36results about How to "Immune to noise" patented technology

There are included a transmission modulator (103) for impulse-modulating the data to be transmitted, thereby producing a subcarrier; a subcarrier control section for controlling the subcarrier to be utilized for communication, depending on the amount and significance of information and on the propagation condition of communication; and an antenna section (101) for radiating the subcarrier signal. This structure allows selection of a subcarrier suitable for information to be transmitted and for propagation environment, and hence allows a communication to be performed which exhibits a high flexibility and a high noise immunity. Thus, there can be provided a communication apparatus that can perform a high-quality, high-stability communication exhibiting an improved interfering immunity and that performs a flexible impulse communication.

Systems and methods for detection of charge switching element failure (e.g., charge FET switching element) in a battery system, such as a battery system of an information handling system (e.g. notebook computer), in which an increase in cell voltage is employed as a criteria for detecting charge switching element failure.

The invention discloses a radar interference signal feature-level fusion identification method based on a deep convolutional neural network, and belongs to the field of radar interference signal identification. The method aims to solve the problems that feature parameters of radar interference signals at present depend on manual extraction, are easily affected by noise and have feature redundancy.The method comprises the following steps: establishing a radar interference time domain data set, extracting feature vectors from radar interference time domain data in the radar interference time domain data set in two different forms, and carrying out series fusion on the two extracted feature vectors; training a support vector machine by adopting the fused feature vectors to obtain a trained radar interference signal feature-level fusion recognition model; and inputting the collected test sample into the identification model to obtain a radar interference signal identification result. According to the method, the CNN is utilized to extract the deep features of the radar interference signals, and different radar interference signal data fusion models are designed at the feature level, so that signal identification is not affected by noise, and meanwhile, the feature redundancy phenomenon is eliminated.

The present invention relates to a DC to DC converter system (100), which comprises converter means (110) and control means (120). The switching sequence of the first, second and third switching means (S1, S2, S3) is controlled by the control means (120) in such a manner that, during the on-time of the second switching device (S2), the first current (II) that flows through the inductive storage element (L) of the converter means (110) can be indirectly measured through the first voltage (VC1) across the capacitive storage element (C1), which is being charged with a second current (12) proportional to the input voltage (Vin) of the converter means (110) and provided by the current source (CS). Thus, the on-time of the second switching means (S2) varies inversely proportional to the input voltage (Vin).

An apparatus for controlling a power converter operating in response to a modulating signal during successive switching cycles includes: (a) A signal sensor coupled with the converter and sensing an extant signal during the extant cycle. (b) A signal level predictor coupled for receiving a reference signal and establishing a predicted level for the extant switching cycle. (c) A comparer coupled with the signal sensor and the signal level predictor for presenting a first output signal when the extant signal and the predicted signal level have a first relationship and for presenting a second output signal when the extant signal and the predicted signal level have a second relationship. (d) A control unit coupled with the comparer and with the converter for interrupting presentation of the modulating signal to the converter device when the comparing unit presents a selected one of the first and second output signals.

The invention provides a seismic data processing method based on the alternating direction method and the total variation theory. The method comprises the steps of building a seismic data model based on total variation, using the classic quadratic penalty function method to convert the model into a solving problem meeting the requirement of iteration and alternating minimization, and performing iteration through the alternating direction method to get seismic data after recovery. By using the technology, seismic data can be de-noised and de-blurred (de-convoluted), the signal-to-noise ratio and resolution of seismic data can be improved effectively, and a good processing effect can be achieved even if the signal is seriously polluted by noise.

The invention discloses the measurement and control system of a quad-rotor unmanned aerial vehicle scientific research and teaching test bench. The system is formed by a hardware circuit and measurement and control software. The hardware circuit is formed by an angle encoder, a main control chip, a decoding board, a gyroscope, a sensor, a remote controller, and an attitude adjustment mechanism. Anunmanned aerial vehicle rolling signal is transmitted to rolling angle control software via the angle encoder and the decoding board, and an attitude response curve is displayed and a corresponding PID parameter is adjusted according to a curve change trend. The software is used to feed back a parameter modification instruction to the main control chip, and simultaneously, the desired posture ofa remote control instruction is compared with the real-time posture monitored by the gyroscope, and a result is used as a control input. Through the main control chip, two input instructions are operated and processed, and then a corresponding control command is output. Through the organic combination of the software and hardware, the smooth flight of the single degree of freedom of an unmanned aerial vehicle is realized. The system has guiding significance for studying the control mechanism of a rotor unmanned aerial vehicle, and has an important reference value for the development of an unmanned aerial vehicle control system.

A touch screen digitizing system comprising a touch screen unit and an analog to digital converter. The touch screen unit includes a first resistive sheet with opposing first and second terminals and a second resistive sheet with opposing third and fourth terminals. The first and second resistive sheets are alternatively energized through the first, second, third, and fourth terminals. The analog to digital converter has an input coupled to the third terminal when the first resistive sheet is energized and the second resistive sheet is not, and coupled to the first terminal when the second resistive sheet is energized and the first resistive sheet is not, a first reference terminal dynamically coupled to a first reference voltage, a second reference terminal dynamically coupled to the second reference voltage, and an output terminal providing digital output data.

A touch screen digitizing system comprising a touch screen unit and an analog to digital converter. The touch screen unit includes a first resistive sheet with opposing first and second terminals and a second resistive sheet with opposing third and fourth terminals. The first and second resistive sheets are alternatively energized through the first, second, third, and fourth terminals. The analog to digital converter has an input coupled to the third terminal when the first resistive sheet is energized and the second resistive sheet is not, and coupled to the first terminal when the second resistive sheet is energized and the first resistive sheet is not, a first reference terminal dynamically coupled to a first reference voltage, a second reference terminal dynamically coupled to the second reference voltage, and an output terminal providing digital output data.

The invention discloses a grid-connected control system and method for miniature internal combustion engine generator sets. The system comprises a grid-connected controller and a rotation speed regulator ASR, wherein the grid-connected controller is used for carrying out processing of a grid-connected synchronous control algorithm, carrying out processing of an optimum control algorithm after gridconnection, sending a regulation signal obtained via the processing to a voltage regulator AVR and the rotation speed regulator ASR, detecting and comparing whether frequencies, voltage values and voltage phases of a generator set and a mains supply are the same tor not, controlling a grid-connected switch and sending a grid-connected operation signal to the rotation speed regulator ASR; and therotation speed regulator ASR is used for carrying out processing of a traditional rotation speed PID control algorithm on the regulation signal obtained by the grid-connected controller via the processing and the acquired generator rotation speed before grid connection, carrying out processing of a variable gain control algorithm after the grid connection, and using the result obtained via the processing to control a throttle percentage. The system and method are capable of solving the problem that no controllers are designed for switching between grid connection and grid separation in the prior art, and solving the defect that the electric power waves are easy to fluctuate after grid connection.

The invention discloses a mechanical system of a scientific research and teaching test stand of a four-rotor unmanned aerial vehicle. The core of the mechanical system is that the four-rotor unmannedaerial vehicle, which can do motions by multiple degrees of freedom, is converted into the unmanned aerial vehicle test stand which can only do a rolling motion by a single degree of freedom; the mechanical system comprises the four-rotor unmanned aerial vehicle and an aluminum profile stand; the four-rotor unmanned aerial vehicle part is composed of propellers, a brush-less motor, pipe clips, a carbon fiber pipe, upper and lower central plates, a central shaft, an angle coder and a coder bracket, wherein one end of the carbon fiber pipe is fixedly connected to the central plates by virtue ofthe pipe clips and bolts, and the other end is connected to the brush-less motor by virtue of the pipe clips and bolts; the propeller is arranged on the brush-less motor; the central shaft is fixedlyconnected to the lower central plate; the central shaft is fixedly connected to the aluminum profile stand by virtue of an aligning bearing device; and the aligning bearing device is composed of a vertical bearing seat and an aligning bearing. The aluminum profile stand is defined by connecting a plurality of 30*30 aluminum profiles by virtue of angle pieces, T-shaped bolts and nuts / with the application of the test stand, the most urgent problems in researches of current multi-rotor unmanned aerial vehicles can be solved.

The invention relates to a non-intrusive load monitoring method. The method comprises the following steps: S1, constructing a load monitoring model by using a computer aided design tool of a power system or electromagnetic transient analysis containing direct current; s2, obtaining an identified load and dissimilarity measurement value and an identified wavelet high-pass filter coefficient by adopting wavelet design and a general analysis method; s3, performing cooperative training of machine learning by adopting a decision tree and a nearest neighbor classifier according to results of the step S1 and the step S2; and S4, verifying the training result in the step 3 by adopting a K-layer cross test and a Monte Carlo method, and optimizing the load monitoring model. The method is high in robustness and free of noise interference, and the influence of noise pollution on signals is reduced.

A low-current grounding fault line selection method based on MEEMD decomposition and high-frequency energy comprises the following steps: 1, decomposing and extracting a high-frequency attenuation oscillation component of a zero-sequence current signal, and obtaining energy in a time domain of the zero-sequence current signal; 2, executing line selection according to the high-frequency energy ratio of each line and all the lines; 3, carrying out simulation experiments under different fault conditions. In order to avoid modal aliasing and endpoint effects, reduce sensitivity to noise and solvethe technical problem that zero-sequence current frequency bands cannot be accurately distinguished in the prior art, the invention provides a low-current grounding fault line selection method based on MEEMD decomposition and high-frequency energy.

The invention, which belongs to the technical field of mechanical part fault diagnosis, particularly relates to a variable-working-condition bearing fault diagnosis method based on fast kurtogram anddeep residual learning. According to the method, a vibration acceleration signal is converted into a series of spectral kurtosis graphs through a rapid spectral kurtosis graph method; processing is carried out to obtain a data set comprising a training set test set and a verification set; and a deep residual error network based on a residual error block and a deep convolutional neural network is constructed, the deep residual error network is trained by using the data set to obtain a trained bearing health state classification model, and the bearing health state of a to-be-detected signal is carried out by using the bearing health state classification model. According to the invention, characteristics generated by the method, namely the spectral kurtosis graphs, are basically not interfered by noises, and the spectral kurtosis graphs of the same fault under different operation conditions have considerable similarity; in the aspect of the classifier, a deep residual learning technologyis adopted, and the robustness of the method to noise and different working conditions is further enhanced by using the strong feature self-learning characteristic of the classifier.

An apparatus for controlling a power converter operating in response to a modulating signal during successive switching cycles includes: (a) A signal sensor coupled with the converter and sensing an extant signal during the extant cycle. (b) A signal level predictor coupled for receiving a reference signal and establishing a predicted level for the extant switching cycle. (c) A comparer coupled with the signal sensor and the signal level predictor for presenting a first output signal when the extant signal and the predicted signal level have a first relationship and for presenting a second output signal when the extant signal and the predicted signal level have a second relationship. (d) A control unit coupled with the comparer and with the converter for interrupting presentation of the modulating signal to the converter device when the comparing unit presents a selected one of the first and second output signals.

The invention discloses a quiescent voltage stability margin evaluation method and system based on GBRT. The system is characterized by comprising two parts, i.e., an offline part and an online part,wherein an offline training database is established offline so as to cover as many grid structures as possible and to improve the generalization ability of GBRT; a nonlinear relationship between measured values and quiescent voltage stability margins is established based on the GBRT and optimal power flow; online evaluation is to predict the current quiescent voltage stability margin of the powersystem according to the nonlinear relationship established by the offline part and the real-time measurement condition of a current PMU, and to give necessary reminders to dispatchers when stability is insufficient. In addition, a node importance index is calculated, and the PMU is preferentially arranged at the important node, so a prediction index within an error tolerance range is obtained withlower economic cost and better economic performance is obtained. According to the invention, the advantages of mechanism modeling and data mining are combined, so real-time performance is good, and accuracy is high.

The invention discloses a real-time fluorescence imaging intelligent enhancement method and device, and the method comprises the steps: training a neural network model in advance through self-supervised learning, loading the trained neural network model to an arithmetic unit, and enabling the loading process to comprise the optimization of the neural network model; converting the image through an imaging system interface to obtain a continuous data stream, inputting the continuous data stream into an arithmetic unit, processing the continuous data stream through a neural network model trained on the arithmetic unit, and outputting the processed data in the form of the continuous data stream, wherein the data processing time sequence is processed, so that real-time data processing and synchronous display are realized. According to the method, image enhancement can be carried out in real time while the fluorescence microscope is used for shooting, so that the signal-to-noise ratio of data is improved.

The embodiment of the invention provides a determination method, device and system for a virtual mouse position. The method comprises the steps that a laser spot position of a non-visible light frequency band on a projection screen, a boundary point position of a projection image on the projection screen and a screen display resolution of a computer projecting the image onto the projection screenare acquired; and the virtual mouse position is obtained according to the laser spot position, the boundary point position of the projection image and the screen display resolution of the computer. Through the scheme, the position of a laser pen spot relative to the projection screen can be calculated quickly and accurately, noise interference from visible light is avoided, the scheme has the advantages of a small calculated quantity, rapidity and accuracy, dual display of the laser pen spot and a mouser pointer cannot occur, and visual interference on field audiences is avoided.

A bus arrival prompting system comprises a platform terminal, a bus terminal and a portable mobile terminal. The bus terminal and the portable mobile terminal are interacted with the platform terminalthrough a wireless network; the platform terminal and the portable mobile terminal are interacted to obtain numbers of a bus which a blind wants to take, and the numbers of the access bus are obtained in real time through interaction of the bus terminal; and when the bus which the blind wants to take accesses, the platform terminal timely prompts the blinds about the vehicle arrival information through wireless communication. The platform terminal, the bus terminal and the portable mobile terminal can allow the blind to obtain the number of the accessing buses at the platform so as to avoid the condition of taking the wrong bus and provide convenience for the blind.

The invention provides an audio playing method and apparatus, and belongs to the technical field of a terminal. The method comprises the steps of obtaining a first audio data stream of a to-be-playedaudio and a noise data stream of an external environment; performing spectral analysis on the first audio data stream and the noise data stream, and determining whether spectrums of the first audio data stream and the noise data stream meet a preset interference condition or not; when the spectrums of the first audio data stream and the noise data stream meet the preset interference condition, performing spectral translation on the first audio data stream to obtain a second audio data stream; and playing the second audio data stream. The first audio data stream of the to-be-played audio and the external noise data stream are subjected to the spectral analysis; and when the spectrums of the first audio data stream and the noise data stream meet the preset interference condition, the first audio data stream is subjected to the spectral translation to be played, so that the played audio is not interfered by noises, a user can hear audio contents clearly, and the audio playing validity andintelligence are improved.

The invention provides conversion which converts the flow speed of water flow into the rotational speed Omega of an impeller by a rotation impeller, and realizes the conversion from the rotational speed Omega of an input shaft to the rotational speed hOmega of an output shaft. The h value is sent to a variable speed case by utilizing a liquid level meter to measure the liquid level h and by mechanical connection so as to realize the conversion the rotational speed from Omega to hOmega. A flowmeter manufactured according to the invention has the following characteristics: no power is used in the calculating process, the flowmeter does not rely on external conditions, is not affected by external interference and thunder and light, and is not interfered by noise. As long as the selection of manufacturing material is appropriate, the manufacturing technique is reasonable and the mechanical structure is not damaged, the flowmeter can stably work for a long term.

The invention discloses a quiescent voltage stability margin evaluation method and system based on GBRT. The system is characterized by comprising two parts, i.e., an offline part and an online part,wherein an offline training database is established offline so as to cover as many grid structures as possible and to improve the generalization ability of GBRT; a nonlinear relationship between measured values and quiescent voltage stability margins is established based on the GBRT and optimal power flow; online evaluation is to predict the current quiescent voltage stability margin of the powersystem according to the nonlinear relationship established by the offline part and the real-time measurement condition of a current PMU, and to give necessary reminders to dispatchers when stability is insufficient. In addition, a node importance index is calculated, and the PMU is preferentially arranged at the important node, so a prediction index within an error tolerance range is obtained withlower economic cost and better economic performance is obtained. According to the invention, the advantages of mechanism modeling and data mining are combined, so real-time performance is good, and accuracy is high.

The invention discloses an integrated rotation-vibration sensing assembly, which is used for measuring vibration information of at least one rotating device. The rotating device is provided with at least one rotating shaft, and the integrated rotation-vibration sensing assembly comprises at least one rotation detection device for detecting the rotating speed of the rotating shaft, converting the rotating speed into rotating information and outputting the rotating information; a vibration measuring device used for outputting vibration information; and at least one processing device, used for receiving the rotation information and determining a vibration basic frequency band according to the rotation information so as to filter the vibration information and output measurement information corresponding to the vibration basic frequency band. In addition, the invention also discloses a digital vibration sensor with a rotating speed signal.

The invention discloses a residential building with rich spatial levels and avoiding from floor noise effects and belongs to the construction field. The residential building with the rich spatial levels and avoiding from the floor noise effects is characterized in that a standard layer comprises four layers; a first standard layer and a second standard layer are lower layers and a third standard layer and a fourth standard layer are upper layers; the standard layer is formed by large layer high rooms, small layer high rooms, standard layer high rooms and double standard layer high rooms; the large layer high rooms comprise living rooms and living room balconies; the small layer high rooms comprise living rooms, primary bedrooms, secondary bedrooms and bedroom balconies; the standard layer high rooms comprise rest rooms, kitchens and studies; and the double standard layer high rooms are indoor staircases. The residential building with the rich spatial levels and avoiding from the floor noise effects has the advantages of being tall in hall and eliminating the depressive feeling; avoiding from the floor noise effects; being obvious in energy-saving, noise reduction and fire prevention effect; and significantly improving the privacy.