro-grid boundary quantitative evaluation method and system for multi-dimensional analysis and dynamic verification. The method comprises the steps that power distribution network and micro-grid scheme data are acquired, and scene recognition and decoupling modeling are carried out after standardization processing; the method comprises the following steps: extracting end supply-preserving scene data, analyzing cost elements, and generating a critical cost threshold table through normalization processing and a threshold approximation algorithm; performing multi-dimensional parameter correlation analysis and integrated learning training based on the table, and constructing a multi-dimensional boundary index model; performing benefit matching calculation and green value accounting according to the result to generate an economic benefit decomposition structure; real-time streaming data processing and stability verification are combined to generate an economical efficiency boundary index set; and finally, generating a standardized evaluation file through matrix mapping and weight dynamic adjustment. According to the method, dynamic quantitative evaluation of the economy boundary of the micro-grid is realized, the capacity substitution benefit and the green value are effectively integrated, and an accurate basis is provided for planning decision.

The present invention provides a method for optimization of active and passive beamforming and signal reception configuration of a dual IRSs-assisted ISAC system. By jointly optimizing active beamforming at a base station, the reception of a sensing signal by the base station, and passive beamforming at an IRS, the achievable rate of communication users is maximized while ensuring that the signal-to-noise ratio of the sensing signal meets the minimum requirement. In the present invention, to solve the complex nonconvex problem generated, first, fractional programming is used to decouple an optimization problem, then, a successive convex approximation algorithm and an alternating direction method of multipliers are used to transform an intractable nonconvex problem into multiple tractable subproblems, and finally, an alternative optimization method is used to efficiently solve for a high-quality sub-optimal solution. Simulation results show that the provided solution has good convergence and effectiveness, and the solution can effectively improve the performance of IRS-assisted ISAC systems.

A method for the joint active and passive beamforming and received signal optimization in an ISAC system assisted by dual IRSs is provided. The method jointly optimizes active beamforming at Base Station (BS), reception of sensing signals at the Base Station (BS), and passive beamforming at IRSs, so as to maximize communication sum-rate of users while ensuring that SNR of sensing signals meets a minimum requirement. To address the complex non-convex optimization problem, the method first applies fractional programming to decouple problem, then adopts successive convex approximation algorithm and alternating direction method of multipliers to transform intractable non-convex problem into multiple tractable subproblems, and finally employs an alternating optimization method to efficiently acquire the high-quality suboptimal solutions. The simulation results demonstrate that disclosed scheme exhibits satisfactory convergence and effectiveness, and can significantly improve the performance of IRS-assisted ISAC systems.

The invention belongs to the technical field of wireless optical communication, and particularly relates to a mobile edge computing task unloading and resource allocation method in wireless optical communication assisted by an unmanned aerial vehicle. According to the method, the data model, the queue model and the communication model are constructed, a task unloading strategy is optimized, and efficient and low-power-consumption edge computing resource allocation is achieved; an optimization technology is adopted to decompose a long-term optimization problem into sub-problems of each time slot, and the sub-problems are solved through a continuous convex approximation algorithm and a BFGS algorithm; finally, joint optimization of task unloading, computing resources, transmitting power, task return transmitting power, time slot scheduling and UAV trajectory is realized, and an optimization framework with more fine granularity in multi-dimensional allocation is formed. The characteristics of flexibility of the unmanned aerial vehicle and high bandwidth of wireless optical communication are utilized, the long-term throughput of the system is improved, the long-term power consumption of Internet of Things equipment is reduced, the high real-time performance of task processing is ensured, and the method is suitable for outdoor mobile Internet of Things application scenes with strict requirements for low power consumption and high real-time performance.

The invention discloses a digital twin high-voltage motor-based strong current operation state monitoring and maintenance system, which comprises the following steps: collecting current, temperature rise and vibration signals in motor operation, carrying out high-precision fitting processing on the signals by using a Chebyshev polynomial approximation algorithm, carrying out trend offset detection on a fitting result based on a CUSUM algorithm, and carrying out high-precision fitting processing on the fitting result based on a CUSUM algorithm; therefore, tiny abnormal changes are identified. And meanwhile, a prediction signal is generated in combination with a digital twinborn model, residual calculation is performed on the prediction signal and the fitting signal, fault features are further analyzed, and early warning information is output. Real-time sensing, trend judgment, fault prediction and intelligent early warning of the running state of the high-voltage motor are achieved, and the system is suitable for motor monitoring and maintenance in an industrial scene.

The invention discloses a two-stage Raman hyperspectral imaging method, which relates to the technical field of spectral signal analysis and comprises the following steps: S0, preprocessing Raman hyperspectral data to remove abnormal values; the method comprises the following steps: S1, taking a wave number where a target characteristic peak is located as a target wave number, and based on a neighborhood of the target wave number in Raman hyperspectral data, performing spectrum denoising on the Raman hyperspectral data by using an adaptive low-rank matrix approximation algorithm to obtain the Raman hyperspectral data after spectrum denoising; and S2, using an improved BM3D method based on a rotating block to carry out image denoising on a Raman image after Raman imaging is carried out on a target wave number in the Raman hyperspectral data after spectrum denoising. According to the method, the neighborhood of the target wave number is used as the target area, the adaptive low-rank matrix approximation algorithm is used for spectrum denoising, then the improved BM3D method based on the rotating block is used for image denoising, Raman hyperspectral fast and efficient imaging is achieved, and the method has the advantages of being good in denoising effect, high in speed and light in weight.

The invention discloses a symbiotic radio network hybrid beamforming method and system based on assistance of a reconfigurable holographic surface (RHS), and the method comprises the steps: constructing a communication architecture with the RHS as a transmitting antenna, and eliminating dual path loss; a dual-mode signal demodulation strategy is designed, the strategy I adopts MMSE-SIC combined with MRC technology to realize demodulation of an environment signal, deletion of a received signal from an SN through SIC and demodulation of a backscattering signal one by one, and the strategy II adopts design of the environment signal and a transmission frame structure thereof, combination of zero-forcing beamforming and demodulation of the backscattering signal to avoid transmission of an environment signal demodulation error; the transmitting power is optimized through a gradient descent method, the digital beam forming non-convex constraint is processed through fractional programming, the RHS amplitude control matrix is optimized through a sequence parameter convex approximation algorithm, and efficient alternating optimization of the three is achieved. The method has the advantages that energy efficiency of a multi-user scene is improved, system power consumption is reduced, dynamic network topology is supported, high reliability is kept, and limitation of an existing single-user scheme is broken through.

The present invention provides a single-stage high-frequency isolated DC / AC inverter, comprising a primary-side bridge inverter circuit, a high-frequency transformer, a series resonant cavity and a secondary-side AC circuit which are connected in sequence. The inverter is modulated by a control circuit, and an inverter modulation method is set in the control circuit and comprises the steps of: determining an inner phase shift angle between primary-side bridge arms, an outer phase shift angle between each primary-side bridge arm and a secondary-side bridge arm, and switching frequencies of a primary-side switching transistor and a secondary-side switching transistor on the basis of a fundamental harmonic approximation algorithm, so as to modulate control variables of the inverter, wherein the control variables include a power transmission characteristic and a soft switching characteristic; and calculating precise current values at switching instants of the primary-side switching transistor to compensate the control variables on the basis of the precise current values.

The invention provides an Internet of Vehicles DNN collaborative reasoning method based on a rate splitting multiple access technology. The method comprises the following steps: S1, establishing an edge computing system model; s2, establishing a communication model; s3, based on a rate splitting multiple access strategy, a road side unit RSU divides the information sent to each group of vehicles into a public flow and a private flow, and distributes a precoding matrix for each group; s4, in combination with a given DNN partition strategy and an RSU computing resource allocation method, calculating the communication rate received by each vehicle by using a continuous convex approximation algorithm; s5, on the basis of the communication rate and RSU computing resources, constructing a directed acyclic graph model of a DNN reasoning task, expanding the model into a network flow graph, and performing task partitioning on the model by adopting a genetic algorithm to minimize reasoning time delay to obtain a corresponding edge-vehicle partition node set; s6, converting an optimization problem into a KKT condition, and solving an optimal allocation scheme of RSU computing resources; and S7, repeating the steps from S4 to S6, and carrying out iterative updating to obtain an optimal solution of the system.

The invention discloses a construction method and evaluation method based on a vehicle network interaction multi-dimensional evaluation model, and the method comprises the steps: automatically recognizing all influence factors in a region through a machine learning algorithm based on intelligent power grid big data and geographic information data, and determining region-level evaluation parameters; fusing the evaluation parameters by a multi-dimensional data fusion method based on deep learning to obtain a comprehensive evaluation index; then, the number of input variables is reduced through feature selection and dimension reduction technologies, meanwhile, an approximation algorithm is adopted to replace complex mathematical operation, and a lightweight assessment model is constructed; and constructing a vehicle network interaction multi-dimensional evaluation model based on the lightweight evaluation model in combination with resource value information and multi-scene requirements. According to the method, the functions of differentiated scheduling, peak clipping and valley filling, frequency adjustment and emergency response are realized, the problems of resource mismatching, scheduling lag and poor regional adaptability caused by traditional static assessment are solved, the power grid stability, the scheduling efficiency and the user participation enthusiasm are improved, and the application of the V2G technology is promoted.

The invention discloses a dynamic beam hopping and resource allocation method for NGSO satellite security communication, and belongs to the field of sixth-generation mobile communication security communication and wireless resource allocation. According to the method, significant non-uniformity of distribution of ground flow requirements in a time domain and a space domain is considered, and a satellite-ground security communication network dynamic beam hopping and resource allocation problem model is constructed by considering ground user flow requirements and composition elements and channel characteristics of an NGSO multi-beam satellite network. Through joint optimization of satellite hopping beam scheduling, power resource allocation and auxiliary interference unmanned aerial vehicle deployment strategies, service requirements of different users and ground eavesdropping environments are dynamically adapted, and the safety throughput and queue delay fairness of the system are improved. According to the method, mixed integer linear programming modeling is adopted, and a low-complexity approximation algorithm is combined, so that the method not only has optimality guarantee, but also can be deployed and operated in an actual satellite communication system, and thus unification of secure communication and efficient resource scheduling is realized.

This invention discloses a floating-point-fractional hybrid anchoring method and computational system for suppressing iteration errors, belonging to the field of high-precision numerical computation technology. It solves the technical problems of pseudo-chaos caused by the accumulation of iteration errors in pure floating-point calculations, the computational power explosion in pure high-precision fractional calculations, and the inability to simultaneously achieve both computational power and precision. This invention completes high-speed iterative computations of nonlinear systems using a general floating-point format. After each floating-point calculation, an optimal rational number approximation algorithm is used to convert the calculation result into a simplified rational number with a denominator not exceeding a preset threshold, eliminating the original errors introduced by floating-point truncation and rounding. The purified rational number is then converted back to floating-point format to participate in the next iteration, relying on a bounded rational number anchoring mechanism to progressively block the error amplification chain. The denominator threshold is set to 10¹²~10¹ based on the effective number of bits of the floating-point number. 5 This invention can adaptively adjust to various scenarios such as embedded terminals, supercomputing, and industrial control, and optimizes the continued fraction algorithm to achieve the globally optimal rational approximation under a given upper limit of the denominator. It requires no dedicated large number hardware and combines the advantages of high-speed, low-load floating-point operations with zero approximation error in rational numbers. It can be stably used in nonlinear computing scenarios that are sensitive to initial values ​​and require long-term stable iterations, such as chaotic simulation, vehicle trajectory prediction, industrial PID closed-loop control, 5G / 6G communication channel simulation, microscale meteorological simulation, and large-scale number theory zero-point verification. It effectively eliminates numerical trajectory drift, reduces equipment computing power consumption, and minimizes the workload of manual calibration and data screening, possessing extremely high engineering and academic application value.

This invention discloses a multi-clinical-staging disease auxiliary classification method and system based on federated learning, belonging to the fields of big data and medical technology. To improve the accuracy of the classification model and the security of privacy data, this invention collects case data to construct a case dataset. This dataset is input into a client for analysis and processing to obtain an optimized case dataset. The client trains an XGGridBoost model on the optimized dataset. The trained model parameters are then encrypted and compressed, and sent to a central server for decompression. The decompressed model parameters are then securely aggregated, and an approximate algorithm is used to determine the optimal split point. The central server sends the optimal split point to multiple clients. Upon receiving the split point, each client determines whether its local decision tree has reached its maximum depth and performs iterative training or outputs the trained model. This invention solves the problem of data silos and improves the security of privacy data.

The present invention discloses a method for managing partially shared inventory of spare parts based on horizontal transfer, comprising the following steps: real-time monitoring of the degradation status of equipment, triggering preventive or post-event maintenance according to the equipment degradation status; after maintenance is triggered, if the local warehouse has sufficient spare parts inventory, directly performing the maintenance; if the inventory is insufficient, initiating a horizontal transfer request to other warehouses; using a Markov model to describe a multi-warehouse system; and using a decomposition approximation algorithm to solve the Markov model to derive key performance indicators of the multi-warehouse system. Compared with the prior art, the present invention optimizes critical thresholds based on the derived key indicators to achieve the lowest-cost inventory sharing strategy, thereby providing an efficient and scalable solution for maintenance management of modern industrial equipment.

The application discloses a seismic risk evaluation method and system, relates to the technical field of seismic numerical simulation, and receives a three-dimensional geometric model of a gridded fault and sets initial conditions; a simulation operation equation is constructed to solve a shear stress loading rate, a normal stress loading rate and a state variable, and a common differential equation system is constructed; based on the three-dimensional geometric model and the initial conditions, the common differential equation system is used to realize accelerated simulation operation of the shear stress loading rate, the normal stress loading rate and the state variable, wherein the method of the accelerated simulation operation comprises: using a method combining an H-matrix and MPI to accelerate simulation operation, comprising obtaining a layered matrix based on a Green function under a half-space condition in combination with an adaptive cross approximation algorithm, comprising a normal stiffness H-matrix and a shear stiffness H-matrix, and using MPI parallel operation matrix-vector multiplication. The application introduces a hierarchical matrix and fuses an MPI parallel computing framework, so that the simulation period is greatly shortened.

This invention discloses an infrasound localization method for minor pipeline leaks, comprising: acquiring the original infrasound signal of the pipeline leak; performing variational mode decomposition on the original infrasound signal to obtain IMF components; based on the IMF components, using an improved generalized cross-correlation function to obtain the time difference between the infrasound reflection to the upstream and downstream ends of the pipeline; calculating the wave velocity of the infrasound according to the pipeline characteristics; and determining the location of the pipeline leak point based on the time difference and wave velocity. The improved generalized cross-correlation function method involves: constructing an analytical signal based on the IMF components; using an improved linear adaptive weighted local approximation algorithm to obtain the local optimal solution of the IMF components; and performing inverse attenuation dispersion compensation analysis based on the generalized cross-correlation function to obtain the time difference between the infrasound reflection to the upstream and downstream ends of the pipeline. This invention can accurately locate the pipeline leak point.

The application discloses an automatic planning method and device for a walking route of a land controllable seismic source, and belongs to the field of physical exploration. The application introduces a TSP idea, and finally obtains a seismic source vehicle driving route which is short in path, less in barrier crossing, less in driving restriction, and in line with the driving operation habit of a driver through steps of constructing a map, setting relevant parameters and weight ratios, and solving the walking route of the land controllable seismic source by using a TNH combination algorithm. The application combines the advantages of a two-approximation algorithm, a Heuristic algorithm and an NSGA_II algorithm to propose the TNH combination algorithm for solving the walking route of the land controllable seismic source, so that the automatic planning of the walking route of the controllable seismic source is realized, the operation efficiency is accelerated, the accuracy of operation results is improved, the calculated route is more in line with the demand of field construction, and the exploration operation is further developed intelligently. The application is applied to the field of physical exploration to plan the walking route of the controllable seismic source.

This invention relates to the field of environmental perception fusion positioning technology, and provides a positioning method, system, and electronic device based on a Bayesian filtering neural network. The method includes: acquiring an observation waveform matrix generated by a base station after receiving a waveform signal transmitted by a target object at each sampling time; inputting the observation waveform matrix into a preset positioning model to obtain the positioning information of the target object; fitting the Bayesian filtering result to the positioning model using a recurrent neural network, which is trained using a cost function including motion likelihood and observation likelihood. The motion likelihood is generated based on the motion model of the target object sample, and the observation likelihood is generated based on the position information of the target object sample determined by a parameter estimation algorithm. This invention addresses the shortcomings of existing technologies where the application of Bayesian filtering algorithms in environmental perception fusion positioning requires approximation algorithms and complex derivations to obtain positioning results, thus affecting positioning speed and accuracy.

The application discloses a cloud-edge collaborative real-time video analysis task execution delay modeling and deployment method, comprising: a video analysis task execution delay model construction method under a cloud-edge environment; the delay model is an additional delay introduced in a video analysis task process, comprising a delay caused by edge end processing, a delay caused by intermediate data transmission between the edge and the cloud, and a delay caused by cloud processing; according to the video analysis task execution delay model under the cloud-edge environment, an approximate algorithm based on the Lyapunov algorithm and the Markov approximation algorithm is used to select a more energy-saving task deployment scheme for the video analysis task, wherein the task deployment scheme comprises edge end machine specification selection, cloud machine instance specification and a task division scheme. The application can quickly adjust the task deployment scheme when the task load of the video analysis task and the cloud-edge network environment change, thereby effectively reducing the long-term operation energy consumption of the system.

The invention belongs to the technical field of power system data processing, and particularly relates to a mixed integer programming-based fully-distributed bad data identification method, which comprises the following steps of: firstly, decoupling a centralized system into a plurality of independent subsystems at a tie line by adopting a node copying and consistency variable introducing method to realize parallel independent calculation; and then establishing a sub-system bad data identification model based on mixed integer programming, and solving by adopting an external approximation algorithm, thereby realizing rapid and accurate identification of measurement bad data of each sub-region of the power distribution network, solving the limitations of a traditional method on calculation efficiency, privacy protection and communication, and providing technical support for realizing accurate perception of the power distribution network. The problems that a traditional centralized method is low in efficiency, difficult in communication and poor in privacy protection are effectively solved, and the timeliness and privacy of bad data identification are improved.

The application discloses a fluid antenna assisted ISAC system joint beam forming and position optimization method based on a quantum particle swarm optimization algorithm. Specifically, an ISAC multi-user system model equipped with a two-dimensional fluid antenna is established, and a beam forming optimization problem is constructed; a weighted least mean square error algorithm is used to convert a communication and rate target function, so that a complex fractional problem is converted into an integral problem, and solving is simplified; an alternating optimization algorithm is used to decompose the original joint optimization problem into multiple sub-problems; an auxiliary variable and a receiving beam former are alternately solved, an iterative solution of a beam forming vector is solved by using a successive convex approximation algorithm, and a fluid antenna position problem is solved by using a quantum particle swarm optimization algorithm. The method of the application effectively improves the communication and rate under the restrictions of the fluid antenna position, the minimum perceived beam pattern gain and the maximum base station transmission power, and is superior to a traditional ISAC system equipped with a fixed antenna.

The present application relates to the technical field of visual domain analysis of digital elevation model, and particularly relates to a DEM visual domain analysis method and system based on volume shadow, which comprises the following steps: S1, obtaining digital elevation model DEM data and user input viewpoint parameters, and constructing a regular triangulated irregular network model based on the DEM data; S2, calculating the yin and yang properties of each triangular face based on the viewpoint parameters, and determining the shadow edge according to the yin and yang properties of adjacent triangular faces and the horizontal distance to the viewpoint. In the present application, the DEM visual domain analysis problem is converted into a volume shadow problem with the viewpoint equal to the light source, and the interpolation processing of the terrain by the DEM data and the GIS platform is neat and standard, which simplifies the calculation related to the volume shadow, realizes the precision of the accurate algorithm, approaches the operation speed of the approximate algorithm, and can consider the inclined surface generated by the interpolation processing of the terrain.

Embodiments of the present application disclose a data processing device and method based on an NPU processor. After determining a target input value, the device and method introduce an auxiliary value, obtain a first calculation value corresponding to the target input value and a second calculation value corresponding to the auxiliary value based on a Taylor expansion, and finally determine a target trigonometric function value from the first calculation value and the second calculation value according to the size of the target input value. Thus, when running the Taylor expansion approximation algorithm in the NPU processor, the accuracy of the value range of the part of the trigonometric function that is difficult to ensure is ensured by processing the auxiliary value corresponding to the target input value.

This invention relates to the field of detection, and in particular to an online method and system for detecting impurities in calcium carbonate ore. Specifically, it includes: acquiring a continuous multispectral image sequence of calcium carbonate ore on a conveyor belt, and performing radiometric correction and spectral normalization preprocessing; identifying dominant spectral clusters of pure calcium carbonate ore by statistically analyzing the pixel spectra of the preprocessed images, and selecting a subset of high-confidence pixel spectra as the initial background training set; training a background dictionary, applying a set of sparse regularization constraints during training, and using the correlation between spectral bands to constrain the atomic energies of the dictionary to continuous or strongly correlated band groups; for each pixel of the detected image, using a sparse approximation algorithm combined with the background dictionary, calculating a sparse reconstruction coefficient vector and a reconstruction residual vector, and obtaining the reconstruction residual and coefficient activation uncertainty measure, respectively; if the reconstruction residual exceeds a first threshold, or the uncertainty measure exceeds a second threshold, it is determined to be an impurity point, and the location and quantity are summarized to generate an online impurity distribution map.

The invention provides an anti-cooperative detection eavesdropping method and system in an unmanned aerial vehicle hidden security network, and the method comprises the steps: obtaining the maximum security communication rate of each legal user in each time slot under the cooperative eavesdropping of a plurality of illegal users based on an unmanned aerial vehicle uplink communication network scene of the plurality of legal users and the plurality of cooperative illegal users; based on an unmanned aerial vehicle uplink communication network scene of multiple legal users and multiple cooperative illegal users, obtaining a minimum detection error rate of a communication behavior of each legal user under cooperative detection of the multiple illegal users; based on the maximum security communication rate and the minimum detection error rate, establishing a joint optimization problem taking maximization of the minimum security throughput of the system as a target, and setting constraint conditions; and solving the joint optimization problem based on a continuous convex approximation algorithm to obtain an optimized unmanned aerial vehicle trajectory, user transmitting power and time slot segmentation scheme.

The application relates to a direction modulation-based covert secure transmission method and a communication base station, and belongs to the technical field of wireless communication. In view of the case that channel state information is non-ideal and distortion noise exists, a robust direction modulation technology is adopted, a continuous convex approximation algorithm is used to optimize and select a transmitting end, a combination with low secrecy capacity is eliminated, the secrecy capacity is improved, the possibility of eavesdropping by a eavesdropping user is reduced to a certain extent, and physical layer secure communication is further ensured.

The application discloses a G1 continuity curved surface reconstruction and generation method based on boundary constraint, and belongs to the technical field of computer aided design and computer graphics technology. Step 1: taking a three-dimensional industrial model to be repaired as input, a unified normalized parameter mapping is established to obtain a standard boundary data set; step 2: taking the standard boundary data set as input, a basic cross-border tangent vector field is obtained; step 3: taking the basic cross-border tangent vector field as input, an original modified cross-border tangent vector field is obtained; step 4: taking the original modified cross-border tangent vector field as input, a tangent plane geometric constraint mechanism is established, and a final cross-border tangent vector field is obtained; step 5: taking the final cross-border tangent vector field as input, a rational harmonic formula is introduced, and four corner point torsion vectors after rational harmonic are obtained; step 6: a high-order B-spline geometric entity meeting an industrial tolerance is generated through a reverse approximation algorithm, and a B-spline curved surface is obtained. The application can automatically generate a high-quality transition curved surface without high-light tearing under complex and disordered boundary constraints.

The application belongs to the technical field of artificial intelligence and digital image processing, and discloses an extensible farmland boundary regularization simplification method, system, device and terminal, determines a farmland category list to be segmented; uses an existing labeled training segmentation model to obtain segmentation masks of all farmlands; performs binaryzation processing on the segmentation masks to obtain binaryzation images; performs digital image processing on the binaryzation images to extract the contours of each connected domain; uses the contour information obtained through image processing to perform polygon approximation and remove redundant points to obtain farmland boundary representation. The processing method for extracting local farmland regular boundary uses the farmland segmentation result obtained through deep learning segmentation, uses a traditional image processing method and is supplemented with a polygon approximation algorithm to perform regularization processing on the result, and simultaneously simplifies the contour points of the farmland, so that the number of farmland boundary points and boundary description accuracy can be better balanced while reducing labor cost.

The application belongs to the field of precise batching, and specifically discloses an intelligent precise batching method, which comprises the following steps: acquiring the number of saggers and the single-sagger load; acquiring the batching quantity of each material in multiple materials through an approximation algorithm; according to the adding sequence of the materials, performing actual feeding of a first added material, and determining the actual batching quantity of the first added material according to the batching quantity of the first added material calculated in the previous step; determining the batching quantity of a second added material according to the actual batching of the first added material; performing actual feeding of the second added material, and determining the actual batching quantity of the second added material according to the batching quantity of the second added material calculated in the previous step; repeating the steps S4-S5, and sequentially performing actual feeding of a third added material, a fourth added material, and so on until actual feeding of a last added material; and according to the batching proportion, feeding each material into a mixing equipment area in batches according to a set number of times.