51 results about "Continuous dynamic" patented technology

The invention provides a maglev train multi-suspension system predictive control method based on event triggering, which comprises the following steps: establishing a multi-electromagnet suspension system continuous dynamic model based on electromagnet parameters of a maglev train; in combination with the input time delay of the multi-electromagnet suspension system, the continuous dynamic model of the multi-electromagnet suspension system is sequentially processed based on a fourth-order Runge-Kutta method, and a discrete coupling time delay dynamic model of the multi-electromagnet suspension system is obtained; establishing a model prediction controller of the multi-electromagnet suspension system by combining the discrete coupling time delay dynamical model of the multi-electromagnet suspension system and the optimization problem; and updating a control signal of the model prediction controller according to a preset composite event triggering mechanism to obtain an initial control quantity, designing a time delay compensation strategy to carry out input time delay compensation on the initial control quantity, and obtaining a standard control quantity to control the multi-suspension system of the maglev train. According to the method, the system energy consumption is remarkably reduced while the good operation performance of the suspension system in a complex environment is guaranteed.

The invention relates to the technical field of bearing detection, and particularly discloses a deep groove ball bearing clearance assembly error detection method and system.The method comprises the steps that dynamic parameters of a deep groove ball bearing in each assembly stage in the assembly process under preset assembly parameters are obtained, and the dynamic parameters comprise vibration spectrum data, temperature distribution data and shaft diameter displacement; obtaining a dynamic clearance error value according to the plurality of continuous dynamic parameters; and obtaining stress gradient distribution data at a plurality of time nodes in each assembly stage according to the dynamic clearance error value. The deep groove ball bearing assembly quality evaluation and analysis method has remarkable beneficial effects in the aspects of deep groove ball bearing assembly quality evaluation and analysis. Through a series of closely associated steps, the actual state of the bearing is comprehensively and accurately reflected by comprehensively considering multiple parameters to accurately calculate local stress and judge the stress area from accurate judgment of a thermal-mechanical coupling stress area, obtaining of an external dynamic deformation comprehensive index and calculation of a total assembly error value and a clearance comprehensive evaluation index.

The invention provides a rotating machine fault diagnosis method and a related system. The rotating machine fault diagnosis method comprises the following steps: S1, constructing a rotor vibration equation according to a vibration behavior of a rotor system under an operation condition; s2, acquiring real-time information of a to-be-diagnosed rotating machine, performing modal analysis based on a rotor vibration equation according to the real-time information of the to-be-diagnosed rotating machine to determine a sensitive frequency band and key features, and forming multi-dimensional feature data; s3, carrying out denoising processing on the multi-dimensional feature data and carrying out signal reconstruction to obtain a reconstructed signal; s4, inputting the reconstruction signal into a pre-trained rotating machine fault model to obtain a fault diagnosis result; and the rotating machine fault model is fused by adopting a neural network ordinary differential equation and a convolutional neural network-gating cycle unit. The method can accurately simulate and predict the continuous dynamic change of the system, improves the diagnosis precision, and enhances the real-time performance and robustness of the model.

The invention discloses a roadway surrounding rock dynamic accurate control method, and relates to the technical field of underground roadway surrounding rock control. The technical problem that in the prior art, blindness exists in support design in roadway surrounding rock control is solved. The method comprises the steps that a data collection and analysis system is used for collecting and analyzing mine geological data and evaluating geomechanics key indexes; the stability influence factor analysis system evaluates the existing roadway support stability by classifying the roadway surrounding rock stability; a targeted supporting scheme is provided in combination with an existing supporting scheme, related standards are established, and continuous dynamic optimization and improvement are conducted on follow-up supporting design. According to the method, dynamic analysis and targeted design are carried out on the roadway, roadway deformation can be effectively controlled, the comprehensive cost of roadway supporting is reduced, the roadway tunneling efficiency is improved, the problem of coal mine excavation imbalance is solved, safety accidents caused by supporting design are avoided, and dynamic precise control over surrounding rock of the underground roadway is achieved.

The invention provides a finite element simulation method for a ribbed wallboard formed through laser shock. The finite element simulation method comprises the steps that initial laser shock pressure time-space distribution corresponding to laser process parameters is determined; applying the impact pit as a load to a laser multi-point continuous dynamic impact finite element model to obtain a simulated impact pit size; judging whether the size of the simulation and experiment impact pit is within an allowable error range or not; a strain field and a residual stress field after laser shock are obtained, and average inherent strain distribution in the depth direction is output; and introducing the average inherent strain distribution into the layered shell unit model of the ribbed wallboard and solving to obtain a deformation result of the ribbed wallboard impacted by the laser process parameters. Dynamic simulation of laser multi-point continuous impact is realized through display analysis, and deformation simulation of the ribbed wallboard formed through laser impact is realized through thermoelastic analysis, so that effective forming simulation can be performed on the complex ribbed wallboard, and the effects of low cost, high efficiency and accurate prediction result are achieved.

The invention relates to the technical field of power system inertia evaluation, in particular to a random mode switching driven power distribution network equivalent inertia online evaluation method and system, and the method comprises the steps: carrying out the priori definition of a group of hidden operation modes based on the equivalent inertia of a power distribution network, and constructing an unobservable Markov chain; a double-layer hidden Markov jump system model is constructed, the bottom layer is an unobservable Markov chain, and the upper layer is a continuous dynamic behavior model for describing the power distribution network; on the basis of PMU high-resolution time sequence disturbance data, deducing the probability of a system maximum probability dominant operation mode and the probability of each hidden operation mode; and carrying out multi-mode probability adaptive weighting on each equivalent inertia estimation value based on the probability of the dominant operation mode and each hidden operation mode so as to obtain the equivalent inertia estimation value of the power distribution network on line. Through the method, the problem that the equivalent inertia of the power distribution network is difficult to quickly, accurately, dynamically and adaptively assess online under the conditions of high permeability of new energy and frequent switching of working conditions is effectively solved.

The invention relates to a flight control technology of an unmanned aerial vehicle suspension load system, and provides a multi-target nonlinear model prediction control method for an unmanned aerial vehicle-load suspension system, so as to realize accurate track tracking, active suppression of lifting rope swinging and collaborative optimization of flight smoothness. According to the multi-state multi-target model prediction control method for the unmanned aerial vehicle-load suspension system, firstly, a nonlinear model prediction control equation is constructed; discretizing the continuous kinetic model by adopting a fourth-order Runge-Kutta method to obtain a high-precision state transition equation under discrete time as a dynamic constraint; forming an optimal control function of a flight nonlinear model predictive control equation of the unmanned aerial vehicle-load suspension system; adding thrust amplitude constraint and moment boundary constraint as inequality constraint conditions of a nonlinear model predictive control equation; and finally, multi-target optimal control of the unmanned aerial vehicle-load suspension system is realized. The method is mainly applied to design and manufacturing occasions of unmanned aerial vehicle hanging load systems.

The invention belongs to the technical field of aero-engine component test environment construction, and provides a dynamic variable speed air inlet channel test method and system under continuous air supply and exhaust conditions, and the method comprises the steps: constructing a spray pipe profile, a tester spray pipe inlet and outlet pressure ratio, air supply and exhaust pressure and a sub-model which is continuously adjusted along with the speed and is used for balancing the flow; inputting a dynamic change requirement of a target speed along with time in a continuous variable speed test under the condition of not stopping air supply into each control sub-model, and establishing a variable speed regulation model of a spray pipe profile, a pressure ratio of an inlet and an outlet of a tester spray pipe, air supply and exhaust pressure balance, a test piece posture and the like along with time in combination with a control rule of each execution mechanism; and a continuous variable-speed flow field meeting variable-speed and time requirements under the condition of uninterrupted air supply and extraction is carried out. According to the method, the blowing test with the speed continuously and dynamically changing can be carried out under the conditions of not stopping air supply and not interrupting operation of a tester, and reliable test data support is provided for design and optimization of an air inlet channel of an advanced aircraft.

The invention discloses a reservoir scheduling calculation method based on a differential form. The method comprises the following steps: S1, collecting basic data and parameters; s2, establishing a water balance equation in a differential form according to engineering precision requirements; s3, according to different scheduling modes, according to the initial state of the time period, obtaining the reservoir outlet flow at the end of the time period, substituting the reservoir outlet flow into a water balance equation in a differential form to solve the reservoir capacity at the end of the time period, and querying a water level-reservoir capacity relation curve to obtain the reservoir water level at the end of the time period; and S4, taking the calculated water level of the reservoir at the end of the time period and the reservoir outlet flow as initial values of the next differential time period, and repeating the step S3 until analog calculation of the whole scheduling period is completed. According to the method, a zigzag process line is effectively avoided, continuous dynamic simulation of the reservoir water level and the reservoir-out flow is achieved through tiny time period iteration, the critical point of a control operation plan, reservoir-in flow change or gate opening adjustment can be responded in time, and the problem of water level oscillation caused by constant variables in time periods in a traditional method is avoided.

The application discloses an energy balance-based asymmetric temperature control system simulation modeling method, and belongs to the technical field of industrial temperature control system modeling and simulation. Firstly, a temperature rising step experiment and a temperature falling step experiment are performed, steady-state data is acquired, a balance point power function is derived, equivalent verification is performed, a temperature rising curve and a temperature falling curve are derived, and identification parameters are derived. A net power deviation is defined, the identification parameters are subjected to smooth transition, gain parameters are acquired, a first-order linear differential model is established, a first-order backward difference method is adopted for discretization, recursion is performed based on a set simulation step, and a recursion model is acquired. A plurality of recursion models are derived based on repetition of the simulation step, a temperature response curve of the temperature control system is established, and simulation modeling is completed. The method only needs two step experiments, real-time parameters are calculated through the net power deviation, real smooth and continuous dynamic simulation is realized, and a reliable simulation method is provided for temperature digital twinning and temperature controller design.

The invention discloses a next interest point recommendation method based on multi-aspect continuous dynamic preference modeling of hypergraph ODE. According to the method, modeling can be efficiently carried out on multi-aspect dynamic preferences of the user from a continuous angle, and more personalized recommendation is provided for the user. Multi-aspect user preferences are decoupled by constructing a multi-view hypergraph from collaborative, geographic, and transformational aspects. By expanding discrete hypergraph convolution operation into a continuous form based on an ordinary differential equation (ODE), continuous evolution of multi-aspect user dynamic preferences is modeled. User preferences under different views are integrated by adopting a self-adaptive fusion mechanism, and a dynamic relationship between the user preferences is captured. In addition, an inter-view and in-view comparative learning normal form is designed to enhance the expressivity and robustness of interest point representation. A large number of experiments carried out on four real spatio-temporal data sets with different scales verify that the method has remarkable advantages and efficiency.

The present application belongs to the technical field of test environment construction of aero-engine components, and provides a dynamic variable-speed inlet test method and system under continuous air supply and extraction conditions, the method comprising: constructing a nozzle profile, a test device nozzle inlet and outlet pressure ratio, and a speed-continuous adjustment sub-model of air supply and extraction pressure and flow balance; inputting the dynamic change requirement of the target speed with time during continuous variable-speed test under non-stop air condition into each control sub-model, combining the control law of each actuator, establishing a variable-speed control model of the nozzle profile, the test device nozzle inlet and outlet pressure ratio, the air supply and extraction pressure balance, and the test piece attitude with time, and performing continuous variable-speed flow field under the conditions of uninterrupted air supply and extraction to meet the variable-speed and time requirements. The method can perform speed-continuous dynamic change blowing test under the conditions of non-stop air and uninterrupted test device operation, and provides reliable test data support for advanced aircraft inlet design and optimization.

A method for calculating continuous dynamic derivatives based on the principle of periodic cross-correlation belongs to the field of dynamic derivative measurement technology. In order to obtain dynamic derivative data that continuously changes with the attitude angle, the present invention includes performing dynamic derivative tests under windless and windy conditions, and after reaching the target frequency and stabilizing, continuously changing the attitude angle, obtaining the balance load and vibration angle, and deducting the trend item of the balance load changing with the attitude angle, and then performing periodic and discrete cross-correlation processing on the balance load and vibration angle displacement after deducting the trend item to obtain the damping derivative under a single vibration cycle under windless and windy conditions. The average attitude angle under each vibration cycle is obtained, and the corresponding relationship between the damping derivative under windless conditions and the average attitude angle and the corresponding relationship between the damping derivative under windy conditions and the average attitude angle are established respectively; the damping derivatives under windy conditions and those under windless conditions are subtracted one by one to obtain the dimensional aerodynamic damping derivative, and then dimensionless processing is performed to obtain the continuous dynamic derivative.

The invention relates to the technical field of power load prediction, and discloses a federated power load prediction method based on dynamic clustering and layered personalized aggregation, comprising the following steps: constructing a transverse federated learning framework; each client carries out local load data independent training and uploads updated model parameters to the server; the server calculates similarity by using cosine similarity so as to dynamically adjust the structure of the clusters, and deletes inactive clusters by judging whether the clusters are active or not according to training rounds and setting an active period so as to adjust the number of the clusters; the server executes hierarchical federal aggregation on each cluster model based on the current clustering result; and the server respectively issues the aggregated cluster model to each client of the corresponding cluster for next training, thereby realizing continuous dynamic optimization. The method has obvious advantages in key indexes such as prediction precision and convergence rate, and the load prediction performance under the federated framework is improved.

The invention provides a continuous dynamic compaction reinforcement foundation treatment construction method and system, and belongs to the field of building construction. According to the technical scheme, the method comprises the steps of collecting multi-source geological survey data to establish a digital twinborn body; virtual dynamic compaction training is carried out in the digital twin body, and an initial construction strategy is output; obtaining construction response data, and updating state parameters of the digital twinborn body; generating an optimization construction instruction based on the updated digital twinborn body; and the dynamic compaction equipment is controlled to execute the optimized construction instruction to complete compaction operation, and continuous construction is achieved. The method has the advantages that intelligent closed-loop control over the whole dynamic compaction construction process is achieved by combining the digital twinborn body and the real-time data assimilation technology, compaction parameters are adjusted in real time according to the actual response of the soil body in the construction process, the construction effect is ensured, equipment path planning is optimized through the reinforcement learning algorithm, and the construction efficiency is improved. The problem of equipment shutdown standby in a traditional construction method is solved, and the overall efficiency and safety of dynamic compaction reinforcement construction are effectively improved.

This invention provides a method for analyzing the dynamic characteristics of a corrugated sidewall belt lifting system for a ship unloader. The method includes: constructing a fractional derivative viscoelastic model of the corrugated sidewall conveyor belt and establishing the basic dynamic equations of the corrugated sidewall conveyor belt; establishing discrete dynamic equations for the conveyor belt, the drive device, and the tensioning device, and establishing a discrete dynamic model of the corrugated sidewall belt lifting system; establishing continuous dynamic equations for the conveyor belt, setting boundary conditions and initial conditions, and establishing a continuous dynamic model of the corrugated sidewall belt lifting system; solving the discrete dynamic model or the continuous dynamic model of the corrugated sidewall belt lifting system to obtain the acceleration, velocity, and displacement of the corrugated sidewall belt lifting system, thus obtaining the dynamic response of the corrugated sidewall belt lifting system. The use of a fractional derivative viscoelastic model, and the further establishment of a dynamic model, more realistically reflects the dynamic characteristics of the corrugated sidewall belt lifting system at each stage.

The invention provides a multi-rotor aircraft power distribution method and system based on dynamic load perception, and relates to the technical field of aircraft intelligent control. Initializing the system, and loading initial parameters of the aircraft and the power system; collecting flight attitude, rotating speed of each rotor, power consumption and temperature of a motor and control instruction data in real time; estimating the actual thrust of each rotor wing in real time based on the acquired data and a built-in dynamic model; receiving a total thrust and torque instruction sent by a flight control system, and calculating a target thrust sequence of each rotor through an optimization algorithm in combination with the actual thrust estimated by each rotor; converting the target thrust sequence into a control instruction of each motor and issuing the control instruction; real-time state feedback and model updating are carried out, and continuous dynamic optimization of power distribution is achieved. According to the method, the flight energy efficiency, stability and system robustness of the multi-rotor aircraft can be enhanced, and a solid technical support is provided for application of the multi-rotor aircraft in wider fields.

The invention discloses a ship actual load ton continuous dynamic measurement system, which comprises a plurality of measurement devices, a cable, a pulley, a data collector and a data center, and is characterized in that an ultrasonic device and a pressure sensor are integrated in each measurement device. Core measurement components of the device are only an ultrasonic device, a pressure sensor and conventional data collector equipment; the overall cost is low and the economical efficiency is remarkable. According to the invention, the measurement device can be manually and directly controlled to be retracted and released from the ship deck without underwater operation or special hoisting equipment; the folding and unfolding operation is convenient, and the adaptability is high. Data can be collected in real time in the whole cargo loading and unloading process, real-time data support is provided for dynamic regulation and control of loading and unloading operation, and economic losses caused by loading and unloading weight deviation are avoided. According to the invention, a plurality of measuring devices are deployed along the transverse direction of the ship, the data center synthesizes the measurement data of each position, each subtle change of the real load ton is accurately captured, and the accuracy of the measurement data of the real load ton change is improved.

The present invention belongs to the field of industrial Internet and control technology. In view of the phenomenon that continuous dynamic processes, discrete events, and the diversity of controlled objects and multiple time scales coexist in industrial Internet systems, a hybrid singular system modeling and stability analysis method based on multiple sampling rates is proposed to solve the stability analysis and controller design problems of the proposed model, while considering the influence of multiple sampling rates and singular perturbation parameters on system stability. The present invention can achieve the complex characteristics of multiple system couplings such as continuous dynamics, multi-sampling, discrete events and multiple time scales, that is, effectively improve the system operation performance and anti-interference performance. At the same time, the hybrid singular perturbation system model established by the Lyapunov function method is applicable to more complex actual systems and also provides a better compromise between system performance and execution cost, that is, it has strong feasibility, which means that it has good practical application prospects.

The present application relates to the field of prefabricated component transportation scheduling of assembly type building, and particularly discloses a dynamic transportation scheduling optimization method for prefabricated components with interference of assembly type building, which designs an improved NSGA-II algorithm combining a simulated annealing mechanism and an adaptive strategy; for the multi-objective optimization problem of dynamic transportation scheduling of prefabricated components with interference, the present application combs three types of interference events, i.e., emergency orders, order modifications and vehicle transportation time changes, which occur frequently in the transportation process, and formulates dynamic rescheduling strategies for different disturbance types; in combination with a rolling horizon optimization method, the continuous dynamic scheduling problem is converted into a series of static sub-problems, and then static optimization schemes are adopted in each static process to realize dynamic optimization.

The application discloses a world action model enhancement method and device fusing structured prior and continuous dynamics, wherein the method takes human videos in a first view and a third view as input, constructs scene-event graphs and structured priors such as causality, counterfactual constraints and execution constraints; under weak labeling conditions, performs spectrum analysis on local time sequence features of actions, and jointly encodes the time sequence features and space-time features into latent actions; further, continuous time latent dynamics modeling is introduced into a unified world action model through a neural control differential equation, and joint reasoning is performed on video state changes and latent action recovery in combination with task semantics, structured priors and counterfactual conditions. Finally, long-time rolling error accumulation is inhibited through real observation rewriting, action block generation and asynchronous closed-loop execution, so that the feasibility of weakly labeled video action learning, the rationality of action generation and the stability of closed-loop control are improved.

The invention discloses a continuous dynamic loading system for simulating road ice and snow state evolution under the rolling action of a vehicle, belongs to the technical field of railway detection equipment, and aims to solve the problem that the prior art cannot meet the requirements of continuity, multiple working conditions and high-efficiency simulation in the research of a coupling mechanism of traffic load and ice and snow state dynamic evolution. The device comprises a frame unit, a rotating and loading unit and N simulation wheel units, N is a positive integer, the inner bottom of the frame unit is used for bearing an ice coating and snow coating simulation experiment asphalt pavement, and the rotating and loading unit is installed at the top of the frame unit. The execution end of the rotating and loading unit extends into the frame unit and is connected with the N simulation wheel units located in the frame unit, and the rotating and loading unit is used for synchronously applying working loads to the N simulation wheel units and synchronously driving the multiple simulation wheel units to move on the experimental road surface in the same direction. The system is mainly used as a loading system in a simulation vehicle road surface rolling experiment.

The invention relates to a continuous dynamic control method for the steel feeding time sequence of a rolling mill, and the method comprises a dynamic steel feeding function module which is used for analyzing the pass number and the rolling speed in a target steel grade rolling table to generate a continuously adjustable steel feeding signal; the steel feeding signal is used for controlling the time from a target steel billet to a rolling mill; and the steel feeding time input module is used for setting a steel feeding signal to control the time from the target steel billet to the rolling mill. The method realizes continuous dynamic regulation and control of steel supply time, effectively reduces steel billet waiting time and rolling mill idling loss, and greatly improves rolling rhythm and production benefits.

The invention discloses a three-dimensional boxing method, device and equipment based on a large language model and a storage medium, and relates to the technical field of logistics automation, and the method comprises the steps: constructing an environment model of an online three-dimensional boxing problem, and defining a spatial constraint condition and an optimization target; inputting the current state data, the spatial constraint condition and the optimization target of the environment model into a large language model to generate a boxing strategy code; performing grammar verification and logic verification on the boxing strategy code to obtain a verification result; and iteratively correcting the boxing strategy code according to the verification result, outputting a target boxing scheme, and carrying out boxing according to the target boxing scheme. Dependence on high computing power is avoided by introducing a large language model. And through a grammar and logic dual detection mechanism, the generation scheme is ensured to accord with physical reality. Through iterative optimization of the large language model, continuous dynamic evolution of strategy codes is realized, and the adaptivity and the space utilization rate in a complex scene are remarkably improved.

The application provides a hydropower station tailwater level management method and device, electronic equipment and storage medium, relates to the technical field of water conservancy and hydropower engineering operation optimization and design, and the method comprises the following steps: acquiring multi-source data of a hydropower station; taking ecological data of the hydropower station in the multi-source data as a water power constraint condition or a boundary parameter, inputting the ecological data into a tailwater channel water power characteristic dynamic model, and calculating tailwater level data; constructing a constraint condition, and inputting a tailwater channel management scheme in the multi-source data into a multi-objective collaborative optimization decision engine; wherein the multi-objective collaborative optimization decision engine adopts a double-layer nested optimization architecture, takes discrete tailwater channel management schemes and continuous dynamic tailwater level control sequences as coupled decision variables, takes maximum comprehensive benefit as a target function, iteratively optimizes under the constraint condition, and generates a tailwater level management scheme. The application realizes multi-element collaborative iterative optimization, and solves the problems of single data, non-combination of ecological models and fragmented optimization elements in the prior art.

The invention discloses a dynamic auditing system and method for a digital twin intelligent manufacturing quality system, and the system comprises a physical layer which comprises a production line, manufacturing equipment on the production line, and a sensor network disposed on the manufacturing equipment, and is used for collecting the multi-source real-time data of a manufacturing process; according to the method, the crossing from'periodic static auditing 'to'continuous dynamic auditing' is realized, the deviation in the system operation can be found in time, the sampling risk and human subjective factors are avoided, the auditing conclusion is more objective and comprehensive, the quality risk can be predicted, the early warning is given out before the problem occurs, the conversion from'post correction 'to'pre-prevention' is realized, and the safety of the system is improved. By using a digital thread technology, the problem source can be quickly traced, the problem analysis and decision-making time is greatly shortened, the workload of manual auditing is greatly reduced, the auditing efficiency and precision are improved, and the quality cost is reduced.

This invention relates to the field of dynamic compression-shear test control technology, and discloses a damping adaptive control method and system for a dynamic compression-shear testing machine. The method first identifies the steady-state damping parameter set of the current working condition online to predict the transient damping requirement when the working condition changes, generates a transient damping compensation feedforward control sequence and injects it into the servo valve control loop, and then corrects the damping feedback gain parameter in real time based on the transient damping residual to complete the closed-loop adaptive control. This effectively reduces the transient loading error when the working condition changes, shortens the transient process time, and improves the efficiency and test accuracy of multi-working-condition continuous dynamic compression-shear tests.

The present invention provides a novel explicit analytical mathematical modeling method for hybrid automata systems. First, the switching mode of a continuous-time switched system model with differential form is transformed into a constant switching signal that depends on an index at the switching moment. Then, the discrete state nodes and discrete input events in the logical evolution graph of a deterministic finite automaton are represented by their own logical vector forms respectively. In addition, by using the semi-tensor product tool of matrix calculation, an algebraic characterization form of the dynamics of the switching signal based on the automaton is given, that is, a restricted automaton-dependent switching mechanism based on the evolution mode of a deterministic finite automaton is designed. Finally, by organically integrating a continuous switched system and a discrete deterministic finite automaton, an explicit mathematical model of a hybrid automaton with a hybrid form is given. The model constructed by the present invention can directly represent and describe the hybrid characteristic problems related to discrete event-driven and continuous dynamic processes in the fields of computers and control.

This application discloses a method, apparatus, device, and medium for dynamic prediction of nearshore microplastic aggregation ridges, relating to the field of data processing technology. The method includes: determining a mass point library; reading the mainstream orientation of each mass point in the library; assembling each mass point as a node to obtain a two-layer heterogeneous graph, and writing the shear rate and diffusion weight calculated based on the mainstream orientation into the edge attributes of the upper and lower heterogeneous graphs respectively; splitting the two heterogeneous graphs into a directed convergent adjacency matrix and an undirected diffusion adjacency matrix respectively, and then extracting node representations; selecting ridge seed based on the node representations; calculating the optimal ridge based on the ridge seed; constructing driving vectors for the optimal ridges from multiple historical frames, and generating a predicted ridge sequence with a preset future duration based on the driving vectors and the geometric coordinates of the ridge points in the optimal ridge. This application ensures that each mass point simultaneously possesses continuous dynamic and material properties at any given time, and improves the prediction accuracy of microplastic aggregation based on accurate centroid alignment analysis.

This invention discloses a continuous temporal interaction graph representation method and medium. The method includes: acquiring raw dynamic graph data and preprocessing it to obtain structural data, node features, and edge features; partitioning the data to generate a training set, a test set, and a validation set; training a model based on the training set to obtain an initial graph representation model, and testing and validating the initial graph representation model based on the test set and validation set to obtain a final graph representation model; acquiring the dynamic graph to be processed; and outputting the node representation vectors corresponding to the nodes in the dynamic graph to be processed through the final graph representation model. This method can effectively capture the continuous dynamics of node representations and learn their state trajectories, thereby effectively improving the representation ability of inactive nodes and the prediction effect of long-interval interactions.