System structure and design index of optical detection gyro stabilization system

1 System structure and design indicators
1.1 System structure
The system is mainly composed of three parts: platform components, electronic components, and display and control components.
Platform components include frame, pitch/azimuth motor, large/small field of view CCD, laser rangefinder, pitch/azimuth resolver (hereinafter referred to as resolver). The CCD camera is installed on two orthogonal inner and outer frames, and is controlled by two torque motors to scan in the directions of two degrees of freedom of heading and pitching. On the inner and outer frames, the azimuth and pitch motions are sensed by the rate gyro respectively, and the output is sent to the stabilization system regulator, and then amplified by the power to the torque motor, so that the frame rotates (scans) or stabilizes the line of sight according to the instruction.
Electronic components include system motherboard, TV tracker, motor drive and data acquisition and retention circuit. The electronic components correct and compensate the heading, level, pitch, roll and azimuth of the system according to the requirements of the system; perform temperature compensation control according to the temperature drift characteristic curve of the gyroscope; automatically sample and monitor system signals; realize the main parameters of the system Constants are dynamically displayed selectively.

Display and control components include control box and industrial computer. Mainly used to display the images taken by the CCD camera and system status information, and complete the search, lock, unlock and other operations.
1.2 Main design indicators of the system
The design indicators related to the stable axis are as follows: the angular velocity of the azimuth and pitch axis is greater than 40°/s, and the angular acceleration of the azimuth and pitch axis is greater than 60°/s2.
For a disturbance torque of no more than 3 000gcm under static conditions, the angle fluctuation is no more than 30" and there is no static difference after being stabilized.
The technical indicators of stable isolation accuracy are: under the sine disturbance with a swing of 3° and a frequency of 1 Hz, the swing of the platform's pitch channel should be less than 2′; the swing table under a sine disturbance with a swing of 2° and a frequency of 1 Hz. , The swing amplitude of the azimuth channel of the stable platform should be less than 2'.
2 Control system design
When the platform is disturbed by the movement of the carrier, if the distance of the optical axis point of action is far away, even if a small error angle deviation occurs relative to the inertial space, the tracking point outside the long distance will be out of the field of view. Therefore, the system is mainly designed for the requirement that there is no static difference after the platform angle output is stable under the disturbance torque.
In the general rate feedback scheme, the selection of PID correction in the correction link can only achieve the system angular velocity without static error, but not the angle without static error. If the system angle output is to have no static error, it is necessary to include a double integration link in the correction link. Therefore, a control method using the PII2 correction link based on the feedback of the rate gyro is designed. Because the armature inductance of the DC torque motor is usually very small, ignoring the influence of its time constant, the simplified model of the stable loop control block diagram is shown in Figure 1.
(3) Dominant pole: A pair of conjugate complex roots with complex real parts in the characteristic polynomial must be the dominant pole of the system, and α>5 should be satisfied.
Using the pole configuration method to determine the three coefficients of the (proportional-integral-double integral) correction link, we can get:
Kp=7.2, ki=245, ki2=6 500. According to the above parameters, a disturbance torque of 1000gcm is applied. The simulation result in MATLAB shows that the system adjustment time and angle static error meet the requirements, but the system overshoot is too large. Increase the open-loop gain of the system, amplify all three parameters of kp, ki, and ki2 in the system calibration link by three times, observe the closed-loop zero-pole diagram, and find that the complex real part of the conjugate complex root does not become larger after the coefficient is amplified The imaginary part becomes smaller, which can weaken the sinusoidal oscillation in the dynamic characteristics of the system and thus reduce the overshoot. At the same time, the poles on the complex real axis are farther away from the conjugate complex roots after the coefficient is amplified three times, so that the dominant pole position of the conjugate complex roots is strengthened, and the characteristics of the system are closer to the expected characteristics of the design. Under the same disturbance torque of 1 000gcm, the system adjustment time and angle static error meet the requirements. The angle output of the azimuth axis under the disturbance torque of 3 000gcm is shown in Figure 2 (the disturbance torque is applied at 60 seconds, the ordinate is in arc seconds, and the abscissa is in seconds).
In the system feedback control, the inner loop is a current loop of a torque motor, which is used to output a stable torque. The secondary inner ring is the frame inertia rate ring, and the outer ring is the position tracking ring. The feedback element of the inertial rate loop is a rate gyro, which measures the angular rate of the frame relative to the inertial space. The position tracking loop uses the resolver to complete the angle measurement. The gyro stabilization system is a torque balance system. The gyro senses the angular motion caused by the disturbance torque, and generates a control torque through the feedback loop to offset the disturbance torque, so as to achieve the goal of stability. In the control system of the gyro stabilized platform, the motor control mode adopts the torque control mode, so that the output current value of the torque (current) loop is proportional to the set value of the closed loop input voltage, which can significantly improve the inertia rate loop The control effect, thereby improving the stability accuracy.
3 System hardware design
The controller selects TI's TMS320LF2407A digital signal processor, and uses a modular main computer board, display control board, A/D board, R/S board and image tracking board. System resources have a certain degree of redundancy, which improves the reliability of the system. The overall expanded block diagram is shown in Figure 3.
3.1 Gyro signal input interface
The gyro selects the fiber optic gyroscope VG941-3AM of Russia FizopTIka Company, which is used to measure the angular rate of the load frame relative to the inertial space, and output an analog voltage signal (0～3V). The gyro signal is converted into a matching A/D chip through the signal processing circuit. Input level. This system selects two 16-bit A/D chips ADS7805U from ADI Company, which can sample at the same time, and the conversion time is 4μs, which meets the system requirements. The 16-bit signal after the A/D conversion is sent to two latches (SN74HC574), and the DSP controls the 74LS138 strobe latch to read the gyro signal.
3.2 DC torque motor servo drive interface
The motor power amplifier circuit selects IR company's motor drive chip IR2104. IR2104 is a high-voltage, high-speed power MOSFET and IGBT driver with an operating voltage of 10-20V. The system uses two IR2104 to control four N-channel IGBTs (IRF540N) to form a full bridge drive circuit to control a DC torque motor. IR2104 controls the turn-on and turn-off of the upper half-bridges Q1 and Q3 of the full-bridge drive circuit respectively through the HO output, while the LO output of IR2104 controls the turn-on and turn-off of the lower half-bridges Q2 and Q4 of the full-bridge drive circuit respectively. To achieve the purpose of controlling the motor speed and forward and reverse rotation. The motor drive interface is shown in Figure 4 (only one is drawn).
3.3 Resolver signal interface
In this system, the special RDC module 19220 of DDC company is selected to receive the excitation signal from the coarse and fine channels of the resolver. The bit4-bit11 of the fine channel are directly sent to the low-level latch after conversion, and bit1-bit3 are sent to the middle position. The low three bits of the latch, bit1-bit5 of the fine channel, and the coarse channel are combined in the MD27C256 for fine, coarse, and fine, and then sent to the middle and high bits to be latched to form a resolver of 18-bit data with a resolution of 4.94″. Single-channel spin The realization of the variable interface is shown in Figure 5.
4 System software design and function
The software design includes initialization, self-checking, control algorithm, fault handling, and the writing of various functional modules. Considering the high real-time performance of the stable module, the entire system program is designed and written in assembly language, and the servo sampling cycle is 1ms.
The system control command can be issued by the control box or the host computer. The host computer can also set parameters such as drift measurement parameters, position command parameters, etc. to complete the parameter settings required for various system monitoring. At the same time, the host computer can also receive, display and store various real-time information from the platform, including gyroscopes, resolvers, motors, etc. , Used for data processing and analysis and judgment. The software adopts modular design, which is convenient for software debugging, and has strong scalability and portability. The system software block diagram is shown as in Fig. 6.
Compared with other platforms, a major feature of this system is that it has rich functions. The system software has five states such as stable drift measurement, resolver lock, position command, target search, and photoelectric tracking. It can also complete load capacity tests and simulations. System tests such as swing test and bandwidth pre-test.
(1) resolver lock
The system applies torque to the platform according to the resolver information and controls it to the electrical zero position of the resolver. The resolver locked state lasts for 5 seconds before ending and automatically turning to the gyro stable state.
(2) Stable drift measurement
After the operator enters the sampling period and sampling time, the system enters the stable drift measurement state, and automatically compensates the gyro drift after the drift measurement is completed.
(3) Position command
The system receives the three parameters of the target's latitude, longitude, and altitude, and then reads its own roll angle, pitch angle, and azimuth angle from the carrier system, and calculates the pitch, pitch, and azimuth that it needs to rotate based on these six parameters. Azimuth angle, and then control the platform motor to rotate to the corresponding position.
(4) Target search
At this time, the joystick controls the movement of the pitch and azimuth motors. After the DSP receives the instruction code for the target search, it takes out the pitch and azimuth motor speed values given by the ground console, and then according to the joystick. The speed values of the pitch and azimuth motors control the rotation of the platform to search for targets.
(5) Photoelectric tracking
The system receives the instruction code of photoelectric tracking, selects the corresponding tracking method, and then controls the movement of the two motor platforms according to the pitch and azimuth misses from the TV tracker to track the target point.
5 System experiment and conclusion
5.1 Static stability accuracy detection
Turn on the platform to make the platform enter a stable state, and add loads to the azimuth and pitch axis respectively. Under the disturbance torque of 3000gcm, the resolver angle output of the system is shown in Figure 7 (disturbance torque is applied at 60 seconds), which meets the design requirements.
5.2 Dynamic stability accuracy detection
Turn on the platform to enter a stable state, install a double-sided reflector on the pitch stability frame, adjust the double-sided reflector, collimator and photoelectric observer to make the azimuth stabilizing axis enter the stable function state, and make the azimuth swing axis swing according to a sine wave Exercise, observe the reading of the photoelectric observer, if it is less than the required value, it meets the requirements. The measured stabilization platform's azimuth axis swing amplitude is 0.2', and the pitch axis swing amplitude is 0.3', which is much smaller than the design index and meets the technical index requirements.
This article discusses in detail the line-of-sight stabilization and high-precision system and its hardware and software design. The system adopts the classic position rate double-loop control structure, and selects DSP as the digital control system to form a high-precision line-of-sight stabilization system. Various dynamic and static performance indexes were tested, and the expected design indexes were reached.