27results about "Digital differential analysers" patented technology

A system and method are used to maintain a variance in feedback factors of an amplifier between the first and second phases either below a threshold value or within a specified range. The system includes the amplifier and first through third capacitances. The amplifier is coupled between an input node and an output node that operates during first and second phases of operation. The first capacitance is coupled across the amplifier and between the input node and the output node during the first and second phases of operation. The second capacitance is coupled to the input node during the first phase of operation. The third capacitance is coupled to one of the input and output nodes during one or both of the first and second phases of operation.

The invention relates to the technical field of new energy power generation, and discloses a generating capacity calculation method and system based on wind and light resources and equipment data. The method comprises the following steps: acquiring and fusing wind and light resource data and equipment performance parameters to construct a standardized data set; based on the data set, a resource-equipment dynamic coupling potential field is established, and the potential field accurately represents the comprehensive power generation potential in the space-time dimension by fusing the resource space gradient and the equipment response characteristics; carrying out triple integral calculation on the potential field, and controlling an error by adopting a self-adaptive algorithm; and finally, realizing self-calibration of the system through model verification and parameter optimization. According to the method, the defects of insufficient data fusion, model staticization, contradiction between calculation precision and efficiency and the like of a traditional method are overcome, and the accuracy, the adaptive capacity and the engineering practical value of generating capacity calculation are remarkably improved.

The invention belongs to the technical field of synchronous control, and particularly relates to a synchronous control system and method for vibration suppression and quality monitoring of twisted pairs, and the method comprises the following steps: S1, obtaining the measurement data of an acceleration sensor and a laser displacement sensor array on a twisted pair production line, a real-time vibration vector field of the twisted pair in the monitoring area is constructed through data fusion; based on the real-time vibration vector field, utilizing a Hamiltonian agent model to predict the vibration energy state of the twisted pair in a future prediction time window; and S2, adjusting a weight coefficient for minimizing the weighted sum of the predicted vibration energy and the control increment norm in real time according to the twisting quality characteristic index calculated in the previous control period. According to the method, conversion from passive response to active predictive suppression is realized, optimal control measures can be taken before vibration deterioration, and the timeliness of vibration suppression is improved.

The present disclosure relates to the technical field of automatic differentiation, including an automatic differentiation method of parallelized reverse mode, a storage medium and a product. In an automatic differentiation conversion process on a source program, a tensor tape corresponding to each reading position of a declared intermediate variable is generated to generate a forward propagation program and a backward propagation program based on the tensor tape; if the reading position belongs to a first loop structure in the source program, the shape of the tensor tape is determined based on the loop boundary of the first loop structure, and the stored intermediate value is indexed by the loop variable of the first loop structure; this makes the data retrieval in the backward propagation process no longer dependent on the execution order of the control flow of the forward propagation of the program, and on this basis, a second loop structure in the backward propagation program can be subjected to loop transformation, and then a target program executable by a processor is generated; the problem of parallel conflict naturally introduced by automatic differentiation is solved by rewriting the backward propagation program, parallel resources are fully utilized, and parallel efficiency is improved.

The application discloses a physical enhancement-based fluid dynamics small sample solving method, including: acquiring sparse fluid observation data in a complex geometric calculation domain, and mapping the observation data to a statistical average flow space described by a Reynolds-averaged Navier-Stokes equation; constructing a neural network model for predicting an average velocity field, an average pressure field and a spatially continuous and learnable eddy viscosity coefficient latent field; when calculating spatial derivatives in the Reynolds-averaged Navier-Stokes equation, a differentiable spectral operator module is used to replace a traditional automatic differentiation mechanism; based on geometric distance information of each spatial point in the calculation domain to a solid boundary, a geometry-aware mask is constructed; the mask is used to distinguish the calculation domain into a near-wall shear dominated region and a main flow region; a joint loss function containing a data term and a physical residual term is constructed, and the physical residual term is preconditioned; and a prediction result is output; the application realizes numerical stability and high prediction accuracy.

The invention relates to the technical field of hydrological sediment monitoring, and discloses a sediment content calculation method, which comprises the following steps: acquiring a single-point flow velocity vector, a single-point original sediment content and a total water depth of a measuring point; constructing a non-constant inertia index combining the flow velocity change rate and the single-point flow velocity modulus; determining a dynamic lag time window by using a non-constant inertia index and acquiring a reference sand content; calculating a vertical density gradient and a gradient Richardson number by using the reference sand content, and determining a turbulence suppression factor; respectively correcting a vertical flow velocity distribution function and a sand content distribution function by using a turbulence suppression factor and a non-constant inertia index, and constructing a vertical flow velocity distribution data column and a sand content distribution data column through iterative calculation; and integrating to obtain the single-width sediment transport modulus. According to the method, sediment response lag is corrected by introducing a non-constant inertia index and a dynamic lag time window, the damping effect of density stratification on turbulent flow is quantified by using a turbulent flow suppression factor, and the full-water deep-water sediment structure under the non-constant flow condition is accurately inverted by using single-point data.

The invention provides a method for calculating sound and vibration characteristics of a multi-section rotary shell under light fluid, which comprises the following steps of: splitting the multi-section rotary shell into a conical section with a linear generatrix and a hyperbolic section with a nonlinear generatrix according to curvature change characteristics of a meridian line of the shell, and generating node distribution meeting differential quadrature operation in each section; a control equation of the shell is established based on the first-order shear deformation theory, boundary conditions are simulated through an artificial spring, force and displacement continuity of multiple sections of joints is processed, free vibration and forced vibration characteristics are solved through a differential quadrature algorithm, and global speed response is obtained; a fitting unit radiation superposition method is adopted to enable the surface of a shell to be equivalent to an acoustic baffle, a cylindrical surface or a spherical crown surface is adopted to fit and calculate a sound transmission vector according to the type of the shell, and far-field sound radiation characteristics under light fluid are solved in combination with global speed response. The sound and vibration characteristic analysis method provided by the invention has high efficiency, universality and good engineering practicability on the premise of ensuring the calculation precision.

The invention discloses a switch hand feeling multi-scale modeling method based on a physical information neural network, and the method comprises the steps: firstly building a one-dimensional nonlinear elastic damping partial differential equation PDE and an ordinary differential equation ODE which describe the overall deformation behavior of a key, and forming a hybrid PDE-ODE kinetic model; secondly, according to the dynamic model, designing a Bayesian physical information neural network B-PINN with an uncertainty quantification capability; and then the data sampling density is dynamically improved in a mutation region, training weight distribution is adjusted by adopting an event-driven strategy, and the adaptive training capability of the neural network B-PINN is enhanced. And finally, generating an elastic modulus and damping coefficient probability distribution map by using the trained network model, mapping the subjective hand feeling score of the user into a mechanical objective function, and inverting parameters. According to the method, the defects of poor generalization and insufficient physical consistency of a pure data driving method are overcome, and the reliability of switch design parameters is improved.

The application discloses a gravity data downward continuation method based on a discrete model of a reverse band-limit Poisson integral, equipment, a medium and a product, relates to the technical field of gravity data processing, and performs gridization on discrete gravity point data, determines a grid limit frequency, a truncation order of grid gravity anomaly data and a type of a discrete model of a reverse band-limit Poisson integral, and when a direct discrete model in a removal recovery mode is removed, determines a coefficient matrix, a parameter vector and a constant vector based on the grid gravity anomaly data, a continuation height and an earth gravity field model; determines a target calculation formula based on the coefficient matrix, the parameter vector and the constant vector; when an indirect discrete model in a near-zone integral mode is removed, determines a target calculation formula of the discrete model based on the grid gravity anomaly data and the continuation height; and obtains a downward continuation result based on the target calculation formula of the discrete model. The application can stably and highly accurately perform downward continuation on gravity data.

The invention discloses a space-time fence early warning method and device based on a vehicle position, a vehicle and a medium, and relates to the technical field of vehicles. The method comprises the following steps: acquiring the current position of a vehicle and associated space-time fence information; when it is determined that the space-time fence where the vehicle is located currently is not switched according to the current position of the vehicle and the associated space-time fence information, the current activity duration of the vehicle is accumulated; and controlling the vehicle to give an alarm under the condition that the current activity duration of the vehicle is greater than or equal to a preset activity duration threshold value corresponding to the space-time fence where the vehicle is currently located. Thus, the problems that an existing electronic fence system cannot continuously track the duration of the vehicle in the fence, the dynamic space-time rule is difficult to support, and a rule engine is rigid are effectively solved, precise monitoring and self-adaptive early warning of the continuous state of the vehicle behavior are achieved, and the real-time performance, flexibility and service adaptability of vehicle supervision are improved.

The embodiment of the application discloses a high-pass filter electromagnetic transient modeling method and device and electronic equipment, which can quickly, conveniently and accurately perform electromagnetic transient modeling on the high-pass filter, and is beneficial to electromagnetic transient simulation of a power system. The method comprises the following steps: firstly, determining a frequency domain model of the high-pass filter according to the high-pass filter; secondly, obtaining a time-domain integral model of the high-pass filter in a time period according to Laplace inverse transformation and the frequency domain model of the high-pass filter; and finally, obtaining a segmented time-domain electromagnetic transient model of the high-pass filter according to the trapezoidal rule and the time-domain integral model of the high-pass filter in the time period.

The invention provides a vibration system equivalent dynamic modeling method based on a steel wire rope catenary effect, and belongs to the technical field of dynamic simulation, and the method comprises the steps: discretizing a steel wire rope into a plurality of mass points distributed in the length direction of the steel wire rope, and connecting the adjacent mass points through spring damping force elements, constructing a dynamic model of the vibration system; performing a virtual test based on the dynamic model to obtain equivalent stiffness and equivalent mass of the vibration system at different travel positions of the trailer wagon; according to the relation between the equivalent stiffness and the travel of the trailer wagon and the relation between the equivalent mass and the travel of the trailer wagon, an equivalent model of the vibration system is established, and the equivalent model is a single-degree-of-freedom vibration system with the equivalent stiffness and the equivalent mass changing along with the travel. According to the invention, the efficiency of dynamics simulation control and the accuracy of simulation analysis are improved.

This invention relates to the field of magnetic pole topology design, and more particularly to a method for designing the magnetic pole topology of a finite-angle torque electromagnetic drive. The method includes: meshing the rotor back iron, air gap, and stator domain of a finite-angle torque motor; defining permanent magnet material variables and soft magnetic material variables; based on the material variables, performing filtering, projection, and constitutive equation electromagnetic field mapping through an electromagnetic field finite element model to generate mapped material variables and torque variables; constructing constraints and an objective function; and solving the magnetic field strength characterization variables of the objective function using an optimization solver to determine the optimal combination of the corresponding permanent magnet material variables and soft magnetic material variables. This invention transforms the distribution problem of permanent magnets and soft magnetic materials in the rotor back iron into a topology optimization problem, achieving a high-performance combination of permanent and soft magnetic materials, improving the material utilization rate and torque density of both permanent and soft magnets, and realizing high material utilization of the rotor back iron of a finite-angle torque motor.

The application discloses a kind of fine modeling methods for single-shaft gas turbine gas path system, including establishing single-shaft gas turbine gas path system's step-by-step compression compressor module, establishing single-shaft gas turbine gas path system's combustion chamber module, establishing single-shaft gas turbine gas path system's step-by-step expansion turbine module, based on the calculation method of compressor module outlet temperature, power consumption, the calculation method of each component mass, temperature of combustion chamber module outlet and the calculation method of turbine module outlet temperature, power, using Python simulation platform establishes single-shaft gas turbine gas path system model, obtains rated operating condition and actual operation data, the data of rated operating condition is input into the single-shaft gas turbine gas path system model built, and the dynamic characteristics of gas turbine are analyzed;It aims at solving the problem of insufficient accuracy caused by neglecting the change of specific heat capacity with temperature and the evolution of combustion product composition in the existing model.

The invention discloses a double-reflector antenna electromechanical coupling rapid analysis method and system, and aims to reconstruct a physical optical integral expression. Dispersing the reflecting surface into triangular grids, and performing linear fitting on amplitude and phase functions in each grid; an equivalent wave number / wavelength concept is introduced, a phase fitting error is analyzed, a differentiated grid processing strategy is put forward, a closed expression for calculating a far-field directional diagram is deduced through affine transformation and partial integration, and numerical integration is replaced by analytical calculation; and converting the structural deformation into an additional phase error, and integrating the additional phase error into a rapid calculation framework. The method breaks through the limitation of the traditional electromagnetic analysis on the grid size, greatly improves the electromechanical coupling analysis speed on the premise of guaranteeing the precision, and provides an efficient tool for the design and profile compensation of a high-frequency-band antenna.

This invention discloses a method for synergistic optimization of soil structure and fertility based on straw return to the field to promote decomposition, belonging to the field of soil optimization technology. A systematic technical system is constructed from the dimensions of in-situ straw decomposition and compost humification enhancement. Highly efficient decomposition-promoting microorganisms and humification promoters are screened through indoor cultivation experiments. The regulatory mechanism of decomposition promoters on the compost humification process is analyzed using humification kinetic models and structural equation models, forming a compost humification enhancement technology. Key parameters affecting in-situ straw decomposition are screened using the nylon mesh bag method. Multiple physical, chemical, and biological indicators of the soil are measured through years of field experiments. A comprehensive scoring model is constructed using principal component analysis, combined with multi-factor variance analysis and multiple comparisons, to screen the treatment combination with the highest contribution rate to the synergistic improvement of soil structure and fertility. Technical procedures are formulated and demonstrated. This achieves synergistic fertilization of the topsoil and sub-topsoil of degraded soils, with significant economic, ecological, and social benefits.

The application relates to the technical field of new energy power generation, and discloses a power generation capacity calculation method and system based on wind and light resources and equipment data. The application collects and fuses wind and light resource data and equipment performance parameters to construct a standardized data set; based on the data set, a resource-equipment dynamic coupling potential field is established, the potential field accurately represents the comprehensive power generation potential in the time and space dimensions by fusing resource space gradients and equipment response characteristics; then, triple integral calculation is carried out on the potential field, and an adaptive algorithm is used to control errors; finally, model verification and parameter optimization are carried out to realize self-calibration of the system. The application overcomes the defects of the traditional method, such as insufficient data fusion, static model, contradiction between calculation accuracy and efficiency, and the like, and significantly improves the accuracy, adaptive capacity and engineering practical value of the power generation capacity calculation.

The present application relates to the technical field of centrifugal pump, in particular to a centrifugal pump blade wrap angle determination method based on fluid mechanics numerical integration. The centrifugal pump blade wrap angle determination method is a numerical integration method based on fluid mechanics equation. By combining differential equation and numerical integration, equidistant points are divided on the axial surface streamline in the axial surface projection diagram, and the arc length and radius between each adjacent two equidistant points are calculated. Based on one-dimensional flow theory, the axial surface fluid velocity v m is calculated m , and a curve changing along the axial surface streamline is drawn. According to Euler equation, a linear distribution model of tangential velocity moment v u r is established. After calculating the integral coefficient K, K·△s i is calculated point by point, and the integral is accumulated to directly obtain the centrifugal pump blade wrap angle value. The effectiveness of the centrifugal pump blade wrap angle value is verified by using an example verification method. The present application can realize the design of the centrifugal pump blade wrap angle through parameterization, and get rid of the condition that the centrifugal pump blade wrap angle can only be obtained by relying on experience.

The invention relates to the field of magnetic pole topology design, in particular to a finite angle torque electromagnetic drive magnetic pole topology design method, which comprises the following steps of: performing mesh generation on rotor back iron, an air gap and a stator domain of a finite angle torque motor, and defining a permanent magnet material variable and a soft magnetic material variable; performing filtering, projection and constitutive equation electromagnetic field mapping through an electromagnetic field finite element model based on the material variable to generate a mapping material variable and a torque variable; constructing a constraint condition and a target function; and solving the magnetic field intensity characterization variable of the target function through an optimization solver to determine an optimal combination of the corresponding permanent magnet material variable and the soft magnetic material variable. According to the method, the distribution problem of the permanent magnets and the soft magnetic materials of the rotor back iron is converted into the topological optimization problem, the high-performance permanent magnet and soft magnetic material combination is obtained, the material utilization rate and the torque density of the permanent magnets and the soft magnets are improved, and the high material utilization rate of the rotor back iron of the limited angle torque motor is achieved.

The invention provides a land hydrological load deformation determination method combining a spherical harmonic domain and a spatial domain, and belongs to the field of geodetics. The method comprises the following steps: S1, based on a gravity field model and a land hydrological model, respectively extracting a 60-order spherical harmonic coefficient and a space grid value from a spherical harmonic domain and a space domain; thirdly, calculating the vertical deformation of the regional hydrological load by using the coefficients and the load number in the step S2; and then, in the step S3, the spatial grid value is subjected to spherical harmonic expansion to 60 orders and then is converted into a grid value again, and the grid value is differentiated from an original value to obtain high-frequency mass distribution information. And S4, integrating the high-frequency load mass by adopting a Green function to obtain the contribution of the high-frequency load mass to hydrological load deformation. And finally, S5, integrating the results of the spherical harmonic domain and the spatial domain to obtain the total load deformation contribution. According to the method, the calculation precision and efficiency are improved, and the method is suitable for GNSS load correction so as to accurately extract tectonic deformation information.

The invention discloses a theoretical calculation method for a vehicle collision compatibility index OLC. The theoretical calculation method comprises the following steps: S1, acquiring time-acceleration curve data; s2, integral calculation is conducted, and standard displacement-collision force curve data, the maximum displacement and the maximum acceleration of the front end of the vehicle are obtained; s3, simulating an FWRB collision test through the law of conservation of energy, and calculating the maximum displacement and the maximum acceleration value of the vehicle in the simulated FWRB collision test; s4, comparing and checking; s5, an MPDB collision test is simulated, and the OLC value of the vehicle in the simulated MPDB collision test is calculated; the invention provides a complete calculation method for theoretically estimating MPDB test compatibility index OLC based on real vehicle FWRB collision test data.

The invention relates to a deterministic integral operation method based on an AI inference chip. The method comprises the following steps: receiving a to-be-operated characteristic quantity from an upper layer of an AI inference network; performing modular operation on the to-be-operated characteristic quantity according to a pre-configured modular constant so as to map the to-be-operated characteristic quantity into a finite field to obtain a deterministic address; disturbing the deterministic address based on the configuration parameter to obtain a disturbed address; the perturbation address is input into a table look-up memory, pre-calculation data corresponding to the perturbation address are read from the table look-up memory, the pre-calculation data are obtained through calculation based on a three-valued logic modal operator before tape-out, and an operation result is solidified and written into a storage array of the table look-up memory, so that deterministic integral table look-up access to the three-valued logic operation result is achieved. According to the method, a deterministic reasoning path is constructed through modular mapping and controllable disturbance processing in a finite field in combination with a table look-up mechanism of a three-valued logic modal operator, and the calculation delay is remarkably reduced.

This invention discloses a method, apparatus, and filter for constructing a parabolic window function. The construction method includes performing feedback integration on the input signal to obtain a first integral output signal; performing delay superposition on the first integral output signal to obtain a first adder output signal; and constructing a parabolic window function by adding the input signal after performing a first delay subtraction and the first adder output signal. This embodiment achieves efficient tracking filtering by designing a fastest tracking filter using a parabolic window function, overcoming the limitation of low tracking efficiency in the exponential tracking mechanism of first-order inertial filtering and enhancing the filtering effect.

The application provides a human motion data enhancement method and system based on a virtual physical perturbation field, and the method comprises the following steps: based on input human key point three-dimensional position data, a human physical agent model with preset physical properties is constructed; based on the kinematic state of the model and semantic information of a target motion task, a task uncertainty potential field model is used to calculate a task uncertainty potential field representing a task-related variation tendency; based on the task uncertainty potential field, a virtual disturbance force acting on the physical agent model is generated; the disturbed physical agent model is subjected to dynamic calculation to obtain a globally related linkage displacement noise; and the linkage displacement noise is added to the original key point position data to obtain enhanced human motion data.

This invention discloses a dynamic detection method for abnormal network security operation and maintenance behaviors based on artificial intelligence, relating to the fields of network security and artificial intelligence technologies. It solves the problem that existing detection methods struggle to adaptively adjust detection thresholds according to dynamic environments and business contexts, leading to high false positives and latency. This invention constructs a heterogeneous temporal behavior graph from operation and maintenance audit logs, uses a temporal heterogeneous graph attention network to encode session behavior, and generates a high-dimensional continuous behavior curve. Based on the Flyner frame, it extracts curvature and torsion changes, constructs a distortion co-variable by combining information entropy and dynamic baseline entropy, and integrates it to obtain a dynamic anomaly index. It generates adaptive thresholds by linking the environment context through a neural frequent differential equation, and continuously optimizes the model using alarm verification feedback. This invention achieves multi-dimensional dynamic analysis and adaptive anomaly judgment of complex operation and maintenance behaviors, effectively reducing false positive rates and improving detection real-time performance.

This invention discloses a method, system, and product for estimating the arm length of a Taiji mission considering satellite relative motion compensation. The method includes: reading ephemeris data of key gravitational bodies during the simulation period, establishing the orbital dynamics equations of the satellite formation, solving the equations using a high-precision numerical integration algorithm, and outputting satellite orbit simulation data; generating basic measurements including noise and satellite-to-solar measurements based on the satellite orbit simulation data, and correcting the inter-satellite distance measurements in the basic measurements by optical signal propagation time to achieve satellite relative motion compensation and outputting high-precision observation data; receiving the high-precision observation data, combining it with the orbital dynamics equations of the satellite formation, recursively updating and optimizing the state variables of the satellite formation based on a Kalman filter algorithm, and outputting a high-precision estimate of the inter-satellite arm length; and visualizing the estimated satellite arm length and its error. This invention significantly improves the autonomy and reliability of space gravitational wave detection missions.

The invention provides a parallel computing method and device for Hermite Gaussian integrals in electronic integrals, and the method comprises the steps: determining all target Hermite Gaussian integral items which need to be computed and intermediate integral items which the target Hermite Gaussian integral items depend on, and building a dependency graph; dynamically distributing the target Hermite Gaussian integral item and the intermediate integral item on which the target Hermite Gaussian integral item depends to a plurality of calculation units, so as to enable the calculation load of each calculation unit to meet a balance condition; and configuring a corresponding calculation instruction sequence for each calculation unit, generating the calculation instruction sequence according to the distributed integral item and the dependency graph, and configuring the calculation instruction sequence as a loop-free operation sequence directly executed by the calculation units so as to control the plurality of calculation units to execute the calculation instruction sequences configured by the calculation units in parallel, and completing the calculation of all Hermite Gaussian integral items. The method is used for solving the problems that the hardware utilization rate is low due to instruction level parallelism difficulty of a traditional recursion method, and a new performance bottleneck is formed under the condition of low angular momentum due to excessive FMA instructions in a replaced fast Fourier transform method.