Method of digital flux reconstruction with DC elimination

• US 4,698,577 A
• Filed: 01/16/1986
• Issued: 10/06/1987
• Est. Priority Date: 01/16/1986
• Status: Expired due to Fees

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Ritter, Allen M., D'Atre, John D., Seymour, Robert A., Kellogg, Richard A.
• Primary Examiner(s): Smith, Jr., David
• Application Number:
  US06/819,279
• Time in Patent Office: 628 Days
• Field of Search: 318/800, 318/803, 318/805, 324/158 MG, 364/481, 364/483
• US Class Current: 318/805
• CPC Class Codes: H02P 21/34 Arrangements for starting
• CPC Class Codes: H02P 23/0077 Characterised by the use of a particular software algorithm

Abstract

A method for developing current control signals from motor terminal voltage in the control of alternating current electric motors which utilize electric current control signals oriented to the phase of a flux component in a rotating field for applying current so as to control the flux and torque in the motor. The method includes developing flux signals from measurement of motor terminal voltage and integrating those signals to develop corresponding flux signals. In one form, the method eliminates DC components occuring in the resultant flux signals by locating local maxima and minima of the resultant flux signal and correcting the value of the flux integrator so as to provide maxima and minima flux signals having equal absolute values. The method also includes establishing an initial flux in the motor prior to startup which is oriented in a predetermined manner such that initial current applied to the motor can be applied so as to immediately develop torque, if required. The method further provides for initializing the flux integrators and associated logic such that discontinuities attributable to events such as frequency reversals are avoided.

Claims

1. A method for eliminating direct current components of a flux vector signal representative of a rotating flux field in an alternating current polyphase dynamoelectric machine, the machine having at least one electrical power terminal for each phase, the method comprising the steps of:

measuring machine terminal voltage at each machine power terminal;

converting the measured machine terminal voltage to equivalent first and second two-phase voltage representative signals;

integrating each of the first and second signals for generating corresponding flux vector component representative signals, each of the flux vector representative signals varying sinusoidally;

detecting consecutive maximum and minimum peaks of each of the sinusoidal flux vector representative signals;

developing a signal representative of a direct current (DC) component of each of the flux vector representative signals by determination of a voltage midpoint between the detected consecutive maximum and minimum peaks; and

subtracting the corresponding direct current component signal from a respective one of each of the first and second signals for eliminating any direct current component in the flux vector signals.

2. Method of claim 1 further comprising the steps of:

integrating each of the DC component representative signals to generate corresponding long term correction signals; and

subtracting each of the long term correction signals from corresponding ones of the first and second voltage representative signals.

3. Method of claim 2 further comprising the steps of generating a periodic waveform having no DC component when the dynamoelectric machine is not energized, and using the periodic waveform signal added to the measured voltage at the machine terminals for establishing an initial value of long term correction signal for when the machine is energized.

4. The method of claim 1 and including the steps of:

applying a DC current to the machine terminals for a predetermined time period when the machine is not running; and

initializing the flux vector representative signals in amplitude and phase to match the values corresponding to the flux induced by the impressed DC current.

5. In a system for controlling a polyphase alternating current electric motor using an electric current control signal oriented to a computed rotor flux vector, a method for developing the current control signal from motor terminal voltage comprising the steps of:

sensing polyphase motor terminal voltage and converting the sensed voltage to a two-phase representation thereof;

converting each phase of the two-phase representation to a corresponding pulse train having a variable pulse repetition rate proportional to the amplitude of the sensed motor voltage;

counting the pulses in the pulse train;

sampling the counted pulses at periodic intervals;

obtaining a differentiated value by determining the difference in count value between adjacent samples;

integrating the differentiated value to obtain a signal representative of a corresponding component of flux in the motor;

determining sequential maxima and minima of each flux component signal;

comparing the value of each pair of sequential maxima and minima for obtaining an error value; and

modifying the flux signal by a value proportional to the error value.

6. The method of claim 5 and including:

determining the direction of rotation of a vector representative of the rotor flux;

sensing a change in the direction of rotation of the rotor flux vector;

identifying the sense of the next peak to be detected based upon the present flux vector angular orientation and direction of rotation;

re-initializing the peak detector to detect the next identified peak to thereby inhibit generation of instantaneous errors at reversal.

7. The method of claim 5 and including the step of:

applying a direct current to the motor prior to start-up for establishing an initial rotor flux having a predetermined orientation.

8. The method of claim 7 including the step of:

initializing the flux integrator whereby alternating current applied to the motor is initiated in correct amplitude and phase orientation with respect to the initial rotor flux.

9. The method of claim 8 and including:

applying a simulated flux signal to the flux integrator for establishing initial conditions which correct for initial DC offset.