159 results about "Simulation error" patented technology

A system and method for testing robustness of a simulation model of a cyber-physical system includes computing a set of symbolic simulation traces for a simulation model for a continuous time system stored in memory, based on a discrete time simulation of given test inputs stored in memory. Simulation errors are accounted for due to at least one of numerical instabilities and numeric computations. The set of symbolic simulation traces are validated with respect to validation properties in the simulation model. Portions of the simulation model description are identified that are sources of the simulation errors.

A channel driver circuit includes a differential module and a driver module. In some examples, the channel driver circuit also includes a sigma-delta module. The differential module receives, via a single node of a load, a channel driving signal that is provided to the load at the single node (e.g., that is based on an electrical characteristic of the load) and generates an analog error signal that is based on the channel driving signal and a reference signal. The driver module is coupled to the differential module and generates the channel driving signal based on the analog error signal or a digital error signal corresponding to the analog error signal and transmits the channel driving signal via the single node to the load. The channel driver circuit simultaneously transmits the channel driving signal to the load at the single node and senses the channel driving signal at the single node.

The invention relates to a baseband channel simulation system of a mobile satellite suitable for a Ka frequency band, which comprises a real-time channel control device and a real-time channel simulation device; the real-time channel control device comprises a simulation parameter setting module, an error code characteristic simulation module and an error code threshold computation module; the real-time channel simulation device comprises a channel characteristic generation module and an error code superposition module; and the error code type and the error code threshold of a link scene set by a user can be generated by the real-time channel control device, a link simulation error-code pattern can be generated by the real-time channel simulation device according to the error code type and the error code threshold, and the simulation of signal transmission under the set link scene is finished. According to the baseband channel simulation system of the mobile satellite suitable for the Ka frequency band, the simulation of a channel of a satellite ground station and a link between satellites in systems, such as a communication satellite, a relay satellite and the like, can be truly carried out with lower cost under the environment of a laboratory. The baseband channel simulation system of the mobile satellite suitable for the Ka frequency band is especially suitable for verifying the practicability and the effectiveness of a satellite communication technology with a Ka frequency band.

The invention relates to a structure deformation fiber monitoring-reconstructing method and system. Strain and deformation data is extracted according to a finite element simulation result of a structure, a strain-deformation function relation of multiple lines is fit on the basis of the idea of least squares, and an initial integral deformation field of the structure is obtained by interpolation.A fiber monitoring system is established, strain, temperature and acceleration data of the structure is collected, a strain sequence is reconstructed via an interpolation algorithm according to discrete point data measured by experiments, the strain-deformation function relation is corrected, and further the integral deformation field of the structure is corrected. The system and method are verified in a rectangular plate deformation reconstruction experiment, the deformation state of a tested piece can be obtained effectively via the system and method, the deformation field is reconstructedfor about 2%, and a simulation error is reduced greatly compared with a finite element.

The invention discloses a method for constructing a graded crushed rock micro-mechanical model and calibrating a micro-mechanical parameter. The graded crushed rock micro-mechanical model is constructed through a test on a graded crushed rock basic parameter, simulation of a test mold and generation of the simulation test piece by endowing the micro-mechanical parameter to a simulation test piece; a micro-mechanical model test piece is subjected to a value experiment, so that a graded crushed rock stress strain curve is obtained through stimulation; and the micro-mechanical parameter of the graded crushed rock micro-mechanical model can be calibrated by analyzing the comparison between an actual test result and a simulation structure of graded crushed rock stress strain according to a simulation error precision requirement. By the method, the graded crushed rock micro-mechanical model can be constructed accurately and quickly and the micro-mechanical parameter can be calibrated, so that an effective tool is provided for deep research on a mechanical behavior and a damage mechanism of graded crushed rocks.

The invention discloses a comprehensive operation state evaluation simulation system of a capacitor voltage transformer. The method comprises steps that: by establishing a CVT circuit model and a corresponding relation between an insulation parameter and each environmental parameter, the simulation error of the CVT under different environmental parameters is acquired, error data is recorded and stored, and is compared with the data measured by a on-site CVT error on-line monitoring device, and then a CVT error state is evaluated. According to the invention, influences of internal and externalinsulation factors on a CVT metering state in the actual operation of the CVT are considered, the comprehensive operation state evaluation simulation model of the capacitor voltage transformer is established, and influence rules of factors such as environmental temperature, secondary load, power supply frequency, external pollution and the like on the CVT metering performance can be acquired through simulation analysis, and a theoretical and practical basis for improving the operation stability of the CVT can be provided.

The invention provides a method for judging a remarkable influence period of human activities on river flow. The method includes collecting and studying long-sequence monthly hydrometeorological data in a river basin; preliminarily judging years of change in remarkable influence on yearly flow due to the human activities; screening a hydrologic model suitable for the studied river basin; simulating a full-sequence flow process; counting simulation error standard quantity SREi of each year on yearly flow; drawing a yearly process of the simulation error standard quantity; judging the remarkable influence period of the human activities on the river flow according to the circumstance that the yearly process deviates from a horizontal axis. On the basis of mathematic statistics and hydrologic process physical simulation methods, the influence period of the human activities on the river flow is diagnosed, possible influence of climate factors on the river flow is scientifically abandoned, a diagnosis result is uniform with actual conditions of the human activities in the river basin, and misjudgment on the remarkable influence period of the human activities due to influence, of the climate elements, on a hydrologic sequence when a conventional method is adopted is avoided effectively, so that the given diagnosis result is more scientific and reasonable.

The invention discloses a fire evolution simulation method. The method comprises the following steps: on the basis of a lattice Boltzmann method of a mesoscopic method, establishing a two-dimensionalfire simulation model in view of fire characteristics in subway stations in combination with related theoretical methods, adding a temperature distribution function evolution equation to a basic modelof the lattice Boltzmann method to construct a dual-distribution function thermal lattice method model so as to simulate the heat transfer in the fluid; performing turbulent flow simulation by usinga large eddy simulation method, importing a sub-lattice model, reconstructing a relaxation time in the model by using physical viscosity and eddy viscosity, and constructing a lattice Boltzmann sub-lattice model; and importing a boundary processing format of obstacles in the fluid to simulate the streaming of the fluid around the obstacles; and performing solution and calculation by using a vortex restriction method to reproduce a smoke concentration diffusion process. By adoption of the method, the fire evolution simulation is optimized, the simulation validity is improved, and the simulation error is reduced. The invenetion further discloses a fire evacuation comprehensive simulation method comprising the fire evolution simulation method.

In a client-server game system, or in accordance with a computer-readable program stored on a non-transitory computer-readable medium, a client participates in a game with other clients by performing an in-game simulation. Based on results of the in-game simulation, the client sends error-checking data to a server. The server analyzes the error-checking data to determine whether a simulation error has been caused by at least one of the clients. Upon determining that the simulation error has been caused, the server causes a reference simulation to be performed. Based on the results of the reference simulation, the server identifies the client(s) that caused the simulation error. The server then sends a patch to the client(s) that caused the simulation error. The client applies the patch to correct and reset the in-game simulation.

A method and apparatus for joint training of an analog equalizer (AEQ) and an analog echo canceller (AEC), which both process an input analog signal in the analog domain. In one embodiment, the method includes joint training the AEQ and the AEC using independent analog error signals.

The invention relates to the field of modeling and simulation, and provides an infrared imaging equipment reconnaissance effect simulation method suitable for combat simulation. The method includes such steps as simulating the target searching process, judging whether the observed field of view of infrared imaging device effectively covers the detected target, and detecting the detected target with the field of view of infrared imaging device; secondly, simulating the process of detecting and recognizing suspicious targets, aiming at judging the effect of target recognition based on the mathematic model constructed by the main technical performance of infrared imaging equipment, battlefield environment conditions and target characteristic information; finally, the simulation of target location process being carried out to generate the target location information with reconnaissance error. The invention divides the whole reconnaissance process into several stages such as searching, detecting, identifying, confirming and positioning. The model calculation of each stage mainly depends on the equipment tactical performance index, which can not only better reflect the inherent attributeof the equipment, but also is not complicated. Compared with the test data, the simulation error of the method is small, and the method has a good application prospect in the combat simulation.

The invention discloses a fault-tolerant method based on a high-performance simulation cloud. The method includes the steps that the fault-tolerant type is determined by the characteristics of a computer according to simulation; all simulation members are added into a simulation system, and meanwhile a shared data area is established to record the simulation process; whether all the simulation members operate normally or not is judged through heartbeat detection, if operation is normal, the shared data area is used for recording and storing current data, and if operation is abnormal, the entire simulation system is returned to a previous recorded detection point; the fault-tolerant type of the abnormal condition is judged, a simulation model is redeployed, and after the abnormal simulation members are subjected to off-line error correction independently, the abnormal simulation members and other normal simulation members recover the simulation operation together. By the adoption of the fault-tolerant method, the state before errors can be recovered quickly through self-stored interaction data, operation of the simulation system is recovered again, reloading operation is not needed by the simulation system, and the simulation error probability in the process of operation again is avoided.

The invention discloses a broadband reverberation waveform simulation method based on the normal mode theory. The method comprises the following steps that reverberation scattering elements are randomly distributed in space, a scattering echo range and phase modulation signals are set, simulated broadband signals are subjected to sub-band division, ocean reverberation time sequences of sub-bands are calculated respectively, the ocean reverberation time sequences are superposed to obtain an ocean broadband reverberation time sequence at a certain receiving array element, and the method is used for obtaining ocean broadband reverberation time sequences at different receiving array elements through calculation. The method reflects influences of a channel on different frequencies in a band width, simulation errors are small, various factors in reverberation simulation are comprehensively considered, reverberation spatial correlation, space-time characteristic and other physical characteristics can be naturally reflected, and requirements for reverberation data of space-time self-adaptive processing and other new reverberation restraint algorithms can be met.

The invention discloses an energy interband non-local quantum tunneling simulation method with a current conservation characteristic. The method comprises the steps of generating a linear equation system which adopts a nodal value increment of a solved physical variable as a variable in a dispersing semiconductor device physical region; processing non-local quantum tunneling, determining a public energy zone for defining a first side and a second side; carrying out weight distribution on point-to-point tunneling current density and energy integral volume produced by interpolation; dividing a coefficient matrix of the linear equation system into a principal matrix caused by general differential equation dispersing and an additional coefficient matrix caused by non-local quantum tunneling with the current conservation characteristic, and storing; adopting a Gaussian origin eliminating method to solve the coefficient matrix. According to the energy interband non-local quantum tunneling simulation method with the current conservation characteristic provided by the invention, the non-local quantum tunneling current conservativeness is ensured, and the simulation error and the uncertainty brought by non-conservation are eliminated; by utilizing a method combining the principal matrix and an auxiliary incidence matrix, the coefficient linear equation system obtained through linearizing a non-linear equation system can be quickly solved, and the solving efficiency is improved.

The invention discloses an engineering wind speed short-time prediction method and system based on physical information coupling. The method includes: simulating a large air flow field of the target area by adopting a CFD method so as to calculate response time and simulation errors of each grid division mode of each sub-area for wind speed simulation and a plurality of simulation wind speed sequences of each internal wind measurement device, and screening dominant grid division modes according to the simulation errors; for each internal wind measurement device, obtaining a wind speed conversion model by offline deep learning based on the simulated wind speed sequence obtained by the fusion CFD method; in addition, obtaining a wind speed prediction model based on wind speed test sample offline deep learning; and finally, intelligently matching the output values of the wind speed conversion model and the wind speed prediction model through the spatial relationship between the target prediction place and the nearest internal wind measurement device, and outputting an optimal wind speed prediction value. According to the method, the calculation and storage cost of offline model training is reduced, and the wind speed prediction precision is improved while the real-time performance and the generalization capability of wind speed prediction are ensured.

The invention discloses a chip image super-resolution reconstruction method based on depth learning, which mainly solves the problem of low resolution at the dense part of the circuit of the chip image reconstruction in the prior method. The technical scheme is as follows: 1. dividing the image set and constructing the training data set; 2. training the training data set; 3. estimating the sub-pixel displacement of K low-resolution images and reference images; 4. carrying out up-sampling on that reference image, inputting the reference image into the train model, and outputting the estimated image; 5. degrading the estimated image, and calculating the simulation error between the degrade image and the K low-resolution images; 6. superimposing the simulation error on the estimated image toobtain an improved estimate image; 7. Iteration being performed on steps 5 to 6 until that error function is less than the error threshold, and the resulting improved estimated image being output. Theinvention improves the super-resolution reconstruction effect of the circuit dense part in the chip image, and can be used for detecting the hardware Trojan horse at the circuit dense part of the chip.

The invention relates to a parallel computing network division method capable of absorbing interface delays of different step lengths. The method comprises the following steps: step 1: determining a transmission line port equation of a Thevenin equivalent circuit incorporating into a sub-network; step 2: performing different step length simulation of transmission line decoupling; step 3: performing preset simulation step length discretization on a sub-network I and a sub-network II; and step 4: simulating the discretized simulation step length. The method can realize complete parallel simulation and is an ideal simulation method in real-time simulation. A transmission line suitable for network division in a network needs to be found by applying a different step length simulation algorithm based on the transmission line and is transformed into a transmission line model of a big step length delay, and the network is decoupled through the model. A computing error caused by no interface delays during electromagnetic transient parallel network division of different step lengths can be realized and numerical oscillations and simulation errors caused by the interface delays can be reduced or avoided.

A digital signal processing based feed forward amplifier includes a delay circuit which receives a portion of the input signal and generates a delayed input signal. A first analog to digital converter receives another portion of the input signal and generates a digitized input signal. A main amplifier receives the delayed input signal and generates a main amplified signal. A second analog to digital converter receives at least a portion of the main amplified signal and converts it to a digitized main amplified signal. A digital signal processor receives the digitized input signal and the digitized main amplified signal and generates a digitized error signal. The digitized error signal is converted to an analog error signal by a digital to analog converter. An error amplifier receives the analog error signal and amplifies it to generate an amplified error signal. This error signal includes a noise and distortion component which is substantially the same in magnitude as the noise and distortion component of the main amplified signal. The amplified error signal and the main amplified signal are combined so that the noise and distortion components of each subtract, leaving substantially only the amplified desired signal component remaining in the amplified input signal, which is outputted by the feed forward amplifier. A third analog to digital converter converts a portion of the amplified error signal to a digitized amplified error signal, and a fourth analog to digital converter converts a portion of the output signal to a digitized output signal. The digitized output signal and the digitized amplified error signal are provided to the digital signal processor which generates the digitized error signal in response thereto.

A method and apparatus in a DC-feed controller for controlling a subscriber line interface circuit controlling voltage (alternately, current) on a telephone line comprising a pair of wires. A digital feedback signal is subtracted from a predetermined voltage (alternately, current) limit digital value to provide a digital error signal that represents the difference between the voltage (alternately, current) limit digital value and the digital feedback signal. The digital error signal is multiplied by a predetermined digital gain value to provide a scaled digital error value. The scaled digital error value is integrated over time to generate an integrated digital error signal having a first predetermined number n of significant bits. A modified digital error signal is received, comprising a second predetermined number m of bits, where m<n, of the integrated digital error signal as an input and an output an analog error signal is provided that is representative of the modified digital error signal and that comprises the control voltage. A sense signal is received, the sense signal representing a sensed voltage (alternately, current) on the telephone line, and is converted to a digital sense signal that is a representation of the sense signal. Finally, the digital sense signal is digitally low pass filtered to provide the digital feedback signal.

The invention discloses a one-dimensional and three-dimensional joint simulation method for an automobile engine cooling system. The method comprises the steps of using three-dimensional simulation software for estimating the thermal power of a radiator, an intercooler and a condenser, taking constant values as the thermal power of the condenser, and then obtaining the outlet air volume of a radiator core, the outlet air volume of the intercooler core and the surface temperature average value of a core inlet; performing one-dimensional system simulation calculation by utilizing one-dimensionalsimulation software based on the air volume and the temperature average value to obtain the thermal power of the radiator and the intercooler; obtaining outlet air volumes of a radiator core, an intercooler core and a condenser core and a surface temperature average value of a core inlet by using three-dimensional simulation software according to the thermal power of the radiator, the intercoolerand the condenser; and repeatedly executing the steps to perform iterative computation until convergence conditions are met. The advantages of the one-dimensional simulation software and the advantages of the three-dimensional simulation software are fully utilized, simulation errors caused by heat reflux are reduced through repeated iteration, and the simulation precision of the water temperature of the cooling system is improved.

The invention provides an underwater acoustic target recognition method for optimizing a BP neural network based on a genetic algorithm. The method comprises the steps: extracting MFCC underwater acoustic features of an underwater acoustic signal, and enabling the underwater acoustic features to serve as a sample of a BP neural network classifier; randomly initializing a population of a genetic algorithm; wherein the individuals in the population comprise weights and thresholds of the BP neural network needing to be optimized, the weights and the thresholds serve as target optimization functions of the genetic algorithm, the target optimization functions are substituted into the BP neural network to obtain simulation errors, the simulation error of each individual in the population is optimized, and the optimal initial weight and threshold of the BP neural network are obtained; and finally, training and identifying the underwater acoustic features by using the BP neural network optimized by the genetic algorithm to obtain the accuracy of the classification algorithm. The method uses the genetic algorithm to identify and classify the characteristics of the underwater acoustic target, uses the BP neural network optimized by the genetic algorithm to train and predict the underwater acoustic target to obtain a higher recognition rate, and the recognition performance is more stable.

The invention discloses a CFD and deep learning-mixed extreme strong wind speed prediction method and system. The method comprises the steps of simulating a large airflow field of a target area basedon the wind speed test sample and the wind direction test sample so as to calculate the response time and the simulation errors of each grid division mode of each sub-area and a plurality of simulation wind speed sequences of each internal wind measurement device; training a wind speed conversion model based on the corresponding simulated wind speed sequence and the wind speed test sample, and training a wind speed prediction model based on the wind speed test sample, so that when the target prediction place is predicted in real time, the output values of the wind speed conversion model and the wind speed prediction model are intelligently matched according to the distance relation between the target prediction place and the nearest internal wind measurement device, and the optimal wind speed prediction value of the target prediction place is outputted. According to the method, the deep learning model is trained offline, the prediction instantaneity is improved, meanwhile, by fusing aCFD method, the large airflow field of the target site is simulated, the optimum deep learning model is matched, and the generalization ability of the prediction system is improved.

The invention provides an AEP memory Error Detection function test method and system, a terminal and a storage medium. The AEP memory Error Detection function test method comprises: using an Intel XDPto carry out simulation error injection on an AEP memory, and simulation error information is obtained; Checking the AEP error information in the BMC; And obtaining an Error Determining function testresult by verifying the consistency of the analog error information and the BMC error information. According to the method, the Error Decoding function of the AEP memory can be automatically tested,the test preciseness meets market requirements, the test method is simple, the test efficiency is high, the method does not depend on manual operation, and time and labor are saved.