37results about How to "Eliminates DC bias" patented technology

The invention discloses a closed loop flux linkage estimation method which is applied to a direct torque control system of a permanent magnet synchronous motor. Calculation is conducted through voltage signals and current signals measured in a real-time sampling mode according to a voltage method flux linkage calculation formula so that initial estimation flux linkage can be obtained, then, the initial estimation flux linkage and target control flux linkage are used as input quantity, a PI regulator is used for conducting dynamic state closed loop correction, and accordingly the target control flux linkage can be tracked ultimately. The given control flux linage controlled by direct torque is used as the amplitude of the target control flux linkage in a flux linkage estimator, the rotor position angle is obtained through a position sensor, and the angle information is dynamically compensated to be used as the angle information of the target control flux linkage. Through closed loop regulation to the voltage method flux linkage estimator, flux linkage calculation deviation caused by direct current bias and initial integration in an integration element can be effectively eliminated, the real flux linkage can be accurately estimated, and good rapidity and robustness are achieved in the actual direct torque control system.

The invention relates to a neutral point potential balance control method for a direct current side in an NPC three-level structure. A neutral point clamping type three-level inverter circuit is adopted in a photovoltaic grid-connected inverter, so that neutral point potential imbalance is prone to being caused; a software control algorithm, namely, a control method conducting hysteresis comparison on a neutral point cannot not completely eliminate low frequency fluctuation of neutral point potential; a control algorithm for injecting zero sequence voltage is relatively complex. The neutral point potential balance control method for the direct current side in the NPC three-level structure is characterized in that the difference value of one half of busbar voltage on the direct current side of a photovoltaic inverter and neutral-point voltage, the current value which needs to be injected into the neutral point in a present switching period is obtained by a PI regulator, the acting time of redundant small vectors in each sector in space vector pulse width modulation is used as a modulation method, and the allocation proportion of the redundant small vectors in one switching period is computed. The neutral point potential balance control method for the direct current side in the NPC three-level structure eliminates direct current bias and low frequency fluctuation of the neutral-point potential of a diode clamping type three-level inverter on the existing hardware condition, and neutral-point balance is achieved.

The invention discloses a parameter training algorithm for EEG signals in a multi-classifier single-channel mode, belonging to the technical field of EEG signal processing. The algorithm includes a system initialization parameter setting module, a multi-classifier training module, a compression coding module, and an indexing and the best classifier finding module. The multi-classifier training module includes a convolutional neural network, a support vector machine, a K nearest neighbor, and a feedforward neural network. The input original signal is compressed into index data according to Huffman coding mode, and the Huffman compression algorithm re-encodes the original data by extracting a common identifier. The classification accuracy of the classifier is tested according to the index data synthesized from the original data. The invention can flexibly select the optimal classification model according to different EEG signal characteristics, and the compression index algorithm makes full use of the cache for information storage, reduces hardware overhead, reduces algorithm complexity, and is convenient for subsequent algorithm expansion.

The invention belongs to the field of the bidirectional full-bridge DC/DC converter circuit control, and especially relates to a control method of bidirectional full-bridge DC/DC converter circuit magnetic biasing. The control method comprises the following steps: performing closed-loop regulation on an AC side current average value of a circuit, and adjusting the duty cycle of a circuit secondaryside according to closed-loop regulation. Through the control method disclosed by the invention, the secondary side AC side voltage is adjusted within short time through the closed-loop control on the AC side current, so that the AC side current can quickly recover to the steady-state current, the DC bias of the recovery process is reduced and then eliminated, the control method has important significance for the application of the bidirectional full-bridge DC/DC converter circuit in a power change occasion.

The invention is used to solve the problem of which the DC bias can not be simply and accurately removed from the signal processor in the receiver. It comprises the following steps: the control module supervises the state of each time slot in the logic channel, informs the idle time slot to the calculation module and disconnects the receiving signal of the signal processor in the idle time slot; according to the sampling signal outputted from the signal processor in the idle time slot, the calculation module calculates the DC bias value of the signal processor and outputs the result to the bias removing module; according to said DC bias value, the bias removing module removes the DC bias of the sampling signals from the signal processor.

The invention provides a method and a device for eliminating direct current offset. The method comprises the following steps of: calculating the Euclidean distance between every two of a plurality of quadrature phase shift keying (QPSK) modulation-based demodulation symbols; determining four sets from the plurality of demodulation symbols, wherein every set corresponds to a modulation direction of QPSK modulation; performing Euclidean distance weighted summation calculation on the four determined sets respectively and selecting a demodulation symbol which has a minimum weighted summation value and serves as a rough estimation point of QPSK modulation in every set; re-determining four sets from the plurality of demodulation symbols according to the Euclidean distance between each modulation symbol and the rough estimation point; performing Euclidean distance weighted summation calculation on the four re-determined sets respectively and selecting a demodulation symbol which has a minimum weighted summation value and serves as a precise estimation point of QPSK modulation in every set; and performing direct current offset calculation and compensation according to the precise estimation point. The demodulation performance of a system can be enhanced.

The invention discloses a method and a device for suppressing DC bias of a DAB converter, and a storage medium, and relates to the field of bidirectional active bridge DC-DC converter control. The method comprises the steps: obtaining a first external phase shift angle between two full bridges of a previous steady state in a DAB transient process, wherein the two full bridges respectively output afirst square wave and a second square wave, and a phase difference value between the first square wave and the second square wave is the first outer phase shift angle; when entering a DAB transient process, respectively changing the pulse widths of the first square wave and the second square wave; and obtaining a phase shift angle according to the first square wave and the second square wave after the pulse width is changed. According to the technical scheme provided by the invention, the main purpose is to directly change the power transmission direction and/or size by changing the pulse widths of the two square waves in the transient process after the previous steady state, thereby eliminating the direct current bias generated in the transient process under the traditional single phaseshift.

The invention discloses a DC magnetic biasing suppression method for a bidirectional full-bridge DC/DC converter, and belongs to the field of bidirectional full-bridge DC/DC converter circuit control.The suppression method comprises: calculating a required phase shift angle after power change, in a first cycle after the required power change, quantitatively changing the phase shift angle and dutycycle of a secondary side, and working at a steady phase shift angle in a second cycle after the required power change. The suppression method of the invention can quickly restore the current flowingthrough a transformer to the steady current after the power change by controlling the voltage at a secondary side AC side in one cycle, to eliminate the DC bias of the restoring process, thereby suppressing the magnetic biasing, and is of great significance for the bidirectional full-bridge DC/DC converter circuit in applications where power is changed.

The invention discloses a photodiode circuit based on background noise elimination and a laser ranging system. The photodiode circuit based on background noise elimination, provided by the invention, comprises a photodiode transconductance amplifier module, an operational amplification integrator module and a filtration module, wherein the filtration module is connected to the output end of the photodiode transconductance amplifier module and the input end of the operational amplification integrator module; and the output end of the operational amplification integrator module is connected to the input end of the photodiode transconductance amplifier module. According to the photodiode circuit disclosed by the invention, only an operational amplifier, a capacitor and a resistor can be used by the operational amplification integrator module, only a capacitor and a resistor can be used for a filter circuit, a digital element is not used in the whole circuit, extra noise is avoided, the signal to noise ratio of the circuit can be greatly improved, and the ranging range is widened. Meanwhile, because only the operational amplifier, the resistor and the capacitor are used, the cost is greatly reduced compared with the prior art, and the overall circuit area is greatly reduced, so that miniaturization of the product is facilitated.

The invention discloses a neural signal amplifier based on an MTJ device, and belongs to the technical field of biological signal processing. An amplifier is provided with two amplifier feedback loops, a pseudo resistor RH is composed of an MTJ and an MOS tube, selective amplification is conducted on signals in the frequency band lower than 300 Hz according to the amplitude of the signals, and inother words, only high-pulse neural action signals are effectively recognized and amplified. The pseudo resistor RH is formed by connecting the MTJ and the MOS tube in series, so the problem that thepositive and negative directions of MOS current are not matched is solved, the noise is lower, the distortion of the amplifier is reduced, and the interference to a neural signal is also reduced. Dueto the fact that the pseudo resistor has extremely high resistance, the capacitor C2 can achieve a very low low-frequency cut-off point without a very large area, the chip area is reduced to a certainextent; in addition, the MTJ is directly manufactured in the fourth layer of metal and the fifth layer of metal through the MEMS technology, and the occupation of the chip area is also reduced.

The invention discloses a compensation method for non-linear periodic oscillation of a permanent magnet synchronous motor (PMSM) position signal. The compensation method includes the steps of acquiring the actual angular velocity and the actual angle of a rotor of the PMSM, obtaining an ideal angular velocity according to the actual angular velocity, obtaining the angular velocity difference through the ideal angular velocity and the actual angular velocity, obtaining the angle deviation according to the angular velocity difference, and using the angle deviation to compensate the actual angleof the rotor. According to the compensation method, compensation for the non-linear periodic oscillation of position signals is realized. The compensation method is simple, the operation is convenient, and the influence of the signal nonlinear periodic oscillation on the operation of the motor can be reduced to a certain extent. The actual engineering data shows that DC bias in the three-phase current can be effectively eliminated, the motor loss is reduced, and the operation stability of the motor is guaranteed.

The invention discloses a system with dual coil assemblies for online monitoring chippings in lubricating oil. The system comprises a power module, a sensor module, a signal demodulation module and anupper computer, and is characterized in that the power module is provided with a first positive voltage output end, a negative voltage output end, a second positive voltage output end and a third positive voltage output end, the sensor module comprises a sensor driving module used for producing sensor excitation signals, a first excitation module, a first sensing module, a second excitation module and a second sensing module, and the signal demodulation module comprises a first differential amplification module, a first phase-sensitive demodulation module, a first lowpass filtering module, afirst super servo amplifier module, a second differential amplification module, a second phase-sensitive demodulation module, a second lowpass filtering module, a second super servo amplifier module,an AD sampling module and a DSP control module. The system has the advantages that the signal demodulation module has totally two paths, the detection missing rate is effectively lowered, the reliability of the detection result is improved, and the detection efficiency and detection precision are high.

The invention provides a novel universal four-phase-shift modulation method for improving the dynamic performance of a dual-active bridge. All carrier signals are kept synchronous, and the universal four-phase-shift modulation method adjusts a corresponding phase shift angle by changing parameters of a comparison register in each carrier. The control method provided by the invention is based on a traditional phase shift control method, only input and output voltages need to be measured, and an auxiliary circuit does not need to be added. Compared with other control methods, the method provided by the invention has the advantages that dynamic response is completed instantly, a transition process and inductive current detection are not needed, and the whole control process is irrelevant to an inductance value. The method provided by the invention reduces the cost and calculation complexity, achieves the quick dynamic response of the circuit, eliminates the DC bias of the inductive current in a response process, and can be suitable for all application scenes of hard switching, soft switching, unidirectional power transmission, bidirectional power transmission and the like. Meanwhile, the method is also suitable for all control methods such as single-phase-shift modulation and three-phase-shift modulation.