66 results about "Flux observer" patented technology

The invention provides a method for controlling direct power of a grid-connected inverter without any non-AC voltage sensor and belongs to the field of electric energy transformation. The invention aims to solve the problem that because an initial value of an integrator in a virtual flux observer of a voltage vector on the network side is improperly selected, the grid-connected inverter cannot normally operate. The method comprises the following steps of: by transformation, obtaining components i alpha and i beta of the three-phase current value in a two-phase stationary coordinate system; by calculation, obtaining components us alpha and us beta of the inversion voltage value of the current grid-connected inverter in the two-phase stationary coordinate system; after the integration of the integrator, calculating components psi alpha and psi beta of the magnetic flux linkage of the voltage on the network side in the two-phase stationary coordinate system; calculating the instantaneous output active power p and reactive power q of the grid-connected inverter, and obtaining components usp * and usq * of the control voltage set value of the new grid-connected inverter in a two-phase rotary coordinate system; and finally obtaining the control signal of all power devices of the grid-connected inverter. The method can be used for direct power control of the grid-connected inverter.

A system (10) for controlling the torque of an induction motor (12) utilizes a flux observer (14). The flux observer (14) receives stator current inputs (16, 18) and a rotor speed estimate (80) and then outputs a rotor flux estimate (20) that provides increased motor control stability at all speeds.

An NPC (neutral point clamped) three-level inverter vector control system based on a novel flux observer provides a simplified SVPWM (space vector pulse width modulation) algorithm. A three-level space vector is decomposed into a two-level space vector; a reference voltage vector is translated; a two-level SVPWM algorithm is used to obtain action time of each basic vector and to choose a right switching status; and the three-level inverter vector control calculation is reduced. Meanwhile, in the invention, the novel flux observer is employed to calculate the rotor flux, so that a current model and a voltage model are used respectively at a low speed and at a high speed; and when a rotation speed is between the low speed and the high speed, an invented hybrid flux observer is used to calculate the rotor flux.

The invention discloses a single-winding bearingless motor torque and suspension force direct controller and a construction method. The controller comprises a reference value calculation part and an observation value calculation part, wherein the observation value calculation part comprises a winding flux observer, a suspension force observer, a torque observer and two matrix transform modules; output of the first matrix transform module is connected with input of the winding flux observer; the output of the winding flux observer is connected with the suspension force observer and the torque observer respectively; the output of the second matrix transform module is connected with the winding flux observer and the torque observer respectively; a observed winding real-time flux is applied to the torque observer, the suspension force observer and a torque and suspension force voltage reference value generator; a voltage static coordinate instruction value is generated in the torque and suspension force voltage reference value generator according to a flux amplitude instruction value, real-time winding torque component flux amplitude and phase and an actual feedback current; and the torque and rotor radial suspension force are directly controlled.

The invention relates to an online motor parameter identification method, in particular to an online rotor resistance identification method of a non-speed sensor vector control system. The method is characterized in that, during dynamic adjustment of a system, the online rotor resistance identification can be realized by using transient flux change process and by adopting an adaptive forgetting factor recursive least square algorithm in combination with an integral-improved voltage model flux observer independent from a full-order adaptive state observer; during stable operation of the system, the rotor flux is subjected to a specific harmonic transiently at a certain time interval to generate dynamic flux change, and the rotor resistance can be identified by adopting the adaptive forgetting factor recursive least square algorithm in combination with the integral-improved voltage model flux observer so as to online refresh the rotor resistance in the full-order adaptive state observer. The online motor rotor resistance identification method provided by the invention can calculate the motor parameter change so as to regulate a motor controller, thereby improving the performance of the motor controller.

A system and method for determining a rotor time constant of an AC induction machine is disclosed. During operation of the induction motor, a flux signal is injected into a rotor flux command so as to generate a time-variant rotor flux. A voltage-current flux observer determines amplitudes of rotor flux variations resulting from the time-variant rotor flux, with the amplitudes of the rotor flux variations comprising an amplitude of a rotor flux variation based on a current model of the voltage-current flux observer and an amplitude of a rotor flux variation based on a combined voltage-current model of the voltage-current flux observer. A rotor time constant of the induction motor is then estimated based on the determined amplitudes of the rotor flux variations.

The invention discloses a squirrel cage asynchronous motor equivalent circuit parameter identification method based on measurable electrical capacity and belongs to the field of on-line identification of squirrel cage asynchronous motor parameters. The squirrel cage asynchronous motor equivalent circuit parameter identification method based on measurable electrical capacity does not depend on a velocity sensor and a motor flux observer does not need to be designed. The voltage data and the current data of a stator of a motor are measured directly, a motor parameter identification model based on a least square method under a dq0 coordinate system is adopted, parameters of the stator and a rotor of an asynchronous motor equivalent circuit can be identified on line through a recursion augmentation least squares algorithm on the basis that the factors which influence identification precision are fully considered. The squirrel cage asynchronous motor equivalent circuit parameter identification method based on measurable electrical capacity is simple in algorithm, small in calculated amount and suitable for on-line identification of the motor parameters. Therefore, the purpose that the real-time parameters of the motor are accurately obtained is achieved.

The invention relates to a speed sensorless starting method for a synchronization motor. The method includes performing two processes of DC positioning before starting, applying id current in a 90-degree direction and a 0-degree direction separately, and applying iq current when starting, and performing closed-loop control on id and iq current at the same time; when the motor reaches a first switch rotation speed point, reducing the iq current gradually, keeping the id current unchanged, and comparing to obtain a difference value between an output angle and an open loop given angle of a flux observer; when the motor reaches a second switch rotation speed point, switching directly and entering into closed-loop control. According to the invention, new id axial current is introduced and a positioning effect is achieved for the motor. Smooth switching can be realized during the switching process of the motor. Besides, even a load has comparatively large fluctuation during the switching process, motor out-of-step can also be prevented. The method has excellent stability and switching quickness.

A full-order flux observer feedback matrix acquisition method and a non-speed sensor belong to the technical field of motors. The method obtains a feedback matrix G under the condition of state stability requirement of full-order flux linkage observer and stability constraint of speed estimation of asynchronous motor. The obtained feedback matrix is applied to non-speed sensor control of asynchronous motors to improve the instability region of asynchronous motors in the low-speed power generation region. By selecting the designed feedback gain matrix, the invention can improve the instabilityproblem when the asynchronous motor is in the low-speed power generation region. By selecting the parameters of the configuration feedback gain matrix, the critical frequency can be reduced and the unstable region can be reduced.

The invention relates to the technical field of asynchronous motors, and particularly relates to a DSP-based rotation speed estimation link PI parameter quantitative setting method. Firstly, parameters of a control object motor, the synchronization frequency and controller parameters are acquired; according to the original transfer function in the speed estimation link, the gain K of a first-order simplified model is calculated; according to the complex vector model of an asynchronous motor full-order flux observer, the time constant T of the first-order simplified model is calculated; according to the acquired gain K and the constant T, the first-order simplified model for a speed estimation link transfer function is obtained; and according to requirements of an open-loop bandwidth omega<c><*> and damping xi<d><*>, after the natural oscillation frequency omega<n><2> meeting the requirements is acquired, and the final PI parameter for speed estimation link is solved. Thus, the DSP-based rotation speed estimation link PI parameter quantitative setting method solves fuzziness and blindness during the traditional speed adaptive PI parameter configuration process.

The invention discloses a bearingless asynchronous motor direct suspension force control method based on model prediction. The method comprises the following steps: obtaining radial suspension force amplitude and phase xi after PID adjustment and polar coordinate conversion of difference values of displacement signals x and y and given displacement signals x* and y* through suspension control; calculating synthesis air gap flux linkage psi 1m and phase eta through torque winding flux observer and vectorial resultant of three-phase current; inputting radical suspension force amplitude and phase xi, air gap flux linkage psi and phase eta into a suspension force calculating module, carrying out rectangular coordinate conversion treatment to obtain a suspension winding flux linkage reference value, screening a difference value with a suspension winding flux linkage reference value at the next moment through a target function to obtain a flux linkage in an optimal switch state, obtaining the voltage signal of a drive inverter after SVPWM modulation of the flux linkage, and finally obtaining three-phase voltage required for a control motor so as to realize bearingless asynchronous motor direct suspension force control. The method can solve the problem that precision is insufficient and is greatly influenced by motor parameters caused by traditional suspension force indirect control.

The present invention discloses a motor driving system, and a method and a device for synchronous computation through reconstructing a phase current and a phase voltage by employing a motor driving system. The method comprises the following steps of: performing sampling of a direct current bus current, and performing reconstruction of a phase current according to the direct current bus current toobtain a three-phase current of a motor; in a reconstruction process of the three-phase current, obtaining two sampling currents in each PWM period to take as a two-phase current of the motor, and obtaining an instruction voltage or a sampling voltage; and performing synchronous conversion processing of the two sampling currents, the instruction voltage or the sampling voltage to allow a current vector and a voltage vector input to a flux observer to be in the same moment. According to the invention, the accuracy of an output angle of the flux observer can be ensured, the normal operation of the motor can be guaranteed, the calculated amount of the calculation method is small, and the calculation method is easy to implement.

A motor protection method for an automobile based on rotor magnetic steel temperature includes adopting a rotor magnetic steel temperature on-line estimation system to realize rotor magnetic steel temperature monitoring, and protecting the automobile motor according to current rotor magnetic steel temperature. The rotor magnetic steel temperature online estimation system comprises a flux observer,a temperature estimation module and a motor protection module. The input signal of the flux linkage observer is stator voltage and current, the output signal is rotor magnetic steel flux linkage signal, the rotor magnetic steel flux linkage signal is calculated according to the stator voltage and current, and the rotor magnetic steel flux linkage signal is output to the temperature estimation module; the temperature estimation module receives the output signal of the flux linkage observer, calculates the rotor magnetic steel temperature, and transmits the calculated rotor magnetic steel temperature to the motor protection module; and the motor protection module receives the rotor magnet steel temperature and takes protective measures for the automobile motor according to the rotor magnetsteel temperature.

The invention provides an asynchronous motor low-speed performance improving method and system. A stator flux linkage psi s and stator current is of a motor are adopted as state variables, an asynchronous motor state equation is constructed, state estimation and state feedback in the modern control theory are utilized, the state equation of the asynchronous motor is used for constructing a rotating speed self-adaption flux observer, and according to the Lyapunov stability law, the rotating speed self-adaption law is derived. Due to the fact that a true value of the motor rotor flux linkage cannot be obtained, a traditional rotating speed self-adaption law ignores the rotor flux linkage value, when the motor is operated at low speed, a big error can be made, and the system is not stable. Inorder to solve the problem, the rotating speed self-adaption law needs to be improved, one compensation dosage is introduced, and the self-adaption flux observer can be stably operated when the motorrotates at low speed.

The invention discloses an asynchronous motor control method without speed sensor driving. The method comprises the steps: according to the reference rotating speed of a motor, adaptively and smoothlyadjusting the reference flux linkage current and the injection voltage of a flux linkage observer and a rotating speed observer; performing virtual voltage injection on the flux linkage observer andthe rotating speed observer; therefore, according to the method provided by the invention, the rotating speed control precision of the asynchronous motor driven by the speed sensorless control systemduring medium-high rotating speed operation is ensured; when the motor operates at a low synchronous rotating speed or even a zero synchronous rotating speed, 200% of rated torque can be stably outputfor a long time, and smooth switching can be realized, so that the rotating speed of the motor can be accurately controlled within a full-speed rotating speed range, and the motor can stably operate.Moreover, the method provided by the invention can be operated on the basis of any one of the motor flux observer and the rotating speed observer constructed based on the motor model, and is wide inapplication range.

The invention discloses a bearingless permanent magnet synchronous motor controller based on a flux observer, which is used for direct torque and direct suspension force control of a bearingless permanent magnet synchronous motor. Through a torque winding flux observer, a direct torque controller and a torque winding voltage source inverter, the torque of the bearingless permanent magnet synchronous motor is directly controlled; through a suspension force winding flux observer, a direct suspension force controller, a suspension force estimation model and a suspension force winding voltage source inverter, a suspension force of the bearingless permanent magnet synchronous motor is directly controlled; and first-order delays of part of input signals are used as inputs of a neural network, and novel flux observers of a torque winding and a suspension force winding are designed by adopting the improved BP neural network, so that the flux size, the phase and the torque of the torque windingand the flux size and the phase of the suspension winding can be accurately estimated, and the stability of a control system is improved.

The invention discloses a PI parameter setting method. The PI parameter setting method is suitable for a full-order flux observer-based asynchronous motor speed-sensorless speed adjustment system. Themethod includes the following steps that: step 1, motor parameters, full-order flux observer parameters and speed loop adjustment parameters are given; step 2, time Tsf of the conversion of a speed actual speed value into a speed estimated value is determined; step 3, the total delay time of a speed loop is determined, wherein the total delay time is represented by an equation that Tomega=Tsf+pic; step 4, the proportional coefficient ksp and integral coefficient ksi of a speed loop PI controller are calculated; and step 5, the feasibility of the parameter ksp and parameter ksi are verified, if the parameter ksp and parameter ksi do not satisfy a speed loop control requirement, the proportional coefficient kp and integral coefficient ki of a rotational speed adaptive law, and a speed loopbandwidth coefficient h are adjusted, the step 2 3, and 4 are repeated again until the parameter ksp and parameter ksi meet the requirement. According to the method of the invention, the transfer function of the speed actual speed value and the speed estimated value is established; and the delay time Tsf between the actual speed value and the estimated value is obtained through a numerical value calculation method; and the PI parameters of the system speed loop are calculated on the basis of an engineering setting method.