1316results about How to "Improve universality" patented technology

According to the goaf detection method based on transient electromagnetic and seismic wave field joint inversion, data acquisition and joint inversion are carried out on the goaf through the combination of the transient electromagnetic detection system and the reflection seismic detection system, the complementary advantages of the two detection methods are played, the bottlenecks of a single method in the aspects of resolution and signal-to-noise ratio are broken through, and the goaf detection method based on transient electromagnetic and seismic wave field joint inversion is realized. Fine characterization of the structure, the physical property and the stability of the goaf is realized; a machine learning classification algorithm is introduced to analyze a goaf multi-attribute feature data set, and intelligent discrimination of the type, scale and stability of the goaf under complex geological conditions is realized; establishing a quantitative response relationship between goaf physical property parameters and goaf electrical and elastic difference characteristic parameters, constructing a goaf transient electromagnetic and seismic wave field joint forward modeling numerical model, and constructing an electromagnetic-seismic joint inversion objective function; and calculating and solving to obtain goaf resistivity, seismic wave velocity three-dimensional distribution and uncertainty quantitative evaluation results, and reducing blind areas of goaf judgment.

The invention discloses an ROV positioning navigation system based on vision inertial navigation fusion and pose correction, which comprises an inertial measurement value and image preprocessing module loaded on an ROV, the inertial measurement value and image preprocessing module is connected with a vision and inertial navigation data fusion module through a wire, and the vision and inertial navigation data fusion module is connected with a pose correction module. The invention also discloses a positioning navigation method of the ROV positioning navigation system based on visual inertial navigation fusion and pose correction. According to the method, the inertial data provided by the IMU and the environment characteristic information captured by the visual sensor are fused, high-precision estimation of the position and attitude of the ROV can be realized, the sensor data are optimally fused by using the MSCKF algorithm, the influence of the error of a single sensor on the positioning result is reduced, and the positioning precision and reliability of underwater operation are remarkably improved.

The embodiment of the invention discloses a bridge crack width monitoring method. The method comprises the following steps: acquiring a continuous video image of a bridge surface; generating a pixel-level segmentation mask image of the crack by using the target detection model; when the target detection model is pre-trained, optical flow features between adjacent frames in the first data set are extracted through an optical flow calculation algorithm, and deep fusion is performed on the optical flow features and the image features; contour pixel points of the crack are extracted based on the crack segmentation mask pattern, and the width change information and the width change rate of the crack are obtained; according to the width change information and the width change rate, performing trend prediction on the crack width change by using a time sequence prediction model to obtain a prediction result of the crack width change; in the training process of the time sequence prediction model, a basic model and a complexity capture module are designed, and the weight of a main model is optimized by using a weighted fusion strategy. According to the invention, efficient and accurate crack width monitoring is realized through crack identification and prediction.

The invention is applicable to the technical field of biological information, and provides a drug target prediction method based on a graph neural network, which comprises the following steps of: respectively taking a drug and a target node as a core, and automatically extracting local sub-graphs of the drug and the target node to capture local context information of the nodes in a complex biological network; taking different types of edges as feature information and fusing the feature information into a feature learning process, dynamically adjusting contribution degrees of the different types of edges to node representation through an attention mechanism, introducing residual connection and a normalization processing strategy, and fusing node features; and performing deep mining on the fused node features by using a two-layer graph convolutional network, and calculating a potential interaction prediction score between the drug and the target. The invention further provides a drug target prediction model based on the graph neural network. According to the method, independent feature extraction and training are carried out on predicted drugs and target spots, and the accuracy and richness of node representation are effectively improved while the flexibility of the model is kept.

The invention provides a multi-degraded image restoration method based on self-adaptive prompt. The method comprises a two-stage hidden space diffusion model, and various degraded images can be flexibly processed. Specifically, the model generates a degradation cue by learning the difference between a degraded image and a clean image, and encodes the cue information in a hidden space. In the first stage, a degradation prompt is obtained through a degradation adaptive prompt learning mechanism as a key feature for guiding the diffusion model in the second stage to carry out image restoration. Finally, prompt information is integrated into a Fourier transform restoration network, and accurate restoration of different types of degraded images can be realized. The method not only improves the accuracy of image restoration, but also significantly reduces the calculation complexity, and is suitable for image restoration tasks in various real world scenes, including denoising, deblurring, defogging, rain removing, snow removing, raindrop removing and low illumination enhancement.

The invention relates to the technical field of battery control, in particular to a new energy automobile battery device and a control method thereof. The method comprises the following steps that dynamic parameter monitoring is conducted on an automobile battery, a battery state parameter set is obtained, and the battery state parameter set comprises a residual electric quantity continuous parameter, a voltage continuous parameter, a current continuous parameter and a temperature continuous parameter; according to the battery state parameter set, working condition identification is carried out on the automobile battery, a battery working condition type is obtained, and the battery working condition type is any one of a discharging working condition, a charging working condition and a standing working condition; if the battery working condition type is a discharge working condition, performing driving behavior data acquisition on the target automobile to obtain a driving behavior data set; and performing driving behavior pattern recognition on the target automobile based on the driving behavior data set to obtain a driving pattern type. The use efficiency and the safety of the battery are improved.

The invention discloses a slope stability evaluation method based on micro-seismic signal double-dominant-frequency mechanism characteristics, and relates to material analysis and application. The method is based on the objective principle that rock acoustic emission dominant frequency characteristics and micro-seismic signal dominant frequency characteristics have consistency and relevance, and the phenomenon of double dominant frequency characteristics existing in acoustic emission tests and micro-seismic monitoring. A rock sample dominant frequency characteristic rule obtained by an acoustic emission test is used for processing micro-seismic signal waveform data obtained by micro-seismic monitoring, so that a dominant frequency characteristic evolution rule of slope rock damage can be obtained through a micro-seismic signal, and the stress condition of the slope rock is more accurately judged based on the dominant frequency characteristic evolution rule; the stability of the slope is evaluated, so that the accuracy of an evaluation result is remarkably improved; the evaluation method combines an acoustic emission test and micro-seismic monitoring, has the advantages of high evaluation accuracy and good universality, and is suitable for large-scale application in monitoring and prevention of geological disasters such as landslide and the like.

The invention discloses a waterworks double-coagulant accurate prediction method based on automatic machine learning, and the method comprises the steps: firstly obtaining the water quality data of a waterworks, the agent dosage data of two coagulants, and other basic data, and carrying out the preprocessing of the data; an automatic machine learning method is adopted, and a tap water double-coagulant dosage prediction model under the production condition is built; the model is optimized according to an actual effluent target value, and accurate prediction of the dosage of the two coagulants is realized by using the optimal model. According to the invention, accurate calculation and intelligent regulation and control of the dosage of the coagulant in the waterworks are realized, and the predictive capability of the double coagulant is particularly emphasized, so that the innovation point ensures that the effluent quality reaches the drinking water standard, meanwhile, the dependence on manpower is remarkably reduced, the operation cost and the medicament cost of the waterworks are saved, and the economic benefit is increased. And double improvement of economic benefits and environmental benefits is realized.

The utility model relates to an automobile sealing strip fatigue resistance test device, the test device is used for carrying out fatigue resistance test on a sealing strip on an automobile window assembly, and the automobile sealing strip fatigue resistance test device comprises a test bench; the fixing mechanism is arranged on the test board and is used for fixing the vehicle window assembly; the pressing and releasing mechanism comprises a driving assembly and a profiling part connected with the driving assembly, the driving assembly can drive the profiling part to repeatedly press the sealing strip, and the profiling part is provided with a profiling face matched with the surface of the sealing strip. According to the fatigue resistance test device for the automobile sealing strip, through cooperative arrangement of the test bench, the fixing mechanism and the pressing and releasing mechanism, the fatigue resistance test can be performed on the sealing strip on the automobile window assembly in a low-cost and high-efficiency manner, and the fatigue resistance test device has relatively good universality and compatibility.

The invention discloses a laser welding spatter real-time recognition method based on OCT keyhole depth measurement. The method comprises the steps of S1, keyhole depth data collection based on OCT measurement and data set construction and S2, construction training of a deep learning model and spatter defect recognition. According to the method, an SD-OCT measurement technology is combined with a composite deep learning model based on 1DCNN-BiLSTM splash defect recognition, a Dropout technology and adaptive moment estimation are adopted to prevent model overfitting, and a TPE algorithm is adopted to carry out hyper-parameter optimization on the model, so that the trained 1DCNN-BiLSTM splash defect recognition model can accurately predict the splash state in the dynamic laser welding process, and the accuracy of the dynamic laser welding process is improved. Manual intervention is not needed, and the laser welding spatter defect can be rapidly and accurately recognized.

The invention belongs to the technical field of anchoring strength detection, and discloses an underground anchor rod anchoring strength prediction method and system based on a support vector machine method. The method comprises the steps of cleaning data, processing missing values and standardizing features; using correlation analysis or other feature selection methods to select features with high correlation with anchoring strength; performing model training by using a support vector machine method SVM, and optimizing hyper-parameters by using grid search; completing the construction of a support vector machine model; and carrying out model evaluation and result visualization. Key parameters in the mixing process are monitored in real time through the sensor network, the system can obtain anchoring strength data in real time, the detection period is greatly shortened, and the construction efficiency is improved. The support vector machine algorithm is adopted to analyze and predict the collected data, personal errors can be eliminated to a greater extent, and the accuracy and reliability of anchoring strength detection are improved.

The invention discloses a generative adversarial network sonar image denoising method and device fused with comparative learning, and the method comprises the steps: constructing a generative network, inputting a noisy image, and carrying out the image denoising processing through a fusion attention mechanism module and a comparative learning module; constructing a discrimination network, inputting a noise-free image without noise and a de-noised image generated by the generation network, and carrying out image discrimination; s3, training an adversarial network composed of the generative network and the discriminant network, and storing final parameters after training is finished; and S4, inputting the noise image into the trained generative network to obtain a denoised image. The fusion attention mechanism module and the contrast learning module can improve the ability of the generative network to understand and process the input image, enhance the ability of the network to capture and retain the image details, and accurately restore the main body features and detail information of the noise image.

The invention relates to the technical field of computers, in particular to a model confidence evaluation method and device, equipment, a medium and a program product, and the method comprises the steps: obtaining an information evaluation set; inputting test questions in the information evaluation set into the first model and the second model respectively to obtain to-be-evaluated answers output by the first model and reference answers output by the second model; inputting the reference answers and standard answers of the test questions in the information evaluation set into a pre-trained evaluation model, and outputting first deviation information between the reference answers and the standard answers; when determining that the test question and the standard answer meet a preset confidence condition based on the first deviation information, inputting the to-be-evaluated answer and the standard answer into an evaluation model, and outputting second deviation information between the to-be-evaluated answer and the standard answer; generating a confidence evaluation report of the first model according to the second deviation information; according to the scheme, the confidence test of the test questions and the standard answers is realized, and the universality and the mobility of the evaluation method are improved through the arrangement of the evaluation model.

The invention discloses a multi-degree-of-freedom ultrasonic-assisted laser cladding remanufacturing device and a working method thereof, and belongs to the field of laser cladding. Comprising a workbench, a to-be-machined workpiece arranged at the top end of the workbench through a positioning mechanism, a laser cladding mechanism erected above the to-be-machined workpiece and focusing type ultrasonic vibration mechanisms symmetrically arranged on the two sides of the laser cladding mechanism through a multi-degree-of-freedom variable-angle clamping mechanism. And focused ultrasound emitted by the focused ultrasonic vibration mechanism is aligned with a cladding area of the laser cladding mechanism. By adopting the multi-degree-of-freedom ultrasonic-assisted laser cladding remanufacturing device and the working method thereof, high-strength focusing of ultrasonic waves is realized through a multi-stage focusing method, and ultrasonic vibration can be uniformly applied during laser irradiation, so that the defects of cracks, air holes and the like are effectively reduced, and the processing quality and efficiency of laser cladding remanufacturing are improved.

The invention relates to the field of power systems, in particular to a power system load type intelligent identification method, system and equipment. The method comprises the steps that clustering centers of a plurality of load type subgroups are determined respectively, and each clustering center represents a typical load type calibrated in advance; utilizing a dynamic time warping (DTW) algorithm to calculate a DTW distance between the to-be-identified load node and each clustering center; according to the DTW distance obtained through calculation, distributing the to-be-identified load node to a subgroup of the nearest clustering center, and determining the load type of the to-be-identified load node; and periodically scanning and identifying load type changes of the load units in the domain, and updating a load type identification result. The invention further comprises a power system load type intelligent identification system and electronic equipment, the identification accuracy of the multi-type industrial load composite bus is improved, and misjudgment and deviation are reduced. The bus load type can be dynamically identified and changed and can be updated in time.

The invention provides an in-hole large-diameter main drive disassembling device and disassembling method, and relates to the field of shield tunneling machines, the in-hole large-diameter main drive disassembling device comprises a base platform with an upper platform and a lower platform, the lower platform is provided with a sliding block, the sliding block slides in the length direction of the lower platform, and the upper platform is provided with two bearing holes; the connecting assembly comprises a first connecting element, a second connecting element and third connecting elements, the first connecting element and the second connecting element are used for being fixed to the main drive, the two third connecting elements are fixed to the upper platform in the width direction, and the third connecting elements are hinged to the second connecting element; the disassembling assembly comprises a first telescopic element and a second telescopic element, the first telescopic element is connected with the first connecting element and the sliding block, and the second telescopic element is hinged to the sliding block; the supporting assembly is located below the bearing hole. According to the disassembling device, the first telescopic element and the second telescopic element are matched, so that the disassembling device can select elements with smaller lengths and strokes, the size of the disassembling device is smaller, and the disassembling device is more suitable for disassembling the main drive in a hole.

The invention belongs to the technical field of wireless communication, and discloses an optimization method based on active reasoning in an MEC-enabled low-altitude element cosmic system, which realizes low energy consumption and low delay of Internet of Things equipment on the premise of ensuring user service quality through joint optimization of a flight path and a resource allocation strategy of an unmanned aerial vehicle, and improves user experience. And meanwhile, an auxiliary means is provided for improving the economic benefits of operators. Aiming at the dynamic nature of the network, the uncertainty of information acquisition and the incompleteness of data use, an optimization problem is converted into a Markov decision process, an improved deep learning algorithm based on active reasoning is adopted, factors except the reward dimension are additionally considered on the guidance of dynamic learning, and the dynamic learning algorithm is optimized. And a personalized strategy is explored according to the preference of the intelligent agent.

The invention provides a method for synthesizing cyclic carbonate by catalyzing conversion of carbon dioxide with a ZIF-8-coated In2O3 composite material, and belongs to the field of catalytic chemistry. According to the method, carbon dioxide and an epoxy compound are used as raw materials, a ZIF-8-coated In2O3 composite material is used as a catalyst, tetrabutylammonium bromide is used as a cocatalyst, cycloaddition of the epoxy compound and normal-pressure carbon dioxide is performed to synthesize cyclic carbonate, and the ZIF-8-coated In2O3 composite material is prepared by in-situ compounding of ZIF-8 and In2O3 of a hollow prism structure. The ZIF-8-coated In2O3 nano-composite catalyst provided by the invention is simple to prepare, the raw materials are cheap and easy to obtain, the catalytic performance is excellent, and high-selectivity synthesis of cyclic carbonate under mild conditions can be realized. In addition, the catalyst is stable in structure, easy to separate from a product and good in recycling performance, and belongs to an environment-friendly catalyst.

The embodiment of the invention provides a code generation method and device, equipment and a storage medium, and relates to the technical field of computers.The method comprises the steps that a target language text corresponding to a target service is obtained, the target language text is compiled through a field specific language DSL, the target language text is analyzed, metadata corresponding to the target language text is obtained, and the metadata are stored in a database; codes of M different programming languages are generated according to the metadata and preset code template files of M different programming languages, M is a positive integer, and the code template file of one programming language comprises a calling structure of the programming language, a grammar structure of the programming language and template variables used for generating code snippets of the programming language. Therefore, the codes of various different programming languages can be generated based on the target language text corresponding to the target business, and the universality and efficiency of automatically generating the codes are improved.

The invention relates to the technical field of communication, in particular to an RoCE protocol-oriented reliable multicast transmission method, which comprises the following steps of: S1, dividing a multicast source and a multicast group in a hierarchical manner; s2, a sending end sends control type information through a UDP protocol, a receiving end monitors a control message, connection is established through an RDMA CM API, and QP ready state conversion is completed; s3, the sending end and the receiving end adopt an RC mode to carry out reliable data transmission, a forwarding node carries out concurrent forwarding in a Reactor multi-thread mode, and a main thread polls and receives a CQ event, determines a forwarding target node, copies message content to a sending buffer area and initiates asynchronous operation; and S4, carrying out API (Application Program Interface) development, and realizing SEND-RECEIVE bilateral operation and READ-WRITE unilateral operation in a multicast scene. The method has the advantages that a reliable multicast transmission function is realized, and a many-to-many concurrent forwarding scene is supported.

The invention discloses an electrostatic protection characteristic analysis method, device and equipment of a TVS (Transient Voltage Suppressor), and relates to the field of electrostatic discharge protection analysis. The method comprises the following steps: acquiring information data of a TVS tube; inputting the information data into the combined model to obtain electrostatic protection characteristic data; the electrostatic protection characteristic data is used for representing the electrostatic protection characteristic of the TVS tube; the combined model comprises an RLC circuit model and a double-capacitor model; the RLC circuit model is a physical simulation model obtained by carrying out optimization fitting by adopting a particle swarm optimization algorithm based on a parasitic effect; the double-capacitor model is a simulation model which adopts simulation software and realizes current waveform reproduction corresponding to electrostatic excitation based on electrostatic voltage; according to the invention, the accuracy and universality of electrostatic protection characteristic analysis can be improved.

The invention provides a wearable motion capture method and system based on a flexible mechanical sensor, and relates to the technical field of motion capture. The wearable motion capture method based on the flexible mechanical sensor is realized by the flexible mechanical sensor, a signal acquisition and wireless transmission unit and an intelligent analysis and interaction unit. The method comprises the following steps: a flexible mechanical sensor monitors an individual joint movement signal in real time, and transmits the movement signal to a signal acquisition and wireless transmission unit to obtain monitoring data of the flexible mechanical sensor; the monitoring data are transmitted to the intelligent analysis and interaction unit, the corresponding relation between the individual motion signals and the motion amplitude and the motion direction is established by calibrating the motion amplitude and the motion direction of the joints, and a calibration result is obtained; and according to a calibration result, three-dimensional motion modeling and attitude restoration of the individual are realized by utilizing a computer technology. According to the invention, the capability of finely capturing the local motion of an individual can be improved, the wearable performance can be improved, and the anti-interference capability is enhanced.

The invention relates to the technical field of brain-computer interfaces, in particular to a self-adaptive calibration method and system for a brain-computer interface earphone, and the method comprises the following steps: obtaining electroencephalogram data; based on the electroencephalogram data, constructing an initial model through federal learning; selecting a user specific sample by using an active learning strategy; performing dynamic parameter optimization on the initial model; based on the result of the dynamic parameter optimization, adaptive personalized calibration is executed; performing anomaly detection and correction on the result of the self-adaptive personalized calibration; performance evaluation is carried out on the result obtained after anomaly detection and correction, the evaluation result is fed back to the previous steps, and the steps form a closed-loop optimization process; through the combination of federal learning and active learning, the calibration time is shortened from 45 minutes of a traditional method to about 5 minutes, and the efficiency is improved by nearly 9 times; the accuracy rate of the method reaches 95.8% and is improved by 8.6% compared with that of a traditional method.

The invention discloses a code generation method and device, equipment and a storage medium, and relates to the technical field of computers. The method comprises the following steps: screening out a target reusable code meta-model corresponding to code generation demand information from a pre-constructed code model library according to the code generation demand information; the code model library comprises reusable code meta-models corresponding to various businesses, and the reusable code meta-models comprise one or more of an application template class model, a function module class model and a code snippet class model; verifying the target reusable code meta-model by using a pre-constructed rule base to obtain a verified target reusable code meta-model; and combining all the target reusable code meta-models passing the verification to obtain a program code corresponding to the code generation demand information. On the basis of ensuring the accuracy of code generation, the generation of multi-field program codes can be realized, and the universality of code generation in the application field is improved.

The invention provides a high-dynamic five-axis laser processing device based on double optical wedges, which can effectively improve the performance of five-axis laser processing equipment, and comprises a laser generator, a light field regulation and control module, a double-optical-wedge light beam offset module, a focusing and beam expanding module, a spectroscope, an observation system, a scanning module, a focusing module, a motion platform, a control system and operation software. The laser generator generates a laser beam, and the laser beam enters the double-optical-wedge light beam offset module after being regulated and controlled by the light field regulation and control module. The double-optical-wedge light beam deviation module deviates a laser beam from an optical axis, and the laser beam is parallel to the optical axis and enters the focusing and beam expanding module; the focusing and beam expanding module amplifies the off-axis amount of the laser beam and changes the divergence angle and diameter of the laser beam; then the laser beam enters the scanning module through the spectroscope; the laser beam is not parallel to the optical axis after passing through the scanning module, and finally the laser beam is transmitted to the machining surface on the motion platform after being focused through the focusing module.

The invention discloses a distributed photovoltaic system DC arc fault identification method and device in the technical field of photovoltaic systems, and the method comprises the steps: carrying out the non-overlapping sliding window sampling of the arc current of a target line, and generating the current data of each time window; calculating the current self-correlation quadratic sum of each time window according to the current data of each time window; normalizing the current self-correlation quadratic sum of t time windows before each time window as a standard value, and calculating the arc fault characteristic quantity of each time window; if the arc fault characteristic quantity of more than m time windows in the continuous n time windows exceeds the characteristic threshold value, judging that the arc fault exists; wherein t, n and m are constant parameters. According to the invention, direct current arc fault identification can be accurately and efficiently realized, and safe operation of the distributed photovoltaic system is guaranteed.

The high-voltage power distribution facility detection robot comprises a mobile platform, an inspection platform and an inspection mechanism, the inspection platform is arranged on one side of the mobile platform in the thickness direction, a bearing table is arranged on one side, away from the mobile platform, of the inspection platform, and the inspection mechanism is arranged on the bearing table. Wherein the inspection mechanism comprises an image acquisition part, an operation interaction part and a grounding switch execution part, the image acquisition part is used for acquiring image information, the operation interaction part is used for interactively operating a to-be-detected power distribution cabinet, the grounding switch execution part is used for controlling opening and closing of a grounding switch on the to-be-detected power distribution cabinet, the bearing table comprises a main table body and a lifting table, and the lifting table and the main table body are separately arranged; the operation interaction part is arranged on the main platform body, the image acquisition part is arranged on the lifting platform, a sub platform body is further arranged on the side, close to the movable platform, of the lifting platform, and the sub platform body can move relative to the lifting platform in the second direction.

The invention discloses a water quality comprehensive toxicity high-throughput determination method, which comprises the following steps: performing exposure contamination on a tested organism by using a sewage sample to obtain phenotypic characteristic data of the tested organism; constructing a toxicity matrix; establishing a machine learning model, and determining the water quality comprehensive toxicity of the sewage sample in combination with the toxicity matrix. The water quality comprehensive toxicity determination method provided by the invention does not need sample concentration and enrichment, ensures the comprehensiveness and accuracy of the detection result, and can be applied to the comprehensive toxicity high-flux determination of the sewage in the whole process of a sewage treatment plant.

The invention relates to a deep pseudo video detection method based on multi-modal comparative learning, and is suitable for the field of deep counterfeit detection. The method comprises the following steps: inputting a to-be-detected video into a trained video authentic identification model, and outputting a detection result of the to-be-detected video; the video authentic identification model comprises a visual encoder which is taken from a trained audio and video contrast learning model and is used for extracting visual features of a face sequence in a to-be-detected video; the audio encoder is taken from the trained audio and video comparison learning model and is used for extracting audio features of the to-be-detected video; the cross-modal feature fusion module is used for fusing the visual features extracted by the visual encoder and the audio features extracted by the audio encoder to obtain a feature fusion result; the spatio-temporal feature extraction module is used for extracting spatio-temporal features from the feature fusion result; the fusion features not only comprise visual features, but also comprise audio features, so that the features are richer; and the classifier is used for classifying the to-be-detected video as a real video or a forged video based on the spatial-temporal characteristics.

The invention discloses a computing power resource allocation control method and device, electronic equipment and a storage medium. The method comprises the steps of determining a resource utilization rate deviation corresponding to a current moment according to an actual resource utilization rate and a set resource utilization rate corresponding to the current moment of at least one graphic processing process associated with a graphic processor; acquiring a resource allocation control quantity corresponding to a previous moment and a kernel function cumulant corresponding to a current moment; and according to the control parameter prediction model, the actual resource utilization rate, the set resource utilization rate, the resource utilization rate deviation, the resource allocation control quantity and the kernel function cumulant, determining the resource allocation control quantity at the current moment, so as to determine an execution decision corresponding to the kernel function to be executed at the next moment of the current moment based on the resource allocation control quantity. According to the technical scheme, the PID control algorithm is optimized based on the neural network model, and the computing power resource allocation process of the graphics processor is dynamically controlled based on the optimized PID algorithm.